University of Southern California Dissertations and Theses

The data dilemma: sensemaking and cultures of research in the media industries

(USC Thesis Other)

The data dilemma: sensemaking and cultures of research in the media industries

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE DATA DILEMMA:

SENSEMAKING AND CULTURES OF RESEARCH IN THE MEDIA INDUSTRIES

by

Natalie Jonckheere

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)

August 2022

Copyright 2022 Natalie Jonckheere

ii
Acknowledgements
This dissertation has my name on it, but it would have been impossible for me to finish it
without the help of several people.
First, I thank my dissertation committee members: Professor Henry Jenkins and my co-
chairs Professors Patricia Riley and Dmitri Williams. Henry, thank you for your support and for
pushing me to learn more about audience studies, which I hope will make a bigger appearance in
next stages of this project. Dmitri, thank you for guiding me for the first half (plus) of my
dissertation studies and for pushing me on several points; your feedback made this draft better
and will make the next stage even better. Patti, thank you for guiding me during the last stages of
the project and for introducing me to the world of organizational communication; your advice
has turned this project into what it is.
I would also like to thank Professors Nitin Govil and Christopher Smith for their input
during the early stages of the project as members of my qualifying exam committee. Your advice
and feedback on my prospectus gave my project a solid footing.
Next, big thank to all my industry contacts who made this possible. Thank you to the
Advertising Research Foundation and to Cynopsis for allowing me to do research at your events.
And a massive thanks to all the wonderful people at my case study organizations. You know who
you are and know the project would not have been possible without you. Thanks also to my
former colleagues at NBCUniversal, as the early ideas for this project were hatched when I was
working with you.
Finally, a huge thank you to my wonderful family and friends, both old and new. You
have supported me in so many ways, and I simply would not have crossed the finish line without
you.
iii
Table of Contents
Acknowledgements ......................................................................................................................... ii
List of Tables ................................................................................................................................. iv
Abstract ............................................................................................................................................v
Introduction ......................................................................................................................................1
Chapter 1: Contextualizing Media Research and Data Work ........................................................20
Chapter 2: Disruptive Technology and Audience Evolution .........................................................38
Chapter 3: Culture and Organizing in Times of Change ...............................................................52
Chapter 4: Sensemaking Data Problems ........................................................................................82
Chapter 5: Sensemaking Digital Divides .....................................................................................103
Conclusion ...................................................................................................................................119
References ....................................................................................................................................127
Appendix: Interview Guide..........................................................................................................143

iv
List of Tables
Table 1: Participant Breakdown.....................................................................................................16
Table 2: Typology of Data Problems .............................................................................................84
Table 3: Typology of Data Problem Triggers ................................................................................84

v
Abstract
In the past several years, many changes have rocked the media industries. These include the rise
of streaming and subsequent decline of audience measurement firm Nielsen’s dominance,
announcements of the upcoming end of third-party cookies, and the COVID-19 pandemic. This
study sought to understand these changes through the eyes of research and data workers, who
collect, analyze, and/or package data within media organizations. The project had two
overarching goals: to understand how research and data workers are finding new ways of
organizing and solving data problems in creative and innovative ways, and in turn, how these
practices are impacting the industries. This study was conducted through fieldwork at two media
industry conferences and interviews with research and data workers at four media organizations:
two data providers and two media publishers. Results are outlined as follows. This study
uncovered how the media industries still cling to legacy thinking, particularly related to data use,
like the role of Nielsen in audience measurement and the role of third-party cookies in targeted
ads. This study also discussed how COVID-19 has created new opportunities for organizations to
fix and maintain cultures that stimulate creativity and innovation; these cultures are necessary
because new data streams create new opportunities to solve data problems. Finally, this study
discussed how changes on the horizon, including the end of third-party cookies and the growing
centrality of data science to the industries, will perpetuate digital divides. This study therefore
has implications for understanding organizational change and how the media industries will
adapt to the post-Nielsen, post-cookie, and post-pandemic futures.
Keywords: media industries, creativity and innovation, organizational change,
organizational culture, data
1
Introduction
On Wednesday, September 1, 2021, the United States television industry received major
news. The Media Rating Council, which oversees media measurement and data systems,
removed its accreditation for Nielsen’s national television ratings (Steinberg, 2021a). It was the
death knell for an ossified system, yet trouble had been brewing for some time. Streaming had
been around for almost 15 years already, beginning with experiments by CBS and ABC to
stream programming and later growing with the launch of Hulu. Netflix’s release of House of
Cards in 2013 sent the streaming revolution into overdrive (Smith & Telang, 2016). Eight years
later, the industry was in the middle of what we still call “the streaming wars,” in which every
major media company has launched its own streaming service with a plethora of original and
library content.
Amidst these battles on the content front, another battle had been brewing on the
measurement front. Time shifting technologies like the DVR and streaming—and audiences’
subsequent adoption of them—meant that the traditional ways of collecting television ratings, so
carefully implemented by Nielsen, needed to adapt. For its part, Nielsen tried. It attempted an
ambitious project called Total Content Ratings, which was to marry traditional and digital
viewing measurement into one holistic metric (Nielsen, n.d.). That project, however, was
plagued by data quality issues, and its release was ultimately delayed (Poggi, 2016, 2017). Years
later, major media conglomerates like NBCUniversal and WarnerMedia were taking
measurement into their own hands by devising ways to track their audiences (Steinberg, 2021b,
2022). They felt that it was the measurement giant’s job to lead the industry to a new currency—
a new metric to both understand audiences and transact these audiences in advertising deals—but
grew weary waiting.
2
As a result of Nielsen’s slow moving, it has a number of competitors in the measurement
space. Comscore is Nielsen’s primary competitor and specializes in digital measurement, but
other measurement companies like TVSquared, TVision, 605, and iSpot.tv are touting their own
effective measurement methods to compete with Nielsen. Furthermore, Nielsen competes with
multichannel video programming distributors like cable companies, which have proprietary data
on their subscribers (Prescott, 2014). To quell concerns of its impending irrelevance amidst the
competition, Nielsen announced a new service called Nielsen ONE in 2020 (Nielsen, 2020). Not
to give up on its fight against Nielsen, Comscore announced plans for a similar service called
Comscore Everywhere (Schiff, 2022). As a result, while Nielsen is battling media publishers, it
is also battling several measurement and data upstarts who wish to unseat it. As the television
industry undergoes the streaming wars, the measurement wars are heating up.
The measurement wars are all about how the industry is trying to play catch up to
changes in technology and audience. Without Nielsen ratings to anchor understandings of
audiences, there is no single source of information for the television industry; rather, there are
multiple imperfect metrics. Similarly, in the digital advertising industry, technology companies
like Apple and Google have decided to block third-party cookies in the coming years. These
decisions will limit the amount of data shared with other players in the advertising ecosystem.
Much like what is happening with Nielsen in the television industry, these changes will impact
the types of data that are used and how research and data workers use them. Alongside these
industrial changes, the COVID-19 pandemic has led to many changes in consumer behavior,
including media consumption, further creating an environment in which research and data work
must keep pace with audience behavior.
3
These changes have unleashed new data dilemmas for the individuals within the media
industries who have to use these imperfect and incomplete data. These individuals undertake a
sensemaking process to understand these data which in turn has created new ways of organizing
and new ways of using data. These processes are the focus of this dissertation. Amid industrial
change and increasing datafication—all of which will be amplified in the post-Nielsen, post-
cookie, and post-pandemic futures—it asks these questions: how are research and data workers
finding new ways of organizing and solving data problems in creative and innovative ways, and
in turn, how are these practices impacting the industries? The answers are told through the eyes
of those tasked with working with audience data in this industry: individuals who work in
research, measurement, insights, analytics, and data science, whom I dub “research and data
workers.” These individuals have an important vantage point to understanding these changes
because they are on the ground working to use these imperfect and incomplete data.
The rest of this introduction is structured as follows. To understand the story of the data
dilemma, I begin by explaining how metrics are used in the media industries, focusing on the
role of traditional Nielsen metrics and what has happened as they lose their luster. I then
introduce key conceptual frameworks that will help us understand these problems: big data and
datafication, sensemaking, creativity and innovation, culture, and media work. I then explain the
method I used to examine these issues and provide a chapter overview of the rest of the
dissertation.
Background: The Role of Metrics in the Media Industries
Traditional metrics used in the media industries, including linear television ratings, have
helped the industry understand its audience. What happens, then, when there are multiple metrics
that each tell a different story about the audience? This discussion focuses on the television and
4
ratings industries of the United States, but ratings providers in other areas of the world are facing
similar challenges (Bourdon & Méadel, 2014).
In the United States, ratings grew out of the expectations of the commercial radio
industry. In the 1920s, radio networks were releasing their own data, but advertisers found that
they could not trust what the networks were saying, so they called upon the industry to create
standardized measurement, which would be syndicated or third-party ratings (Buzzard, 2012).
These standardized ratings began as telephone surveys and later moved to diary methods, in
which audiences were asked to write down what programs they listened to; diaries held over into
the creation of television ratings (Buzzard, 2012). In the 1980s the PeopleMeter, a device that
tracked television viewing without people needing to remember what they watched, was created
(Buzzard, 2002). The key aspect of these ratings methods—phone surveys, diaries, and meters—
is that they rely on a sample of the national population and use those people’s media
consumption to project out to create ratings. They are as sound as the sample used and never
perfect. As a result, a common view is that the industry in effect creates the audience (Ang,
1991; Ettema & Whitney, 1994; Hartley, 1987). More specifically, by deciding what is counted,
the ratings industry governs what is included in its audience, meaning that there is a difference
between the measured audience and the actual audience (Napoli, 2003), wherein the measured
audience fills in for the actual audience, which is unknowable.
Some within the ratings industry have moved beyond those traditional ratings, which
relied on sampling. As a result, many newer measurement methods use so-called “big data”
(Kelly, 2019). This is not without issues, however, as big data have certain biases (boyd &
Crawford, 2012). Specifically with regards to television audience measurement, big data analysis
techniques are often biased towards real-time insights, which privileges live and event
programming, create an overabundance of data, and create new data divides (Kelly, 2019).
5
Usually, big data measurement systems use proprietary data, or data that are available to
organizations about their own users. As a result, those with access to more data have more
power, which creates a data divide. Those without data are effectively in the dark. This shows
why it is now so difficult to create standardized measurement systems across media platforms,
especially now when there are a mix of commercial and subscription linear and streaming
providers, many of whom have proprietary data. Nevertheless, the industry is calling for
standardized measurement so that everyone can have the same information and be on a level
playing field, much like they were in the traditional broadcasting environment.
This is important because traditional Nielsen television ratings are often considered
“market information regimes,” or metrics that help players within a field understand itself
(Anand & Peterson, 2000). In a study of the music industry, Anand and Peterson introduced the
term “market information regime” to demonstrate that the introduction of new sales figures to the
music industry revealed that country music was much more popular than thought, which then
changed how the music industry understood and treated the genre. In the television industry, one
similar example is the introduction of the Local PeopleMeter in 1999, a move that shifted
television measurement in local television markets from diaries, which relied on audience
members’ recollections of what they watched, to meters, which automatically recorded what they
watched (Napoli & Andrews, 2008). The result was that broadcast ratings declined, while cable
ratings shot up, likely because the diary method favored broadcast because people remembered
to record this more (Napoli & Andrews, 2008). More recently in the television industry, social
television analytics, which measure the buzz surrounding television shows on social media, show
that certain genres, like special events, are more popular than what linear television ratings report
about popular genres (Kosterich, 2016). As Napoli and Andrews (2008) explain, the Local
PeopleMeter met resistance because it disrupted the prevailing narrative that broadcast retained
6
dominance over cable. Because of this potential resistance, we see that despite the industry’s
calls for Nielsen to get with the times, there may not be as much of a commercial imperative to
change measurement, as it might give certain players in the industry news they do not want to
hear. This tension forms the background for some of the issues this dissertation will unpack,
including creativity and innovation in research and data work.
Conceptual Frameworks
The above section set the scene for the rest of this dissertation. In this section, I introduce
the conceptual frameworks that will help us understand the current situation.
Big Data and Datafication
Big data in the media industries is best thought of with regards to the four Vs: volume,
velocity, variety, and value (Stone, 2014). In other words, there has to be a lot of it, it can be
analyzed quickly, it is structured in various ways, and it has to serve an actionable function
(Stone, 2014). Big data as a concept is not a panacea, however. boyd and Crawford’s (2012)
proposed six provocations about big data, outlined as follows: big data change how knowledge
operates, they are likely biased and inaccurate, they are not necessarily better than “small” data,
they rely on their context for meaning, they are not always ethical even if available, and they
create digital divides due to unbalanced access. Despite this, governments and industries have
moved towards increased datafication, in which the collection of data on human behavior is
increasingly normalized (van Dijck, 2014). Datafiction runs the risk of turning into dataism, in
which great trust is placed on data (van Dijck, 2014). However, big data are always subjective
and should not be taken at face value, an idea that the field of critical data studies has furthered
(Iliadis & Russo, 2016). This calls into question whether data ought to be trusted. As we shall
7
see, many research and data workers understand that their data are imperfect, and they reach that
conclusion through the process of sensemaking.
Sensemaking
Sensemaking is the process by which people work through their understandings of what
they know. It is a social process that helps understand how individuals and groups make
meanings and participate in creating their environments (Weick, 1995). The key idea is that
sensemaking requires not just interpretation of one’s state and surroundings but also action to
impact them (Brown, Colville, & Pye, 2015; Maitlis & Christianson, 2014; Weick, 1995).
Through action, sensemaking creates new ways of organizing, which in turn create organization
(Brown et al., 2015).
Sensemaking occurs when there is a particular event that shakes someone out of their
routine, often when a situation is ambiguous or uncertain (Weick, 1995; Weick, Sutcliffe, &
Obstfeld, 2005). Sensemaking thus begins with noticing that something is not as it should be and
then functions to organize this ambiguity or uncertainty (Weick et al., 2005). In Weick’s (1995)
formulation, sensemaking has seven properties: it is “grounded in identity construction,
retrospective, enactive of sensible environments, social, ongoing, focused on and by extracted
cues, and driven by plausibility rather than accuracy” (p. 17). After confronting an event that is
different, an individual undertaking sensemaking will bracket the event for further thought, label
it to understand it, and then move forward to address it (Weick et al., 2005). Analyzing narratives
and metaphors can help understand the sensemaking process (Hill & Levenhagen, 1995; Landau
& Drori, 2008).
While Weick (1995) argues that sensemaking is retrospective, making sense of past
events to impact the present, others have begun to explore the idea of prospective or future-
8
oriented sensemaking. In Gephart, Topal, and Zhang’s (2011) study of future-oriented
sensemaking that occurred through public hearings about local utilities, the public hearings were
discovered to be a site where sensemaking and legitimation happened; experts were gathered to
produce a plan that was then enacted. In this way, future-oriented sensemaking, much like
retrospective sensemaking, creates the future.
Due to sensemaking’s role in creating action, sensemaking is useful to study
organizational crisis and change (Maitlis & Sonenshein, 2010) and creativity and innovation
(Drazin, Glynn, & Kazanjian, 1999; Hill & Levenhagen, 1995; Martin-Rios, 2016), which are
both relevant to this study due to the intense amount of change media organizations have
undergone, which have necessitated creativity and innovation. While sensemaking usually helps
organize chaos, rather than creating it, sensemaking enables change, creativity, and innovation
by creating new meanings, which allow for new ways of organizing (Maitlis & Christianson,
2014). In an increasingly complex and fast-paced atmosphere, many people undertake a process
of sensemaking that has been called simplexity: “a fusion of complexity of thought with
simplicity of action,” which underscores how important it is to balance thought with action
(Colville, Brown, & Pye, 2012, p. 5).
The modern media industries are one such fast-paced atmosphere, and few studies have
looked at sensemaking in the media industries. Anand and Peterson’s (2000) study explained
how metrics that are available to everyone in a field become market information regimes, which
help people make sense of—and therefore create—the field; their source of information was
from the commercial music industry. Recent scholarship has begun to look at sensemaking in
public media organizations, television news production, and film crews, for example (Evans,
2018; Meziani & Cabantous, 2020; Patriotta & Gruber, 2015). This study examines research and
9
data work within the media industries to try to figure out how people are making sense of their
work as creative and innovative in response to change.
Creativity and Innovation
While it is possible to view both creativity and innovation as distinct processes, one
simple way to view them is as stages of a process; in this view, creativity is “idea generation,”
while innovation is “idea implementation” (Anderson, Potočnik, & Zhou, 2014). Others,
however, see creativity as the implementation step (Lewis, 2014). Furthermore, while the above
definition characterizes innovation as a process itself, innovation can also be an outcome,
whether that outcome is a product, business model, or some other outcome. The following
sections unpack these differences and explore different ways to view creativity and innovation.
Creativity
One definition of creativity is the act of engaging in something creative, whether or not
there is a creative outcome (Drazin et al., 1999). In general, three factors contribute to creativity:
creative or imaginative thinking, expertise, and motivation, which can either be extrinsic or
intrinsic (Amabile, 1998). This implies an emphasis on individual factors, yet there are team or
organizational levels to this as well (Amabile, 1998; Drazin et al., 1999). In fact, many studies
assume that creativity is simply an outcome and occurs at the small group or project level, but it
may be generative to imagine creativity as a process that occurs at multiple levels of an
organization (Drazin et al., 1999).
Innovation
A key distinction in understanding innovation is the difference between innovation as a
process and innovation as an outcome. The definition cited above—in which innovation is “idea
implementation”—builds on the idea that innovation is part of a process that begins with
10
creativity (Anderson et al., 2014). In general, process innovation is concerned with processes
within an organization, including increasing efficiency (Damanpour, Walker, & Avellaneda,
2009). A specific type of process innovation is the technological or technical process innovation,
which introduces new technology to increase operations (Damanpour et al., 2009).
In general, innovation as an outcome is more developed in the literature and is primarily
concerned with what the innovation is (Crossan & Apaydin, 2010). The different ways of
looking at innovation as an outcome include its form, its magnitude, its referent, its type, and its
nature (Crossan & Apaydin, 2010). Understanding innovation by its form or type is primarily
concerned with the tangible outcome of the innovation, such as a new product or service
(Crossan & Apaydin, 2010). Innovation by magnitude is usually thought of with regards to
incremental change, which perfects existing practices, or radical change, which introduces new
practices (Crossan & Apaydin, 2010). Finally, because innovation is often thought of as
something “new,” the referent of a product innovation refers to who decides what is new; this
could be the firm or the market (Crossan & Apaydin, 2010).
Using the typologies above, innovation as applied to media organizations and industries
is usually thought of as an outcome and is often incremental or sustaining (Krumsvik, Milan, Ní
Bhroin, & Storsul, 2019). Innovation in audience measurement and data use is more difficult to
define. One of the few exceptions is Buzzard’s (2002) study of technological innovation and
subsequent diffusion of the PeopleMeter in the period from the mid-1980s through the 1990s.
Napoli and Andrews’s (2008) study of diffusion and adoption of two new ratings systems, one
for books sales and one for television, contends that innovation in audience measurement
systems, while new technologies, are still incremental rather than radical because they do not
overhaul the entire measurement system. Furthermore, the authors argue that innovation in
11
audience measurement should not primarily be considered a process, product, or service
innovation; instead, because their ultimate goal is to provide the market with information, they
are better understood as market information regimes. Examples from the study demonstrate how
the two measurement innovations highlighted changed the way the market views audiences.
Culture
This dissertation draws on Eisenberg and Riley’s (2001) typology of frameworks to view
organizational culture, which focuses on the communication dimensions of culture. Here, I
introduce the three themes I employ: culture as text, culture as climate, and culture as
effectiveness. One relevant dimension of culture as text is treating spoken discourse as culture
(Eisenberg & Riley, 2001). Culture as climate refers to how culture affects the overall
atmosphere of an organization, often through its attributes (Eisenberg & Riley, 2001). The
communication dimensions of climate—support, trust, and openness to new ideas—are
especially relevant (Eisenberg & Riley, 2001). Culture as effectiveness refers to the ways an
organization’s values or practices contribute to its accomplishments and how to manage those
values and practices (Eisenberg & Riley, 2001). In particular, this dissertation is interested in
cultures that foster creativity and innovation, which are often those that match the
communication dimensions of climate, such as support, trust, and openness to new ideas
(McLean, 2005; West & Sacramento, 2012).
Approaches to Media Industries and Media Work
So far, in describing the conceptual frameworks above, I have explained how they related
to the media industries. In this section, I introduce the conceptual and methodological
frameworks of media industries studies, which coalesced as a distinct field relatively recently
(Holt & Perren, 2009). The media industries tradition generally emphasizes qualitative
12
approaches, including “top-down” industrial approaches and “bottom-up” ethnographic
approaches (Holt & Perren, 2009). It also draws on a number of existing frameworks, including
an emphasis on the intersection of creativity and commerce found in the cultural and creative
industries tradition (Hartley, 2005; Hesmondhalgh, 2019; Holt & Perren, 2009).
Specifically, I follow frameworks for two strands within the media industries tradition:
critical media industries studies and studies of media work. Critical media industries studies
focuses on midlevel fieldwork and the so-called “helicopter approach,” which is between macro-
level industrial analysis but slightly above micro-level ethnography (Havens, Lotz, & Tinic,
2009). Gitlin’s (1983/2000) Inside Prime Time is a notable example of a study that uses the
helicopter approach (Havens et al., 2009). This follows earlier studies like those of Epstein
(1973) and Gans (1979), which took deep dives into the production of news, both television and
print. Critical media industries approaches also emphasize human agency (Havens et al., 2009),
and this project centers the agency of media research and data workers.
The media work approach follows this emphasis on human agency by exploring the work
of making media. In Deuze’s (2007) definition, “Media work can be seen as a particular (and
popular) set of professional values and practices within a wider context of culture production” (p.
54). A few of the main trends in media work are that they gather in urban areas, involve risk,
usually require some level collaboration and project-based teams, and deal with technology and
information management (Deuze, 2007). One of the key peculiarities of this is that media
workers face increasing precarity, even as average consumers use media more and more (Deuze
& Steward, 2011). Some strands of media work scholarship have begun to explore specific
professions. For example, Roussel’s (2017) study of Hollywood agents draws on many of the
ideas from the media work traditions, including the emphasis on qualitative methods. By
13
focusing on Hollywood agents, Roussel (2017) exposes an area that had been under-researched
and under-theorized before, a role within the system that fits between the creative side and the
commerce side. Studies like this are important because they show that media work and media
professions extend beyond that which people commonly think of as “media work”: the work of
the creative professionals only. In fact, corporate players like agents play a vital role in the
production process in Hollywood (Roussel, 2017). This study does the same for research and
data work by explaining its role in the industry.
Method
For this study, I employed qualitative methods: virtual fieldwork at two media
measurement conferences and semi-structured interviews with research and data workers,
including individuals at four media organizations that serve as the primary case studies for this
project. Due to challenges and limitations stemming from the COVID-19 pandemic, I was unable
to do in-person fieldwork, interviews, or participant observation. However, I was able to bring
my own background and knowledge to my research, having worked in digital advertising sales
research at NBCUniversal between 2014 and 2017. In those years, I saw a lot of the very issues I
discuss here and attended some of the same events I included in my fieldwork. Where prudent, I
draw from these experiences. Nevertheless, the bulk of this project stems from the online
fieldwork and remote interviews I conducted for my research.
Virtual fieldwork took place at the Advertising Research Foundation’s
AUDIENCExSCIENCE Conference in September 2020 and the Cynopsis Measurement & Data
Conference in June 2021.
1
For both, I attended panels, workshops, and presentations live online

1
I thank the Advertising Research Foundation and Cynopsis for granting me the opportunity to conduct
research at their events.
14
and took field notes about the experience. For the ARF conference, I also attended extra events
asynchronously and interacted with participants using the conference’s networking platform.
Interviews were conducted with individuals at four media organizations: two data
organizations and two media publishers. Below, I provide additional information about the
interview procedure, explain who the participants are, and create an outline of the four
organizations. Each organization has been given a pseudonym.
Interview Procedure
I made initial contact with the participating organizations through industry conferences I
attended or from cold calling the companies. After a brief conversation about the goals of the
project, the contact person at each organization made arrangements with their organization,
which included getting authorization from the relevant office, which might be Human Resources,
the Office of the President, etc. This individual then recommended other participants, who were
told that participation was entirely voluntary. Each participant received an information sheet
about the study ahead of their interview.
Interviews were conducted remotely March through May 2021, using Zoom or another
video conferencing system. While interviews were conducted with video, only audio was
recorded with consent from each participant. I opened with questions about each participant’s
work background and current position. Because the interviews were semi-structured, there was a
set of common topics including organizational and industrial change, challenges and disruptions
to work, examples of creative and/or innovative projects they had worked on, descriptions of
creativity and innovation, and issues related to data and audiences. Other issues were added as
needed to explore responses to the standard questions. See the Appendix for an interview guide.
The analysis of data problems in Chapter 4 draws from both descriptions of projects they thought
15
were creative or innovative and from their explicit definitions of creativity and innovation, which
I requested after they described projects they believed were most relevant.
Interview Participants
Participants were eligible to be interviewed if they worked in or managed media research,
insights, measurement, analytics, or data science at one of the participating organizations. As a
shorthand, when discussing this group as a whole in this paper, I use the term “research and data
workers,” but given the range of experience presented here, these workers are broken out by type
as follows for the purpose of analysis:
● Research and Insights: Individuals involved in pulling data, packaging it, and creating
key “takeaways” for the organization. They will typically use data from research vendors
like Nielsen or Comscore or from proprietary tools.
● Analytics and Data Science: Individuals more “on the ground” working with data. Their
work will typically involve Tableau, SQL, R, or advanced programming.
● Product and Measurement: Those individuals involved in conceiving and developing
measurement systems and products.
Interviewees fit within a wide range of experience, from data analyst to CEO. For purposes of
categorization, they are broken out in the data as follows:
● Frontline Employees: Analysts, senior analysts, and analytics leads, who typically do
not manage others.
● Middle Management: Managers and directors
● Upper Management: Vice presidents, senior vice presidents, and C-suite executives
One interview was conducted with each of the 22 participants. Interviews ranged from 40 to 60
minutes each, for an average of 53 minutes. Participant breakdown is below:
16
Table 1
Participant Breakdown
Organization Type Total
Participants
Participant
Breakdown
by Level
Participant
Breakdown by
Function
Bullseye
Platforms
Data 7 4 frontline, 2
middle
management,
1 upper
management
2 research and
insights,
5 analytics and
data science
Connected
Media Research
Data 6 1 frontline, 3
middle
management,
2 upper
management
3 product and
measurement, 3
analytics and data
science
USA Media Publisher 2 2 middle
management
2 research and
insights
Ballast
Communications
Publisher 7 5 middle
management,
2 upper
management
7 research and
insights

Data Analysis
I transcribed the audio of the interviews and conducted qualitative analysis (Miles,
Huberman, & Saldaña, 2020) on the interview transcripts and my field notes from the
conferences into NVivo. For items related to culture, I used in vivo codes. For other items related
to big data and digital divides, I used codes inspired by boyd and Crawford’s (2012) six
provocations for big data. For items related to creativity and innovation, I started with various
typologies of creativity and innovation (Crossan & Apaydin, 2010; Krumsvik et al., 2019) and
17
typologies of organizational practices that allow for creativity and innovation (McLean, 2005;
West & Sacramento, 2012).
Overview of the Four Cases
Bullseye Platforms
Bullseye Platforms is an advertising technology organization with offices across the
United States and a handful of locations abroad. It has under 1,000 employees and prides itself
on being a smart and innovative company. I spoke to seven participants from offices across the
east coast, Midwest, and west coast of the United States. They were all part of one large team
that had two branches: an analytics group and a group that specialized in one of Bullseye’s
proprietary research tools.
Connected Media Research
Connected Media Research is an analytics and measurement organization that is one of
several attempting to fill a hole in the measurement space that incumbent Nielsen is not filling.
The organization has under 100 employees and is based on the east coast of the United States,
with some employees abroad. I spoke to six participants who were based on the east coast of the
United States; one was the CEO, and the others were part of one large team that specialized in
analytics and product development.
USA Media
USA Media is a broadcast media organization based in the United States that has
television and digital offerings and that works with local stations across the country. It has a few
thousand employees. I spoke to two members of the business intelligence team based on the east
coast of the United States.

18
Ballast Communications
Ballast Communications is a legacy media publisher with assets covering film, linear
television, and streaming. It has over 10,000 employees across the globe. Despite being a legacy
media conglomerate, it has embraced streaming as part of its business plan. I spoke to seven
participants from an advertising sales research team based on the east coast of the United States.
I spoke to the head of the team and members of three groups within that team that had different
specialties within advertising sales research.
Chapter Overview
This dissertation is structured as follows. Chapter 1 gives an overview of the research and
data world in the television and digital media industries. It describes the culture of these
industries, the culture of research and data workers, and the role that these workers play in these
industries. Chapter 2 chronicles the changes in technology and audience behavior that have
created the period of flux leading to the measurement wars. Chapter 3 examines culture in times
of change. It is organized around four mini case studies, one for each participating organization.
It chronicles how participants described each organization’s culture and their culture of their
team within the organization. It then describes changes that have occurred in the organization
and how these changes are creating new opportunities to improve culture through new ways of
organizing. Chapter 4 involves sensemaking data problems. It introduces a typology of new data
problems that commonly arise in research and data work and explain how they are moments of
sensemaking that trigger creative and innovative solutions. Chapter 5 takes a future-oriented
approach to explain how research and data workers make sense of upcoming digital divides,
including third-party cookie deprecation and the growing importance of data science and
analytics to the television and digital media industries. Finally, the conclusion offers key
19
takeaways about the role of research and data work in the media industries, notes limitations and
future directions, and anticipates the future of research and data work.

20
Chapter 1: Contextualizing Media Research and Data Work
During my time working in advertising sales research at NBCUniversal, I had a
conversation with a colleague about our status as researchers within a media organization. We
were talking about a problem we had with one of our internal sales clients. “I don’t get it,” my
colleague said. “Sales thinks we’re just data monkeys, that we’re here to just pull data in
isolation and send it off to them without thinking.” In reality, she went on, our job was to
explain, contextualize, and interpret data for sales—in other words, to package it for them to help
them tell stories. To my colleague, simply pulling a number from a vendor and sending it off
without context was not worthy of being called research.
The above anecdote demonstrates that even within media organizations, the role of
research and data workers is not always clear. If internal clients who have to use data do not
understand what research and data workers do, it is probable that many people outside the
industry, including many readers of this dissertation, do not know either. This chapter seeks to
reverse this issue by contextualizing how research and data work fits into modern media
industries. This chapter has two parts. First, it discusses the culture of the television, digital
media, and advertising industries, focusing on a key tension between the culture being fast-paced
and open to change yet still beholden to lingering legacy thinking. I underscore the particular
role measurement firm Nielsen plays in this legacy thinking. Second, this chapter overviews the
role of the research and data workers within television, digital media, and advertising, using the
frameworks of occupational community and professional subculture. These two sections set the
stage for the following chapter, which will highlight how changes in technology and audience
behavior have changed the media industries and the particular role that research and data work
plays in keeping up with the times.
21
The Industry: Television, Digital Media, and Advertising
The segment of the media industries that this dissertation focuses on is the intersection of
the television, digital media, and advertising industries—where the legacy television industry
interfaces with streaming and other digital media, which can be subscriber- or advertising-based,
and where the types of data used to make business decisions continue to evolve dramatically.
These industries are made up of various stakeholders. For the purposes of this analysis, three key
stakeholder groups are relevant. First are media publishers, who are involved in the production
and/or distribution of media content. These are television networks, cable channels, and
streaming services, for example. Second are advertising agencies, who execute advertising
strategies on behalf of brands. Third are data vendors, who collect data and sell them to clients at
publishers or agencies.
Research and data work provides an important part of the television, digital media, and
advertising industries. In particular, ratings and research are an important part of how television,
digital media, and advertising decide what to do and demonstrate the intersection of creativity
and commerce within this industry (Ellis, 2000; Gitlin, 2000; Wyatt, 2014). More broadly,
television, digital media, and advertising usually fall under what we would call the “cultural” and
“creative industries.” While the former usually emphasizes culture as the output and the latter
creativity as the input, in many respects, the industries they encompass overlap, and both are
concerned with the intersection of creativity and commerce (Hartley, 2005; Hesmondhalgh,
2019; Oakley & O’Connor, 2015). But beyond that, this industry is defined by several factors. In
addition to the creativity and commerce mix, this industry is known to be risky and to have a
high level of ownership concentration, leading to conglomerates (Hesmondhalgh, 2019).
22
While discussing the culture of an industry as broad as the intersection of television,
digital media, and advertising is difficult, a few key themes have emerged. Specifically, there are
new imperatives to be dynamic and fast-paced, but these imperatives battle legacy thinking that
is reluctant to change. This discussion highlights a few of these industry-wide trends and
demonstrates how they are borne out in the four cases chronicled in this study.
A Fast-Paced Culture
Individuals in the media industries have to keep up with a rapid change in these
industries. Many scholars have highlighted the period of industrial change in which modern
media industries are in the midst (Järventie-Thesleff, Moisander, & Villi, 2014; Krumsvik,
Milan, Ní Bhroin, & Storsul, 2019). Many media organizations are reckoning with creative
destruction, described by Joseph Schumpeter (1950) as the process whereby old systems are
destroyed and new ones take their place. These disruptions will be discussed more in the next
chapter. For now, however, it is important to understand that many see this constant state of
change as an immutable part of the industry’s culture.
Participants in my research described the feeling of constantly having to be on one’s toes.
Bobby from Ballast Communications explained to me:
I think change, there is going to be, in even the next couple of years, it's going to be so
rapid that every week we have to expect something new…. So, like, in terms of media,
the media landscape, especially as a TV publisher that has digital offerings that well,
yeah, I'd say like almost every week we have to pivot and reflect on what we're trying to
do.
This participant feels that the change is so constant that they have to adjust weekly. Some
participants even see this reflected in their organization’s culture. Erica from Connected Media
Research had previously worked for a legacy media publisher; she described how Connected was
different from that organization:
23
In the previous role, I always was itching for something new and I was bored and I've
never been bored at [Connected]. We're constantly doing new, interesting concepts that
continue to push my skill set forward. And I think we move very fast as a company
because we have to do that. We have to constantly be first in the market to keep our name
out there.
These are examples of the industry cultures impacting the organizational culture. But not all
organizational cultures keep up with the change of pace. As this participant later described, her
former employer was set in its ways and adapted slowly. This is where we see the other major
trend in the media industries—legacy thinking—still exerting its influence.
Lingering Legacy Thinking
Legacy media organizations, like Erica’s former employer, are more likely to stick with
what is established and less likely to embrace disruptive innovations (Krumsvik et al., 2019;
Küng, 2011). The uneasy relationship between linear television (i.e., broadcast and cable) and
digital media is one of the places where most of this tension exists. There is a split between the
legacy, old guard thinking and the new thinking; linear television is on one side, and streaming
and digital media more broadly are on the other. In this view, television is personified as old and
digital as young. Donovan from Connected Media Research told me, “We skew older for a
startup because, you know, TV is a pretty old business. A lot of those kind of gray hairs are
around.” At the ARF Conference, the head of an advertising technology company explained that
television is “a much more mature piece of the media ecosystem” than digital and that therefore
digital media does have the same baggage that television does and can pivot more easily to
change. Referring to legacy media publishers, Donovan from Connected told me:
I found that they were…a lot more stubborn and just kind of like the old guard thinking,
than I would have expected… Within the TV industry as well, there's always this 800
pound gorilla in the room that is Nielsen, and especially for a start up that one day hopes
to kind of replace the currency, knowing that they're there, knowing that no one is really
happy with them, but they all kind of have to use them. There's like this shared delusion
that like Nielsen data has to be used because it may be wrong, but at least it's hopefully
wrong consistently for everyone. People are very reluctant to use new technology. It's
24
tough to change. There's so much money on the line that people don't want to be wrong
about trying something new. And so I think that's where that stubbornness might come
from. It is just better not to rock the boat for these places.
According to a lot of these sentiments, if television were a person, they would be old, gray-
haired, stubborn, set in their ways, and lugging heavy baggage—not someone ready to be nimble
and adapt to change. Donovan’s perspective is noteworthy because his work at Connected is his
first in the television industry; he had worked in other industries before. Because of this
perspective, he can easily see and describe some of television’s idiosyncrasies. To do so, he used
metaphorical language, personifying television as “gray hairs” and “old guard” and Nielsen as an
“800 pound gorilla.”
Other participants described the role of Nielsen in perpetuating this legacy thinking. This
reflects a view that the Nielsen ratings way of viewing the world—where what matters is
exposure to media content on a television screen first and other screens second—still holds
power in some segments of the industry. Some of the participants’ descriptions of Nielsen and its
acolytes used language related to animals or creatures. Erica from Connected echoed Donovan’s
characterization of Nielsen as a “gorilla” by calling Nielsen a “giant.” These personifications of
Nielsen as a powerful creature illustrate how difficult it would be for a newer organization like
Connected to topple Nielsen. Peter from USA Media also reflected on the role of Nielsen in his
work and described how people who use its data considered themselves superior. Peter works
primarily on the digital assets of USA Media and has to collaborate with others on his team who
work on the linear television assets, meaning that they work more closely with Nielsen data than
Peter does. He described a situation in which there was tension between him and one of his linear
colleagues. I shared my own background working in digital research and collaborating with our
linear counterparts, describing how that work relationship often involved territorialism. Peter
agreed with me and reflected on how difficult it is for the television people to change: “Plus the
25
fact that for a long time, linear Nielsen folks really ruled the roost and were not afraid to tell
people that were involved in digital that didn't matter.” Peter uses another creature metaphor—
“rule the roost”—to describe the dominance of the Nielsen way of thinking and how it continues
to linger. Lawrence, the CEO of Connected, also explained that the Nielsen way of thinking is a
trap:
If you look at people [from] Nielsen or Kantar, none of them [are] able to view any
creative, innovative company, because if you're that like a dominant number one player
for like a 35 years, I guess you try to focus on maybe cutting the cost or try to make some
minor improvement instead of trying to build something completely different.
All these characterizations paint a picture of legacy thinking impeding established organizations
from wanting to rock the boat. Lawrence’s comments echo the findings of Napoli and Andrews
(2008), whose study showed that new measurement systems were unpopular among legacy
media organizations. In this view, legacy thinking stifles innovation.
The Profession: Research and Data Work
In this dissertation, I discuss the television, digital media, and advertising industries by
focusing on a group I have called “research and data workers.” Defining this group and
providing more context about its role in the media industries are two of the aims of my research.
Studies of media work, notably Deuze (2007), have sought to explain the professional cultures of
different groups in the media industries, like journalists and game developers, for example. Other
studies have illuminated the work of television writers (Phalen, 2018), Hollywood agents
(Roussel, 2017), and individuals involved in film production (Caldwell, 2008). Research and
data work has not yet received similar attention.
To explain and define research and data work, I first overview background literature on
professional subcultures and occupational communities, two separate but related concepts that
explain how workers of the same type form identities within and across organizations. Second, I
26
discuss exactly what research and data workers do, linking it to concepts related to occupational
communities. Third, I drill down to explain the differences between subtypes of research and
data workers. Finally, I illustrate the professional subculture of this work, linking to the concept
of occupational community where relevant.
Professional Subcultures and Occupational Communities
While organizational culture is often discussed as a whole, many subcultures exist within
an organization, and some of these are related to particular professions (Bloor & Dawson, 1994;
Hofstede, 1998). In these cases, individuals who are part of a particular profession bring those
values and norms to the organization and may over time influence the organization’s wider
culture (Bloor & Dawson, 1994). Some of these professional subcultures come in the form of
occupational communities, which are groups of people in the same line of work who share values
and behaviors and may continue these social bonds outside work; members share boundaries,
social identity, reference groups, and social relations (Van Maanen & Barley, 1984). In other
words, these groups set boundaries between themselves and others, derive identity from their
work roles, look to each other to set standards for work, and mesh work and recreational
activities (Van Maanen & Barley, 1984). Both professional subcultures and occupational
communities are also shaped by the professional and job training that members receive, which in
some cases is standardized (Bloor & Dawson, 1994; Van Maanen & Barley, 1984).
While the concept of occupational community has been applied to a myriad of fields, it
has been applied to some aspects of creative media work, including communities of songwriters
and video game developers (Schwartz, 2018; Skaggs, 2019; Weststar, 2015). Like the studies of
media work cited above, these studies often demonstrate how occupational communities support
each other in precarious working environments. The following section applies these concepts to
27
research and data work to describe exactly what these workers do, discussing how closely it fits
into the concepts of an occupational community.
Research and Data Work as an Occupation
Media research and data workers are individuals tasked with collecting, analyzing, or
packaging data; this encompasses people who develop measurement products, data scientists
who wrangle large data sets, and traditional media researchers to scour data sources looking to
tell a story. This work is often fast-paced and client-driven. As an occupational community, they
are scattered across organizations and come together at industry-wide events, like the ARF and
Cynopsis conferences I attended and other events organized by the Coalition for Innovative
Media Measurement. Across all of these roles, many participants cited that understanding
datasets, having some fluency with data analysis tools, and having a balance of skills were most
necessary for the job; communication and storytelling, or being able to explain data, were other
often mentioned skills. The following sections unpack these skills and demonstrate how the
particulars of media research and data work foster an occupational community.
Explaining Data to Clients
Research and data workers communicate about data with both internal and external
clients. This task may involve explaining television ratings from Nielsen, data on video-on-
demand views or website visitors from Comscore, website views or clicks from Adobe Analytics
or Google Analytics, or psychographic data from MRI-Simmons, for some examples. Internal
clients are often sales or programming teams or other research teams; these are teams within the
organization that need to use data. For example, a research and data worker at a publisher may
put together a report on a network’s viewers that involves data from different sources and then
need to explain the meanings and limitations of those data. External clients could be networks or
28
brands for data companies and brands or agencies for publishers; these are other organizations
that need to see the data as part of some kind of business deal. For example, research vendors
may send representatives to publishers who are using their data to field questions on how the
data were collected and provide guidance on best practices for using those data. In the next
sections, I unpack how participants described instances like this, noting a clear distinction
between people who understand data and those who do not.
Participants described to me various ways in which they were called upon to explain data
to clients. As Dusty from Bullseye Platforms explained to me:
So [Bullseye] has this tremendous amount of DSP [demand-side platform] and DMP
[data management platform] data within our system, and most of our clients are usually
overwhelmed with, you know, what do we do with all these data and how do we utilize
this data to achieve better marketing outcomes. And our team would specifically get on
calls to work with our account manager to work with that client to address those
questions.
In Dusty’s organization, the issue is that Bullseye has more data than external clients understand
what to do with themselves, so Dusty’s team will often aid these clients. Other participants
discussed explaining data to internal clients, like sales teams. Bobby from Ballast
Communications described this work as teaching and consultation:
I'd say a lot of what we do is data translation, data translation for sales and internally
different groups. I always say this to our group, is like we are teachers, consultants to
sales, to operations and to rev [revenue] and analytics, because we kind of know a little
bit of everything. So we are almost consultants and teachers to sales.
This gives research and data workers boundaries, to cite one of Van Maanen and Barley’s (1984)
elements that make up an occupational community. In both quotes cited above, the participant
draws a clear boundary between people who understand data and people who do not; it is this
knowledge that differentiates research and data workers from other teams in their organization.
29
When working with clients, one challenge research and data workers face is having
others within and outside their organization care enough to hear the intricacies of what they
explain about data. Courtney from Ballast explained to me:
I think a big challenge among researchers is just getting attention and also having people,
having stakeholders actually listen to what you're trying to say and heed your
recommendation and have that research applied in a meaningful way.
This sentiment illustrates what Van Maanen and Barley (1984) say about occupational
communities forming their own reference groups, which can be driven by the community being
stigmatized in some way. In this case, media researchers feel ignored, and although this is not as
strong as an example of stigmatization as some examples in Van Maanen and Barley’s (1984)
text, it hints that the community lacks respect. In this case, the sentiment reflected is that non-
researchers do not understand research. The participant refers to “researchers” having this
problem, pointing toward group identification.
Understanding Research Vendors
Part of this group identification is that media research and data workers have the
knowledge and vocabulary to understand datasets, whereas others within and outside their
organization do not. Speaking from the data vendor perspective, Parker, the head of the team I
interviewed at Connected Media Research, told me:
I invest a lot of time with everybody on really understanding the building blocks of how
we produced the data from the [Connected Media Research] system standpoint. And my
team has nothing to do with the technology that's producing the data itself. But we do
have to interpret it. We do have to think about different ways that we could assemble or
reassemble what we have and to answer different questions. And so I really try and make
sure they understand how the sausage is made so they can think about making a better
sausage.
The key idea Parker expressed is that it is paramount for people on his team to understand their
data, not only so that they can make the data better but also so that they can answer questions.
Once again, we see the role of research and data workers in explaining data, which has to come
30
from understanding it deeply in the first place. Similarly, speaking from the publisher
perspective, Nancy, the head of the team at Ballast, told me:
I think once you start to work within a media organization [a publisher], which is again, a
big difference from an agency, research takes on a different role because suddenly
research is the steward of how your properties are seen. So you have to really dig in
deeply to understand all of the metrics and how they're developed and see how that
advantages and disadvantages you, and again really understand the methodology so that
you can clearly understand whether the patterns and the numbers are correct, whether
there might be any issue, whether one measurement service is better than another. So it's
a much deeper focus on the quality and the specifics of information.
As Nancy explained, part of understanding datasets is understanding what data are and are not
useful; this allows research and data workers to determine gaps in the data they have and
ascertain what other data sources might fill those gaps. Along those lines, other participants
spoke to me about the need to test different research vendors and see if their data are useful. The
ability to do that comes from a deep knowledge of the ways that data are collected. While
research and data workers at vendors have to understand how their own data are produced, as the
participant from Connected Media Research mentioned, others at publishers have to understand
the intricacies of prominent research vendors like Nielsen and Comscore to be able to judge
supplemental data sources against them.
As Grace from Ballast explained: “We still will gut check against Nielsen because that is
a gold standard.” Nielsen’s dominance, which I explained earlier in this chapter, also relates to
creating the professional identity of some segments of the research and data work community,
especially traditional linear television researchers. After Peter and I shared our experiences of the
linear-digital territorial divide in our interview—in which he described that the Nielsen users
“ruled the roost”—he described linear researchers he had encountered as follows: “Nielsen had a
very specific way of training people to do things that they could never move past, never. They all
report the same, they all talk the same.” The strength of Nielsen thereby creates shared
31
knowledge and vocabulary that certain segments of the industry abide by, leading some segments
of the research and data community to remain as ossified as the greater television industry.
This intricate knowledge of research vendors gives research and data workers a sense of
boundaries and social identity, to reference Van Maanen and Barley’s (1984) elements of
occupational community. Many within the industry use shorthand and jargon to communicate
about these various metrics. Peter’s quote above—that Nielsen people “all talk the same”—
shows that this shared Nielsen vocabulary is one way this inside knowledge is signaled. As the
availability of data vendors and metrics explodes, there is even more jargon to understand.
During interviews, many participants would use acronyms or initialisms and would stop to ask
me if I knew what they were talking about. Because my time spent in advertising sales research
gives me some inside knowledge of this community, in many cases, I understood what they were
describing, but the rapid change within the industry meant that there was some new jargon I did
not understand; as the amount of knowledge needed increases, therefore, more boundaries
develop. Furthermore, boundaries are set up between members of this community. As Peter’s
quote explains, individuals who understand Nielsen data would erect boundaries between
themselves and other members of the research and data community, further entrenching
Nielsen’s foothold on the industry.
Because of Nielsen’s impact on and dominance over the industry, research and data
workers also share frustration regarding the inability of Nielsen to measure audiences adequately
in a changing media environment. Understanding the frustrations of working with Nielsen data
and commiserating with others about it are other signs that one has the social identity of being a
research and data worker. For example, earlier in this chapter, I quoted Donovan, who referenced
32
the “shared delusion of Nielsen,” expressing his frustration and amazement that Nielsen’s way of
thinking still has a hold over the industry.
Having a Balance of Skills
Aside from understanding research vendors, many research and data workers typically
use data analysis tools in tandem with other skills. Participants frequently cited hard skills like
understanding data visualization tools like Tableau and programming languages like SQL, R, and
Python, all of which are helpful to discover and tell stories using data. In many cases, however,
these technical skills are balanced with other “soft” skills. Brenden, the head of the team at
Bullseye, talked about the need to have a balance of three different skills: math and statistics,
programming, and domain expertise, i.e. knowing something about advertising. Brenden talked
about how he builds out his team as follows:
You have a baseline for everybody, but then you can have different levels. And it's the
mix of those that gets you what you want because you're not going to be able to hire our
triple threat every time. You're not even going to be able to hire a double threat all the
time.
Similarly, Dusty, a manager who works under Brenden at Bullseye, told me about a similar
approach he takes:
It's a combination of technical and soft skills together. And we don't emphasize that you
have to be very, very technical, like an advanced SQL writer or whatever, or advanced
coding skills. But I want somebody to be the bridge between the technical portion of the
data versus the business strategy. And that's what I look for, and honestly in this industry,
it's hard to get people to have those skills. So usually you find somebody who is stronger
in one or the other and you try to develop the counterpart.
In order to keep up with the changing industry, research and data workers are thus expected to
have a balance of skills that will keep them nimble.
Research and Data Work: Specific Roles, Specific Skills
The perspectives cited above reflect a range of different types of research and data work.
Through the course of my fieldwork and interviews, I discerned a few different kinds of research
33
and data work. Below, I delineate three broad types of work and explain how they are distinct
from the others.
Research and Insights
Research and insights workers, who may also be thought of as traditional media
researchers, pull data from third-party sources (i.e., research vendors like Nielsen and Comscore)
to create compelling stories. They package data to make it understandable for external clients
like brands, agencies, or networks and internal clients like sales, programming, or corporate
teams. For example, if an alcohol brand wants to advertise on a network or digital asset, a
research and insights worker would put together a story showing the percentage of their audience
who is 21 or over and enjoys drinking. As a result, some of their key skills needed are strategic
thinking, communication and storytelling, and understanding datasets, especially those from
research vendors. For those who work in advertising sales research, who make up the majority of
the participants in this study, this process involves fusing a sales perspective with data analysis
skills. Bobby from Ballast Communications described his job as follows:
We are like a fusion of data scientists and marketers right now. So a lot of projects, we'll
be in between data scientists. A lot of times actually we act as a role of the marketer
because we know the landscape really well and we know some data science… I don't
think in a lot of projects what we do, we don't work in between marketers and data
scientists. In some cases we do. But in a lot of cases we act as a marketer, and we work
with data scientists because we understand the business.
In the case of research and insights work, the balance of skills involves knowing something
about the data they use but also understanding the market and how those data should be used.
This allows research and insights workers to be able to wear different hats and adjust the way
they communicate depending on who they are talking to.

34
Analytics and Data Science
Analytics and data science workers, on the other hand, more often work with first-party
(or internal) data. Their key skills include a knowledge of data analysis tools or programming
languages—including Tableau, SQL, R, and Python—and strategic thinking. Data analysts will
more commonly use tools like SQL to extract information for stories, while data scientists will
work more in Python to structure data and create models. While analytics and data science are
thus two distinct fields, I combine them here to distinguish them from research and insights
workers, who are less on the ground working with large datasets. Participants in this study
included one data scientist, while the others in this category were all data analysts. Ken, a data
analyst at Bullseye Platforms, described his job as such:
It's my current role where we are in charge of aligning ourselves with heads of agencies,
analytics, heads of agencies or the brands themselves to get them data that they don't
have or answer questions that they can't answer themselves with the data they have, do
more complex querying, certainly more Tableau and data visualizations, more longer
term projects.
The distinction between what Ken described and what Bobby described is that Ken’s job
involves more hands-on work with data. Both, however, involve communicating externally,
necessitating the balance of skills described above.
Product and Measurement
Product and measurement workers are often involved in big picture thinking to create
new measurement tools and new data and measurement products. They determine how to collect
the data that some of the others, particularly research and insights workers, use day-to-day. As a
result, these workers are more often found at measurement and advertising technology
companies. Their key skills are strategic thinking, as well as some data analysis and
understanding data. Neil from Connected Media Research described his job as follows:
35
I am both responsible for creating the strategy of the products that we sell to our clients,
which includes brands that advertise on TV, includes TV networks, includes others that
want consumer TV behavioral data as well. And so I also am responsible for the
execution of making sure that the internal teams create products that match the
expectations of the market.
Product and measurement work is thus at a level above the other work. Here, knowledge of the
industry is even more paramount, as it is these workers’ job to recognize holes in the market and
strategize how to fill them.
Research and Data Work: Professional Culture
Within the media industries, research and data workers have their own professional
subculture. They typically describe themselves as intelligent, curious, and detail-oriented. They
also tend to socialize with each other and come together as a research community at various
industry-wide events, like the ARF and Cynopsis conferences I attended.
Stereotypically, research and data workers have had to come up against the idea that they
are “nerdy.” This description reflects a view that research and data workers are interested in
minutiae, whereas other groups are not so much so. Nancy from Ballast described having
previously worked for a data vendor, with researchers as her clients:
That was an interesting one, because my clients are research people. They're very smart.
Right now in this role, my clients are mostly salespeople as well as our advertisers. So it's
a different type of need. When you have research as your clients, they poke and prod.
They're very much trying to look under the hood and understand methodology.
Nancy later described the challenges of so many detail-oriented people trying to solve problems
together:
There's been times where like, you know, maybe someone doesn't agree with the
methodology or the approach. We work through that. But that's typical with any, if you
get a bunch of researchers together, it's always a lot of opinions, you know.
The stereotype is that research and data workers are nerdy and detail-oriented, both of which
may contain some negative stereotypes. While research and data workers’ intelligence makes
36
them seem different from others within media organizations, some push back against this
stereotype. As Lily from Ballast described:
I think, you know, historically, researchers can be seen as these basement elves that just
work by themselves and you're, you know, in front of a screen all day. And that's just not
like who I am as a person. So if you're spending so much time with people, you really
need to get to know how they work and the ways that they work so you can work
together. If we're all working in silos, it doesn't make sense.
This conversation with this participant reminded me of the conversation I cited at the beginning
of this chapter: researchers as data monkeys who exist in isolation. Much like the beginning of
the chapter when we saw how participants used creature metaphors to discuss Nielsen’s
dominance in negative terms, here we see creature metaphors to describe negative stereotypes
about researchers. However, Lily wanted to work to reverse that stereotype; she described
encouraging her team to socialize more. Similarly, many participants discussed events that foster
socializing outside work; as we shall see in Chapter 3, participants at Connected Media Research
in particular discussed socializing events. These events build social relations and develop the
occupational community. As more than one person told me during my time in advertising sales
research, “Research is really incestuous. Everyone has worked with everyone before.” Therefore,
we see that the research and data work community extends across organizations and exists at the
level of industry, creating its own professional subcultures that suffuse within their
organizations.
Conclusion
This chapter has defined the research and data work as an occupational community with
its own professional subculture within the wider media industries. Much like other parts of the
media industries, which attempt to be ambidextrous in balancing a need for change without
wanting to change radically, the community of research and data work still relies on some older
thinking, like the dominance of Nielsen, balanced with an emphasis on staying up to date with
37
newer research vendors and data analysis tools. Because an occupational lens is helpful in
understanding sensemaking accounts (Landau & Drori, 2008), understanding research and data
as an occupational community will help us understand how research and data workers enact
sensemaking, which will be taken up in the later chapters of the dissertation. For now, because
we understand the role of research and data work, we can better understand how this work must
adapt to changing environments. The next chapter overviews the changes in technology and
audience behavior that have rocked the media industries, which have created new challenges for
research and data workers.

38
Chapter 2: Disruptive Technology and Audience Evolution
In February 2013, Netflix released the entire first season of House of Cards on its
streaming platform. Although Netflix had already launched original streaming in the United
States with another series called Lilyhammer on its platform the year before, that one was a
Norwegian co-production that first aired on television in Norway (Greene, 2013). House of
Cards was the first series Netflix specially ordered to be produced for Netflix. The show changed
television and became seen as a key pivot point in the television industry: the beginning of the
streaming era (D’Addario, 2018; Smith & Telang, 2016). While the importance of House of
Cards should not be underestimated, several other developments led to the streaming era, going
all the way back to cable disrupting broadcast television beginning in the late 1940s. Since then,
each new development afforded audiences more media consumption options and forced the
television industry to adapt, prompting a battle between old and new.
For our purposes, both technological developments and changing audience behavior are
factors impacting how research and data work has had to adapt. With each development comes a
new conundrum for the industry to crack, so much so that it is fair to characterize the industry as
always having to play catch up to technology and audiences. This chapter seeks to understand the
triggers of change in media industries. It begins by overviewing common triggers of change
before delving into new technologies and audience evolution. It then describes the changes to the
industry that participants described as most disruptive to their work. By doing so, this chapter
seeks to explain exactly how research and data work has moved into the state of rapid change
and chaos it is now in, which will be described more in the following two chapters.

39
Triggers of Organizational Change
Organizations that do not change risk decline (Alvesson & Sveningsson, 2008). Often, a
certain factor—either internal or external to the organization—prompts such a change. Common
internal factors include new technologies and changes in leadership, structure, and strategy
(Child, 2015; Dawson, 2003). External factors are often those that affect multiple industries and
areas across the world (Child, 2015). For example, common external triggers of change include
policies and other regulations, globalization, major world events, new technologies, and
competition (Child, 2015; Dawson, 2003). Internal and external triggers of change may overlap;
for example, an internal trigger of change may change an organization’s goals, which then may
make it vulnerable to external triggers, or vice-versa (Alvesson & Sveningsson, 2008).
Technology, which appears on both lists, blurs the line between internal and external triggers of
change, and as is chronicled above, it is a key driver of change within the media industries.
In the past, incumbents, including those in the media industries, have been hesitant to
embrace new technology and change due to maintaining a legacy mindset (Curtin, Holt, &
Sanson, 2014; Küng, 2011). Chapter 1 introduced the legacy thinking that governs parts of the
media industries, and some historical examples provide more color to illustrate these legacy
mindsets. For example, a study of the broadcast networks’ responses to cable in the 1980s found
that the legacy television networks began to embrace new technologies as business models only
as a way to hedge their bets, but they still lagged behind in some areas; for example, when
CNN’s launch exposed that audiences wanted news content available all the time, the networks
struggled to change their own news approaches (Litman, 1983). In the earlier days of the
internet, the broadcast television networks cautiously embraced the internet mostly as a way to
increase audience engagement because audiences were online, rather than quickly using it to
40
create new business models (Chan-olmsted & Ha, 2003). These examples demonstrate the power
of audiences to prompt change by changing their consumption habits.
Disruptive Technologies and Audience Evolution
In general, technology often gives more control to some players while disrupting the
function of others, thereby shifting the power balance (Uricchio, 2004). In an overview of
disruptive technology that builds from the work of Clayton Christensen, Danneels (2004) offers a
few different ways to understand disruptive technology. First of all, he summarizes Christensen’s
definition by arguing that disruptive technology is that which replaces an incumbent, arguing that
one way to build on this idea is that disruptive technologies are those that make incumbent
technologies outmoded, which negates any investment made in those incumbent technologies
(Danneels, 2004). Then, Danneels (2004) offers his own definition of disruptive technology: “a
technology that changes the bases of competition by changing the performance metrics along
which firms compete” (p. 249). In all three definitions provided, a key idea is a shift from one set
of practices to another set of practices.
Disruptive technologies afforded the audiences more power to control their viewing.
Some examples of such technologies for the television industry include the remote control and
forms of connected television, including the cable set-top box and streaming (Hesmondhalgh &
Lobato, 2019; Uricchio, 2004). The remote control allowed audiences to change the channel
rapidly, avoiding programming and advertisements they wanted to avoid, the latter of which
especially disrupted broadcasting, which is centered on advertiser support (Uricchio, 2004).
Similarly, the introduction of the cable set-top box launched electronic program guides and
digital video recorders, which enhanced audiences’ ability to scan for what they wanted to watch
and avoid what they did not want to watch, including advertisements (Hesmondhalgh & Lobato,
41
2019). Currently, the explosion of digital video providers, including streaming services like
Netflix, afford audiences even more control and choice over programming (Smith & Telang,
2016). Streaming services like Netflix that do not carry advertisements further disrupt the
traditional broadcasting and cable television businesses. Like e-books and digital music in other
segments of the media industries, digital distribution of video content changed the ease with
which consumers could access content and opened the door for new competitors; more
importantly, it forced incumbent firms to adapt or risk irrelevance (Smith & Telang, 2016). The
smooth integration between smart TVs and the internet made it even easier for audiences and
even more urgent for the industry to adapt.
Largely as a result of these new technologies, audiences are increasingly fragmented and
autonomous, both of which have changed media consumption (Napoli, 2011). The two ideas
suggest a balance between structure and agency factors in governing what audiences consume.
Fragmentation—the result of an increasing proliferation of media options—changes the
distribution of audience attention (Napoli, 2011). Because of this proliferation of options,
audiences also have more options to choose what, when, and how to watch what they want to
watch and, in turn, have more autonomy in making these choices (Napoli, 2011). With more
viewing options, audiences choose what to watch using a mix of personal factors and structural
factors. For example, users may make rational choices, but they also rely on genre preferences,
tastes, repertoires, opinion leaders, and their social networks to aid in these choices; these all
involve some level of agency (Webster, 2014). Some of the structural factors that govern what
people choose to watch include socioeconomic status, time available to consume content, and
culture (especially languages spoken or understood), all of which can constrain what someone
can watch (Webster, 2014). The result of all this is that it has become increasingly difficult to
42
attract a large audience—and, in fact, to measure all of these smaller audiences scattered across
different programming and platforms.
Research and Data Workers Describe Disruption
Before diving into what participants cited as the biggest current disruption to the media
industries, I explain some prehistory, as participants also gave me an overview of some changes
that had occurred in the media industries before the present. In a particular way, the change in
audience consumption patterns forced the measurement industry—that segment of research and
data work tasked with measuring audiences—to adapt. The mid-2000s saw the introduction of
commercial measurement. This is the same time that adoption of the DVR and TiVo technology
that allowed audiences to skip commercials was growing. As a result, it became even more
important for advertiser-supported media to be able to measure not just the programs but also the
ads. To address this issue, Nielsen introduced commercial ratings, and the three-day commercial
rating (called “C3”), rather than the live or program ratings, became the standard rating that
matters. In this way, Nielsen adapted its measurement system to meet the changes caused by new
technologies and changing audience consumption patterns, albeit belatedly.
In the following sections, I highlight two specific disruptions that the participants in this
study cited as the biggest disruption to the current media system: the upcoming deprecation of
the third-party cookie and the streaming revolution. I go into depth about the changes that
streaming brought about but only introduce the third-party cookie issue, on which I will
elaborate in Chapter 5. In this section, I demonstrate the challenges these changes cause for
research and data workers. In many cases, participants expressed that they do not think the
industry is well-equipped to meet these disruptions.

43
The Deprecation of the Third-Party Cookie
Currently, third-party cookies are used by advertisers to show consumers more relevant
ads. As many participants explained to me—and as has been chronicled extensively in the trade
press and in multiple panel discussions at the ARF and Cynopsis conferences I attended—third-
party cookies are due to take a hit in the coming years.
2
This change means that websites will not
be able track visitors’ activity on other websites and thus serve visitors targeted ads, though
platforms like Google and Apple will retain the ability to do so. Some participants referred to
this change as a disruption to identity because it will make it much more difficult to identify
visitors. Rather than being driven by technology, however, this change is being brought about by
other triggers of change, namely policies and regulations, which will be discussed in Chapter 5.
The implications of this change with regards to sensemaking digital divides, including user
privacy and agency over their own data, are extensive and will be also discussed in that chapter.
For now, I introduce how the end of third-party cookies is changing how advertisements
are measured online and causing anxiety for research and data workers. This anxiety was
especially felt at Bullseye Platforms, which as an advertising technology company is the one of
the four cases most affected by these changes. As Brenden from Bullseye told me:
So the biggest disrupter is the disruption to identity. And that's what everybody's been
going nuts about…. And mostly the industry is not well prepared for it. A lot of people
have just been paralyzed with indecision about what to do about it. It's going to create a
lot of change and confusion.
This characterization emblematizes the hand wringing about this many in the industry are going
through. In this case, the industry here is lagging in figuring out how to address the upcoming
change to cookies.

2
Not long after these interviews were conducted, Google announced that it would delay deprecation of third-party
cookies on its platforms (Hercher, 2021). Therefore, the concerns participants express here were not as imminent but
are still important for the future.
44
Streaming, Over-the-Top, and Connected TV
Streaming has changed the television, digital media, and advertising industries and
fragmented the audience, which is now dispersed across several platforms. With linear television
(i.e., broadcast and cable), the audience was fragmented across channels, but viewing across
these channels could be measured the same way. On the other hand, streaming meant that
audiences were on different platforms, each of which had to be tracked separately, and were
moving away from linear television. The C3 ratings designed to reveal who was watching ads on
linear television kept falling as audiences shifted to other platforms. As various panelists at both
the ARF and Cynposis conferences explained, this shift has meant that more and more
advertising dollars are going to over-the-top (OTT) and connected television. Unlike the cookie
issue, the streaming change is a force of creative destruction, with a new disruptive technology
displacing incumbents. As Erica from Connected Media Research explained:
I would say streaming has been a massive disrupter in the way that things have been done
in the past. Linear TV is not going anywhere. I think everyone knows that TV is one of
the best places for advertising and advertising dollars are really what drives television.
That's not going anywhere. But people are changing the way they watch TV. And I think
by changing the way that people are watching TV or by people changing the way they
watch TV, it's really been putting more pressure on the industry to change the way they
measure television. For a long time, it's been accepted as it's been very black and white
on how it's measured and with the concepts of VOD coming forward, even DVR coming
forward and skippable ads… And now you have streaming wars. A lot of people are
becoming cord cutters and are moving to those streaming services.
The big question that streaming introduced for research and data work is how to track viewers
across platforms. Traditional television ratings, even newer ones like the C3 rating, tumbled, and
the measurement industry had to make up for this. Streaming meant that measurement methods
had to be changed fundamentally, and as a result, it required the introduction of different streams
of data.
45
In particular, streaming has necessitated cross-platform measurement, or measurement
that tracks viewership across multiple viewing platforms, such as linear television, video-on-
demand, and streaming on computers, mobile devices, smart TVs, or other connected devices.
This is a major challenge for the industry and one that received plenty of attention at both the
ARF and Cynopsis conferences. Valerie, who works at an advertising agency and spoke to me
during the ARF conference, explained the challenge as follows:
I would say the greatest challenge in the media research space specifically is the
consistency in measurement for TV to more digital formats of TVs…. In terms of what it
means when you're looking at a TV ad exposure, when you're in a Hulu or another video
platform, there's just not exact consistency. And I know some of the organizations like
the MRC [Media Rating Council] having their definitions of what it means to be exposed.
And I know there's dialogue in the industry that is discussing how do we change that?
How do we make it consistent?
This participant’s appeal to the idea of “consistency” demonstrates that such measurement
functions as a market information regime, or how an industry makes sense of a market (Anand &
Peterson, 2000). Without consistent information across platforms, various analytics cannot be
market information regimes. Nielsen was that consistent source of information in the past and to
some extent remains so. As Donovan from Connected mentioned in the last chapter, many in the
industry use Nielsen data simply because it is consistent, rather than because it is accurate. With
the changes that Valerie described, however, there are more and more sources of information,
and there is no way for industry players to ensure that they are comparing platforms fairly. This
is one way that streaming has upended the industry: by making existing market information
regimes like the Nielsen C3 rating moot and forcing another player to devise a new market
information regime.
Streaming therefore has acted as a disruptive technology for the television, digital media,
and advertising industries; it has shifted power imbalances and replaced incumbents, as legacy
46
media companies are trying to catch up to the streaming giants, and it has changed competition
metrics, as new streams of data other than C3 ratings are needed.
Driving Change: Technology or Audience?
While most participants cited the deprecation of third-party cookies or streaming as the
top disrupter to the media environment, some participants used language that implies that it is
audience or consumer behavior that is really changing things—or at least that it is a mixture of
new technology and audience change.
3
For example, even though Erica above cited streaming as
the main disrupter, her answer focuses more on changing audience behavior, such as many
people becoming cord cutters. Technology gives audiences the option to change their behavior.
Echoing this view of technology driving audiences to change, Nancy from Ballast
Communications described technology’s impact as follows:
Before DVR became a mainstay, TiVo is what made people start to skip commercials.
And it was all technology. Now all TVs are now smart TVs and are so easily connected
to your Internet. Before people bought smart TVs, it's not connected to the Internet
because they didn't know how to do it. Now it automatically searches for your Wi-Fi, so
you don't have to really do anything but plug it in. And the TV companies decided, “We
need to make sure there's apps and there's places where we can go once they're in this
smart TV environment.” So technology is one hundred percent, I think, the driver of all
of this. Once technology is available, then audiences follow. And then, of course, content
needs to follow as well.
Grace’s comment that “audiences follow” affords audiences only a sliver of agency; rather, it
depicts audiences as simply going wherever the industry directs them to go through new
technology. Other participants, however, spoke of audience behavior directing what the industry
should do. Interestingly, the two participants I quote below speak not of “audiences” but of
“consumers,” which seemingly paints consumers as having more agency than audiences, who

3
In this discussion, I use the term “audience” wherever possible because that is the term that most studies of
measurement use (e.g., Webster, Phalen, & Lichty, 2014). However, participants at various times used terms like
“consumer” and “users,” so where prudent, I use the same terms as participants.
47
have traditionally been painted as more passive. For example, Sean from Ballast explained the
following:
I think definitely consumer viewing has shifted to streaming so that it's a big disrupter.
And then just cord cutting in general, people canceling their cable subscriptions, has
significantly impacted viewership of linear TV networks. Whether they're going to
streaming or not, I mean, that has happened and that has caused the drop in viewership.
The past few years, anyway, viewership has only gone down every year, so I think that's
certainly a disrupter, that there's just less supply in the marketplace, as a result.
In Sean’s conceptualization, consumers changing their viewing habits is the key driver of
change. Joan from USA Media spoke in similar terms:
I think the disrupter is that people or consumers, I think they will go wherever there is
really great content, and they are seeking curated content, they want really good stuff. So
how do you get their attention? It's getting increasingly hard… So in a way, the disrupter
is the consumer, because they kind of have the power, right? I think all these plus
services that have emerged, Disney+, Paramount+, I mean, they can make a ton of money
off of that, but they know consumers are driving the change and it's happening faster. So
I think things that we thought were five years away are already happening right now.
In Joan’s description, consumers exercise their agency through media choice; she explained that
consumers seek good content wherever it is. Even if changing technology begins this process, as
Grace described, the comments from Sean and Joan underscore the agency of audiences to force
the industry to adapt.
How the Industry Responds to Change
As a result of these changes in technology and audience behavior, research and data work
is playing catch up in two ways: first by trying to devise new measurement solutions that track
audiences wherever they go, and second by trying to use new data streams to tell stories.
Measurement Challenges
As the television industry grappled with industrial change, many changes have occurred
in the measurement industry as well. Others have chronicled these changes extensively (Buzzard,
2012; Webster, Phalen, & Lichty, 2014). To say a word, each time television changed,
48
measurement had to change. As a topline overview, Nielsen’s ratings began as phone surveys,
moved to diaries, and then moved to metered measurement. But as was discussed earlier in this
chapter, the DVR increasing audience choice was the beginning of the end for cable and the
beginning of the end of dominance for Nielsen. Grace from Ballast Communications explained
the circumstances surrounding Nielsen’s measuring DVRs as part of its transition to C3 ratings:
I think in 2007 it was enacted, but there was a good year and a half or so in advance of
that, where all the discussions between agencies and networks were talking with Nielsen
about like, how do we go about doing this? And I think historically, Nielsen has been
very slow to respond to transitional change. You'll hear a lot about that common
commentary from people as well as in the trades. So they are just a slower moving
company, but they just have such a foothold on the industry. [People] both on the
network side and as well as the agency side were like, “What the hell, Nielsen? You guys
have to get your act together. This is how audiences are changing. And you need to start
to be able to measure that. Otherwise, it makes your measurement that you have available
moot or not as valuable.”
She also explained that a similar period happened in the mid-2010s, when cross-platform
measurement became more of a reality. Cross-platform measurement is measuring audiences
across different platforms (television, computer, mobile, connected device, etc.). To implement
cross-platform measurement, data must come from a variety of sources, which introduces several
challenges. One of the chief challenges is, as Grace explained, the “huge operational
commitment” to implement Nielsen’s solutions; essentially, Nielsen’s clients have to manually
sync Nielsen’s solutions with their assets through tagging them on the backend.
While Nielsen had its fair share of trouble, including with the rollout of its cross-platform
measurement system described in the introduction (Poggi, 2017), more competitors came into the
arena. As Lawrence, the CEO of Connected Media Research, explained to me:
The monopoly or the dominance of Nielsen data is getting much weaker. If six years ago
when we talk about, OK, like we want to come up with a new measurement solution in
the United States, the investor really doesn't get it, it's like Nielsen [has been] around for
decades. Like why do you want to disrupt Nielsen? I mean, it doesn't make sense. Now, I
think many, many people realize Nielsen may continue to be an important data source,
but we need other data sets to complement them as well.
49
Here, we see clearly that Nielsen’s dominance has been knocked down. Comscore, another
measurement company, has often been viewed as the chief competitor to Nielsen, but panelists at
the Cynposis conference mentioned that the arena is much wider; it has moved beyond a
monopoly or even a duopoly. Players like Connected Media Research have seized this
opportunity. In this way, the way that viewers have how they watch television—chiefly,
streaming—has also acted as a disruptor for the measurement industry. Incumbents like Nielsen
have fallen, power has shifted to a more diverse set of players, and the metrics used to gauge
competition have changed.
Storytelling Challenges
As a result of these measurement challenges, no metrics are perfect. This idea will be
extended in Chapter 4 when I talk about sensemaking data problems. For now, I introduce how
research and data workers face what they often refer to as “storytelling” challenges due to the
measurement challenges explained above. Melissa, who works at a media publisher and spoke to
me after the ARF conference, explained that tagging and privacy challenges on measurement
services can hamper accurate measurement:
Since OTT [over-the-top] is becoming so big and it's such a priority across not just our
company, but the industry, Comscore's tagging for it is really bad. So some of our
networks are tagged, some of them aren't. Same with competitors. So then our numbers
are undercounted… So we always have to give that caveat when we're giving Comscore
data…, that it's just based on desktop mobile as a proxy, because we don't have any
tagging capabilities for certain networks… And then the second area, kids is a really big
challenge. And because [a younger-skewing network] is a huge focus for me, you can't
measure under 13 for mobile. So that's just like a hard line of we always have data issues.
And so anyone who's worked on the network for a while, once you just get used to it,
there's just an understanding that the data is always going to be a little bit off. Use it
directionally. We use that term all the time, so we use this as a proxy.
Because of issues accurately tagging data, some audiences are not captured properly and are
undercounted. In this case, certain audiences are in fact invisible simply because the technology
to measure them is not yet easily implemented. This creates a storytelling problem because
50
Melissa and others like her know that these audiences exist, but they have no data to paint a
picture for clients to understand what they are doing.
More generally, Joan from USA Media described how the rapid change in audience
behavior causes storytelling challenges:
I think it's keeping up with all the changes, but then also, you know, using data to show
internal stakeholders that like, OK, this is what the marketplace or consumers are
demanding, how are we going to it may not be what you wanted to do in the next two
years, but like this is this is a change that we have to make. Right. Like being convincing
and giving a compelling story and compelling reasons as to why these things need to
happen.
Joan’s quote here, paired with her quote above about how certain changes they thought were five
years away were happening at the present, illustrates how fast everything is changing in this
industry. Joan needs to keep up with new data in order to communicate these changes to other
stakeholders at USA Media.
Conclusion: COVID-19 Accelerates Change
This chapter has explained how new technologies and changes in audience behavior have
created a dynamic, fast-changing environment for research and data work. As audience
consumption patterns continue to evolve, research and data work must adapt to every change,
both in how it measures audiences and in how it uses audience data to tell stories. These changes
generally arise from new technology, which is an external trigger of change. As Child (2015) and
Dawson (2003) explain, major world events can also be external triggers of change. The
COVID-19 pandemic is a major world event that had a big impact on the television, digital
media, and advertising industries. My fieldwork and interviews were conducted between
September 2020 and June 2021, several months to over a year into the COVID-19 pandemic.
While the particulars of the pandemic are specific to that specific moment in the pandemic, they
illuminate how global events can cause change within organizations.
51
As a topline overview of how COVID-19 has changed the media industries, Courtney
from Ballast explained:
COVID is the biggest disruptor in the media landscape. I feel like it's changed everything
from how I work, to what I work on, to what our priorities are as a team and as an
organization. It's just completely changed everything, and we're also seeing a greater
fragmentation of viewing. Because of COVID, people just aren't commuting, they're not
listening to podcasts on the train. They might be streaming more, those kinds of things.
And it's been challenging with time and again, like the whole reorganization, it's been
challenging to find ways to track and measure media that is just on all of these new
platforms and different devices that we're now selling ad space on as an organization.
And now, our team has to figure out how to measure campaign performance on.
Courtney’s comments demonstrate how COVID-19 has accelerated changes that were already
occurring in the media industries. In particular, audience consumption patterns, which were
already leaning towards streaming, changed even more. Because streaming was already difficult
for legacy publishers like Ballast to measure, this created more challenges for media
organizations. On that note, the next chapter will describe how many of these challenges,
including those posed by COVID-19 and remote work, have forced media organizations to find
new ways of organizing.

52
Chapter 3: Culture and Organizing in Times of Change
As we saw in the last chapter, changes in technology and audience behavior created new
challenges for research and data work, and many of these industrial disruptions, like streaming,
were exacerbated by the COVID-19 pandemic. This chapter takes our discussion down from the
industry level to the organization level by examining culture and organizing in times of change.
Specifically, it examines some of the changes that have taken place and how organizations are
looking to implement new ways of organizing to improve culture. As we shall see, many—but
not all—of these new ways of organizing implement practices associated with environments for
creativity and innovation, which will be explored further in the next chapter.
This chapter begins with an overview of cultures that are conducive to creativity and
innovation. The main part of the chapter comprises four mini case studies, one for each
participating organization. Each case study begins by painting a picture of the organization’s
culture and the culture of the team I interviewed, as described by the participants. In these
sections, I underscore each team’s culture as it operates in the larger organization for purposes of
analysis. However, it is important to note that these teams are an integral part in constructing the
overall organizational culture, so pulling each team out for analysis is merely a methodological
tool. The overall organizational cultures and each team culture are also constituted by the
professional subcultures I described in Chapter 1. I also note the extent to which COVID-19 and
remote work disrupted their organizational norms and practices, as some teams were already
geographically distributed and did not experience as much of a disruption.
In each case study, I then lay out how elements of communication serve to reinforce and
build culture. For three of the case studies, I explain how hiring and training also reinforce and
build culture (neither of the two participants from USA Media discussed this, so I do not include
53
a discussion for that case study). This chapter also includes one extra example of new ways of
organizing once a team became all women; the participants are not identified with one of the
organizations due to the sensitivity of what they told me. Finally, this chapter offers some
analysis and illustrates how many of the new ways of organizing are bringing these organizations
closer to established cultural practices that promote creativity and innovation. Many of the
challenges cited in this chapter are a result of COVID-19 forcing remote work; however, as the
last chapter demonstrated, the rapid rate of change in the media industries was already making
research and data work play catch up to changes. COVID-19 merely sped up the rate of change.
Improving Culture
In this chapter, I use the concepts of organizational culture as text and as climate
(Eisenberg & Riley, 2001). More specifically, I describe culture based on what participants say
about their organizational culture, using their words as a text to describe the culture. This is
partially a limitation of COVID-19 preventing me from doing any participant observation, as I
have to rely on interviews to evaluate culture. Second, I describe culture using the
communication dimensions of climate: support, trust, and openness to new ideas (Eisenberg &
Riley, 2001). Describing culture is not my only aim in this chapter, however. More specifically, I
focus on how these cultures are improved through new ways of organizing, especially as a
response to challenges related to COVID-19 and remote work. Here, I take a processual view of
organizational culture change, which Alvesson and Sveningsson (2008) argue sees culture
change as an ongoing process. Even more specifically, I use the change metaphor of “fix and
maintain” (Alvesson & Sveningsson, 2008; Marshak, 2002), rather than a complete culture
change. “Fix and maintain” means seeing change as not necessarily large-scale but more focused
on reinforcing elements that already work through perfecting them. The changes I will describe
54
in this chapter fall under this umbrella: they are focused on improving practices to reinforce
existing cultures.
The question then becomes: fixing and maintaining what kind of culture? As we shall
see, many participants described maintaining cultures conducive to creativity and innovation.
The main factors that foster creativity and innovation at the team and organizational levels are
(1) freedom, autonomy, and trust, (2) resources, and (3) open communication and support. These
match with the communication dimensions of climate (Eisenberg & Riley, 2001). The type of
freedom and autonomy that encourages creativity refers to the freedom to determine how best to
approach a project, rather than the freedom to decide what projects to take on (Amabile, 1998).
Relatedly, a culture of mutual trust, environments that allow employees to take initiative, and
freedom from tight control can also stimulate creativity and innovation (Hofstede, Neuijen,
Ohayv, & Sanders, 1990; Jassawalla & Sashittal, 2002; Malmelin & Virta, 2017). Resources that
allow for creativity and innovation include both time and money, but even the absence of such
resources can create challenges that make individuals work creatively (Amabile, 1998; McLean,
2005). Support can come from the organization, one’s supervisor, or one’s team (McLean, 2005).
A key element of support is open communication among individuals, their team, and their
supervisors (Angle, 1989; Kanter, 1983). Part of this open communication is a focus on
constructive feedback and sharing among team members (Amabile, Conti, Coon, Lazenby, &
Herron, 1996; West & Sacramento, 2012).
Industrial Change and COVID-19: Creating a Reactive Environment
As the previous chapter explained, the media environment has become dynamic and fast-
paced. This has resulted in changing technologies and audience behavior and the COVID-19
pandemic. As I explained in Chapter 1, research and data work is highly client-driven; these
55
workers must respond to the needs of their clients, both internal and external. As a result, this
work can at times become reactive instead of proactive. Joan from USA Media described how
the changes in the media environment have made her team’s work more reactive:
With the BI team, it is a huge infrastructural challenge, because we have to basically find
out where everyone, regardless of age, is watching TV and figure out how to measure
them. And what's hard is that it's very reactive because we have to, like a new piece of
technology is coming out or like, oh, now we've got our [specialized] app or whatever
app on a smart TV. OK, we have to be able to measure that. Right. So it's a challenge. I
mean, I think it's just, it's really hard.
As Joan described, with every change came a challenge for her team; they had to ensure that they
were prepared to answer questions about each new piece of technology. Courtney from Ballast
echoed this sentiment but described it based on changes related to the COVID-19 pandemic. She
explained:
Since the onset of the pandemic, there have been few changes in our organization and sea
changes in the way that people are consuming media, and that has forced us as a research
team within an ad sales body to be hyper responsive. There's a lot of pressure to bring in
more revenue. There's a lot of pressure to bring in more clients to potentially offset any
losses that were experienced when brands just decided they were going to pause or delay
campaigns altogether.
Courtney focused on the ways that her team has been forced to stay nimble to respond to the
changing needs of their organization. Courtney said in another part of her interview that this
environment is so fast-paced that she felt she cannot sit down and focus on being creative.
Because the work environment has become very reactive, finding new ways of organizing to
improve culture—especially through fostering increased support, trust, and collaboration—is
paramount. To meet the demands of change, research and data workers should be able to think
not just reactively, but creatively.
The Four Cases: Improving Culture through New Ways of Organizing
This section explains the four case studies that make up this analysis. All of these teams
consist of groups with different subspecialties within their team, meaning they are functionally
56
heterogeneous, which as a design structure, helps create an environment for creativity and
innovation (Somech & Drach-Zahavy, 2013). Furthermore, many of these teams noted that their
leadership set the tone for their team culture, meaning that culture is leadership-driven and top-
down. Finally, groups that had remote work in place before COVID-19 adapted more quickly to
the remote work scenario and did not experience as much disruption in adjusting to teamwork
during the pandemic.
Bullseye Platforms
Bullseye Platforms is an advertising technology company. The team I interviewed is
composed of two subteams: an analytics group and a business intelligence group that specializes
in one of Bullseye’s proprietary data platforms. I interviewed the head of the overall team, four
members of the analytics subgroup, and two members of the business intelligence group. I also
interviewed one manager from each of the subteams.
Before the COVID-19 pandemic, Bullseye Platforms had offices across the United States
(east coast, Midwest, and west coast), and the team I spoke to had members at four different
locations. At least one participant was working from home well before COVID-19 forced remote
work. As a result, this team did not find that remote work strongly affected their work, aside
from the inability to visit a colleague’s desk and talk. This team talked about how much they use
Slack to communicate amongst each other; they also have manager one-on-one meetings and a
weekly team meeting so that everyone is aware of what everyone else on the team is doing.
Organizational Culture
Participants described Bullseye as having a hardworking and smart culture. Jordan
described all the teams she had worked with as follows:
One thing that I really like about [Bullseye] is every team that I work with, is everyone is
very, very smart, and everyone, for the most part, I would say also cares about their job
57
and performs well… I haven't worked with any cross-functional teams where it's like that
person's a slacker or they are just like not putting in any effort.
Participants see their colleagues at Bullseye as both highly intelligent and motivated. Although
hard work, intelligence, and motivation tie all the Bullseye teams together, the organization has a
myriad of different teams, each with its own culture, according to those I interviewed. As Adria
put it, “I'd say they all are very smart, but it depends who's in each office that you have a
different personality.” This leads Bullseye’s overall culture to come across as siloed. Individuals
from the same team and from different teams are scattered geographically and may not often
interact with each other. Furthermore, many offices will be dominated by one type of team, such
as sales or engineering. Jordan, who is based in the Midwest and who has interacted with other
teams, told me: “[Bullseye] is like a Cheesecake Factory menu of like offerings,” noting that
every office has its own culture. As a result, participants noted that sometimes it is difficult to
describe the organization’s culture because each office and team is distinct. This is especially
important to note because I spoke to individuals who, before the COVID-19 pandemic forced
remote work, had been based in four different physical offices, despite being on the same large
team.
Team Culture
Despite the overall organizational culture coming across as siloed, participants from both
subteams described their team as curious, hardworking, and collaborative. Here, I focus on the
collaborative element. Brenden, the head of the team at Bullseye, explained how he fosters
collaboration on his team, which usually begins with him assigning a project to someone. He
described what happens next as follows:
That person is going to work on it. Now, they might decide that they need input, that they
need to bounce ideas off somebody. And so one thing I've done with my team and that I
think is really important is I try to create a very collaborative atmosphere. I want people
bringing their problems to the team, proposing their ideas, pressure testing things. And so
58
with my team, I've tried to create both the spaces to do that, as in like we're going to have
certain meetings where we're going to nominate topics and everybody's going to dig in on
that topic for this meeting and share ideas or the like.
Brenden’s statement shows that he is following one of the aspects of culture that fosters
creativity: constructive group feedback (Amabile et al., 1996; West & Sacramento, 2012).
Brenden encourages his team to give each other feedback in the hope that it helps everyone do
better work. Stan, who works under Brenden, echoed this view that the culture is one of trust,
demonstrating that Brenden’s tactics work:
I think his [Brenden’s] style is very, very open and honest, like not a lot of spin, he sort of
trusts people to. You know, I think his philosophy, and I think it's a good one, is to try
and get the best people you can in the door and then just give them the space they need to
do what they do. So we are very free to, we were given kind of ambitious projects, maybe
stuff that even on paper, we were not 100 percent qualified for. But then, like, given the
latitude and the support to actually work our way through that stuff.
Freedom, autonomy, and a culture of trust—all of which are reflected in Stan’s description of his
team’s culture above—lead to greater creativity and innovation (Amabile, 1998; Jassawalla &
Sashittal, 2002; Malmelin & Virta, 2017). Other leaders under Brenden try to replicate this
culture to reinforce it. Dusty, a middle manager, trusts his analysts and encourages them to help
each other, facilitating collaboration even at the lower levels:
The way we work is that obviously whenever there's analysis they're the ones executing
them, I'm trusting them to execute things. If they have any question, they will come back
to me or they will try to ask each other and learn from each other. Right now, the junior
one is asking the slightly senior one a little more, obviously. But that being said, when
they encounter, for example, a conversation that will need to be more strategic with the
client, that's where they'll loop me in.
Dusty’s description of his team emphasizes not only trust but also open communication. A key
element of support, which stimulates creativity and innovation, is open communication among
individuals, their team, and their supervisors (Angle, 1989; Kanter, 1983). Here, we see that
Dusty facilitates communication not only between him and his analysts but also between his two
59
analysts themselves. Taken with Brenden’s and Stan’s descriptions above, we see that open
communication is encouraged at all levels of this team.
New Ways of Organizing
Bullseye has taken the rapid rate of change as an opportunity to begin some new ways of
organizing. COVID-19 and remote work became moments for people to participate in the
organizing process of Bullseye. In particular, this falls under two buckets. First is building and
preserving culture at both the organizational and team levels, in part to account for the siloed
structure of the organization. Second is fostering increased camaraderie and collaboration among
team members, which was challenged by remote work. In the discussion below, these two
buckets intertwine because many of the same tactics used to build and preserve culture are also
used to foster camaraderie and collaboration.
While participants at Bullseye noted that remote work was not too much of a disruption
for them because they were distributed well before the pandemic, they experienced some change,
including an increased use of electronic communication. One major change cited by Jordan and
Lacy, among other participants, was that they were now unable to jump over to a colleague’s
desk to ask a question. To account for this, Bullseye has seen a surge in usage of the messaging
system Slack, which Jordan noted they were using before. Jordan described the situation as
follows:
It wasn't Groundhog Day before COVID. I definitely feel like the hours glued into my
screen were less before COVID, and the work amount has gone up post-COVID… And
so now instead of someone coming to my desk, back in March and April, the Slack and
the pings was just overwhelming. Obviously, it's a way to connect. And we had Slack
before… We've always had video conferencing for internal calls. We've always had Slack
and pings. I'd ping the person who sat across from me just to chat about something. So
the technology we used pre- and post-COVID hasn't changed for communication. It's just
the amount obviously has gone up.
60
The team uses Slack not only to facilitate quick communication and collaboration but also as a
venue to facilitate mutual support and culture building, which in this case is important because
participants noted that Bullseye’s overall culture is siloed. Bullseye as a company has facilitated
some of these events on Slack and other channels. Ken described, “We have office managers
who try to promote trivia on Slack or dress up your cat in a weird way contests or just stuff to
win some Amazon gift card or have some levity in a Slack channel.” Adria talked about how she
uses Slack and other virtual events for this:
I'm really focused as a manager, and I'm just really stressed out, especially while
everyone's quarantined, to keep morale high. And so we try to figure out fun ways to
interact, either on that Slack channel or I have been really encouraging everyone to host
happy hours. So I hosted one… It's just kind of an opportunity for someone to say, like,
OK, I have a great theme idea or something like just something silly to try and make
everyone feel connected or have a bit more fun just because we aren't seeing each other's
faces.
We see from these quotes that while Slack had at one point become a distraction and time drain,
individuals within Bullseye, including Adria and the office managers Ken referenced, have
begun to use Slack and other virtual means to promote support and camaraderie among people.
In this way, virtual communication becomes a force for building culture within and across siloed
teams. Dusty added that he was about to begin experimenting with another virtual collaboration
system, pointing to the fact that Bullseye as an organization is open to experimentation:
I think that's the real collaboration, like the small chitchat and being able to pull
somebody aside and start white boarding stuff. I've been thinking about that and over the
weekend, just this past weekend, I saw something. A line that I thought was incredibly
interesting. It's called Gather… Starting this week, I am testing Gather, which will allow
me and my team to be in a virtual office together.
During our interview, Dusty attempted to show me how Gather works, but we were unable to get
it to work. Essentially, Gather is an online platform that creates virtual office spaces for teams to
interact. Dusty’s decision to test it out is along the same line as other individuals within Bullseye
61
who experimented with online platforms to facilitate support and collaboration during the
pandemic.
Hiring and Training
At Bullseye, participants spoke to me about making hiring and training decisions based
on skills, particularly emphasizing a need to find a balance between hard and soft skills, which is
essential to research and data work, as I described in Chapter 1. Adria participated in a hiring
decision for the first time recently. She described the process as follows, noting that she had had
discussions with the hiring manager about “culture fit”:
Culture fit wasn't ever like, oh, you have to be a certain age, certain demographic for
certain whatever. It was really, are they going to be hard working? Are they going to be
communicative? Are they going to be good at presenting? … But the things we looked
for was like with that mind I set out, like, what are the key things we really need in this
role? And it's someone who's good. It's kind of a weird role because it's someone who's
good at doing things maybe by themselves, like doing that hard-core analysis, but then
who is really able to communicate and sort of do this like sales flair in presenting it.
What Adria describes above is not only a balance of skills but a person who fits with the
elements that stimulate creativity and innovation described above: she needs to find someone she
trusts to do a good job and someone who can communicate with the team. As a result, Adria’s
hiring decision served to reinforce Bullseye’s existing culture.
Dusty looks for a similar mix of skills but framed it around having a baseline of technical
skills and communication skills. He looks for a candidate who is strong on one side and works to
develop the other side. He finds that the technical skills, like Tableau and SQL, are easier to
train, and he usually does this by walking them through a current project. In his opinion, the soft
skills are harder to train. He explained how he goes about training soft skills as follows:
I would start off with the person usually to nail down their internal communication right
before they start communicating externally. Internal, you don't have a lot of pressure and
internal you always get very honest feedback about how you can improve…. Presenting
internally first, you're always going to get the first hand, how does the client think and all
62
that, and once they're ready, you start to do more external client meetings and start
getting that.
Similarly to Adria, Dusty describes that training these soft skills involves open and honest
feedback, much like the team meetings Brenden and Stan described above. In both cases, open
and honest feedback is not only encouraged but is expected. Here, we see that even the process
of training someone to communicate well reinforces existing cultural practices, in this case, a
focus on open and honest communication.
Connected Media Research
The team I interviewed at Connected Media Research covers product, analytics, and data
science. I spoke to the CEO of the company, the head of this team, one member of the product
side, two members of the analytics side, and one data scientist. Much like participants at
Bullseye, participants at Connected also noted that remote work was not a large disruption
because Parker, the head of their team, worked remotely part time well before the COVID-19
pandemic mandated it. Because he was not always present with his team, Parker told me that he
had previously worked on a lot of what he described as “social connectivity” among his team to
ensure everyone felt supported, which included open lines of communication with him and
among other team members. Much like the Bullseye example, Connected’s new ways of
organizing involve building and maintaining culture, especially through encouraging
collaboration and facilitating team building events.
Organizational Culture
Connected is beginning to mature from its startup days but still works to maintain a
startup culture. Participants noted that the company had not seen as many layoffs as many
startups during the COVID-19 pandemic, which they took as a sign of its stability. However,
many still enjoy the benefits of its startup feel, as Erica explained to me:
63
I think one thing for me that's been really important, and I think our leadership teams
have done a fantastic job, is maintaining that culture. The benefits of startups is that
you're small, you know everyone, your CEO knows your name.
This feeling of being known makes participants from Connected feel closer to the organization.
However, at least one participant felt that this startup culture made the organization still
somewhat siloed. Skyler described: “It's what I would describe it again, like organizationally,
culturally, at that startup phase, it's very siloed off and working on small, digestible, bite sized
things. But it hasn't gotten to the point where it's like looking larger.” In Skyler’s assessment,
Connected’s startup culture is not an asset but rather a detriment, preventing teams from
collaborating on bigger projects with other teams across the organization.
Aside from the startup culture, participants at Connected stressed a culture of
empowerment, noting that they felt they had control over their own work and often received
recognition for a job well done. Lawrence, Connected’s CEO, explained to me that he built this
employee empowerment into the culture as follows:
Another thing is we try to really empower every employee to give them ownership…. So
it's very agile, it's very nimble, and we believe we try to hire the best people and that the
best people have to come up with their own idea to figure out how to make things work.
It's very different from a big company: “We've got this master plan. There's like three
functions and we've got a detailed plan. And OK, so every employee is supposed to
exactly follow the plan.” If you're a big company, maybe that's the only way to manage it,
but we certainly try to avoid that. We try to build more like a bottom up empowerment
approach to address those issues.
Lawrence wants frontline employees to have ownership over their work, even using the term
“bottom up,” and other participants emphasized that leadership listens to them. Erica explained
to me that the company sends out employee engagement surveys to ensure that the company is
still providing a supportive environment and culture. However, the directive for frontline
employees to take ownership of their own work is still coming from the top. For example,
Connected’s CEO also explained to me that the company gives out an award during company
64
all-hands meetings to the employee who best espouses the values of the Connected’s culture. As
a result, Connected’s leadership both encourages and recognizes excellence. Similarly what Stan
said about his team at Bullseye, Connected also hires people they trust to find solutions on their
own.
Team Culture
Participants at Connected Media Research described their team as collaborative, curious,
and open and honest. Parker, the head of the team, mentioned the importance of giving his team
a sense of autonomy over their own work. He explained:
I'll go to battle with my team, but I need to have my team take ownership of the work that
they're doing. And I think that goes hand-in-hand with the intellectual curiosity thing
because if someone is just executing on a statement of work blindly, to me, that's not
ownership… I think we have a responsibility, if we see something interesting, to pull at
that thread, to ask ourselves why. And I think that comes from that sense of like I want to
know, like I'm not just going to accept that this is the way it is. Doesn't make sense to me.
That's not how I watch TV. That's not like my understanding. Let me see what's going on.
And a lot of learning comes from that, so I think ownership is one of the one of the most
important elements of all of that.
Much like the other teams, Parker wants his team to care about their work, which comes from
them having ownership over it. Parker continued, explaining his philosophy for leading his team:
We have no ego on our team, so what I mean by that is if we're looking over work that
someone's done and there's feedback, like one of the things that I really push is like we
have a responsibility as a team to give that feedback, like it is a gift to the to the team. It's
the only way we grow is by openly discussing these things. And it's not personal. It is
about the content and it's in pursuit of growth and intellectual honesty.
Parker’s use of the word “responsibility” shows that he entrusts his team to handle work in an
honest and thorough way. As a result, while the literature often cites freedom and autonomy,
many participants in this study rather used words like “ownership” and “responsibility,” pointing
towards an understanding that the employees are expected to operate autonomously, rather than
autonomy being the exception. Parker explained that he worked on creating an open and
supportive environment among his team; in his team, open and honest communication is a gift to
65
other team members. He communicates to his team that honest feedback is expected because it
makes the work that they are doing even better; this aligns with the literature demonstrating the
importance of constructive group feedback (Amabile et al, 1996; West & Sacramento, 2012).
This demonstrates that Parker is concerned with creating an environment that will make his team
as successful as possible.
During the interviews, when I spoke to frontline employees or middle managers, I asked
them if they had ever pitched a project to their managers and how it went. Most of these
participants responded positively, noting that their managers were receptive to new ideas. Erica
from Connected referenced an example and explained the dynamic between her and Parker:
It actually happened today or yesterday. I was on a client call, and I had an idea for how
we could do some work with this one client, but I wanted to get [Parker’s] feedback on it
before maybe pitching it to the client itself. So I pitched him my ideas, we walked
through what works, what doesn't work, and then we basically built off that and actually
formulated a statement of work or a concept of first statement of work that was very
collaborative between my ideas and his ideas. And I would never say that with my
current team, I'm never really put down. All ideas are welcome. And then we elaborate on
what would work, what doesn't work, why does that not work?
This example shows not only that Parker encourages his employees to have new ideas but that he
takes the time to help them develop them. There is an open flow of communication between the
two. From these examples, we see generally that people feel comfortable pitching ideas, which
allows them to work without constraints.
New Ways of Organizing
Like many organizations, the challenges for Connected involved maintaining a sense of
culture through the pandemic. Because some participants saw the organization’s culture as
siloed, we can argue that some of these events are designed more to build rather than maintain
culture, however.
66
First, COVID-19 and remote work caused some disruption to this team. Many
participants described that they could not go over to a colleague’s desk to have a conversation,
much like the participants at Bullseye described. Neil described that this shift meant that
communication became more formal, as a meeting was needed for everything:
I think where the area that I personally struggle is when you're in the office and there
somebody has a conversation, but then it sparks an idea in your head or in somebody
else's head and you kind of have this mini discussion with the people in the office. And
so now to have that, you have to schedule some time to schedule a call. So it's like it's a
formal event, where before it was kind of an organic thing.
Even though participants mentioned the challenges of formal communication, Parker decided to
institute more team check-in meetings. He described two specific types of meetings he began
during the pandemic to foster support and collaboration. The first is a morning meeting:
We instituted a daily morning meeting for the team, which we didn't always have. But I
felt given that we are apart, we're not seeing other people, it was helpful to have that
time. We don't use it every day, but it's always blocked. And if anyone needs it or
anyone, and we've said if anyone just feels like they want to talk to another person just to
say hello, just let people know and everyone will be there for the morning meeting.
These morning meetings serve the purpose of maintaining a supportive culture. Parker makes
himself available to everyone to keep the lines of communication open during the pandemic. The
other type of meeting Parker started was a larger team meeting:
We instituted a more in-depth weekly team meeting, which was a chance for us to
discuss, like one topic in depth so someone would bring a topic, something that they were
working on if they wanted input or they wanted to update the team on what they were
doing. It's a great opportunity for knowledge sharing.
Parker’s description of this team meeting is similar to the team meetings that Brenden described
above. In both cases, Parker and Brenden encourage members of their team to bring project
examples to these meetings and allow other members of the team to give open and honest
feedback, thereby encouraging a collaborative atmosphere.
67
In addition to these changes that have happened at the team level, Connected has
emphasized organizing wider events to build culture. In addition to the company all-hands
meetings that I reference in the previous section, Connected’s head of human resources organizes
social events for team building. Donovan described it to me as follows:
Our HR has actually done a really good job of scheduling events, whether it's like making
a birdhouse, doing a whiskey tasting, doing an escape room. I think we're making cheese
next week. So in that respect, I think credit to HR for doing a wonderful job of making
sure that, you know, we do try to maintain some sort of pride in our work. And I don't
know if that's necessarily the best word, but they've done a good job of trying to look
after us.
Other participants from Connected referenced how effective these activities are. These
sentiments are notable not only because they reflect an unusually positive view of an
organization’s human resources department but also because they illustrate how Connected’s
leadership entices frontline employees to want to do a good job in their work, underscoring
Connected’s commitment continuing to build to bottom-up employee empowerment during the
pandemic.
Hiring and Training
Much like the team at Bullseye Platforms, the team at Connected Media Research
emphasizes a balance of skills when hiring new members for their team. Parker says that there
are three elements that make someone a good candidate for a position on his team: technical
skills, some understanding of the media industry, and intellectual curiosity. When he makes a
hiring decision, he looks for someone with intellectual curiosity first:
What is most important is, and it's much harder to train, and what I tend to focus on is
intellectual curiosity, like storytelling, because a lot of what we're doing, we're going to
like discovery mode. We're trying to create something new. And if you don't have that
intellectual curiosity to start pulling at threads asking why like two, three or four times to
get to the bottom of things, I just don't think that you can make those strides that you
need to make.
68
Unlike Dusty’s emphasis on communication at Bullseye, the key soft skill Parker looks for is not
communication but curiosity. Parker is concerned about how potential team members approach
problems. This emphasis reinforces Connected’s and this team’s focus on empowerment and
ownership. In this case, Parker needs to trust that members of his team have the capacity to ask
questions and find answers for themselves. As a result, even looking for a key innate curious
nature ends up reinforcing elements of Connected’s culture.
Ballast Communications
At Ballast Communications, I spoke to an advertising sales research team. Participants
included the head of the team and members of three subteams that worked on several Ballast
properties. This team at Ballast all described remote work as more of a disruption than other
teams did. Before the pandemic, they were all working in the same office and at offices and
desks relatively near each other, which made in person collaboration easy. They had a major
adjustment to work from home.
Organizational Culture
Some participants found it difficult to describe Ballast’s overall culture because it is such
a large organization, and their team does not interact with every part of the organization.
Participants who did describe the culture of Ballast Communications described it as dynamic.
Reflecting the constant change that occurs in the industry, Bobby described how that affects
work at Ballast:
We are always told, you know, we have to adapt to change. Change is going to happen in
the marketplace, and it happens even further in [Ballast]…. We have to get used to
change in this specific industry, especially with a mix of TV, we have to get used to tons
of change. So, I think change, adapting to change, being innovative, slow evolution, I
mean, an evolution over like complete revolution are like big priorities within our
organization and reflected even in [Nancy’s] org too.
69
Because this participant describes that the organization constantly reminds them of the need for
change, it reflects that there is still legacy thinking at this organization that they are trying to
dismantle.
4
It is also noteworthy that Bobby mentioned this organization is more adept at
evolutionary instead of revolutionary change, which supports the idea that established companies
are better-suited to adapt to evolutionary instead of revolutionary change (Christensen &
Overdorf, 2000). Despite the emphasis on keeping up with the times, then, Ballast prefers to take
its time adapting to such change.
Team Culture
Participants in this advertising sales research team at Ballast described their team as open
and honest, close knit, collaborative, smart, and curious. Many participants referred to their
overall team as “[Nancy’s] org” pointing to team identification that directed from Nancy’s
position of leader. As Grace said:
Out of the nature of how [Nancy] works, we collaborate on a regular basis because we
meet weekly and we talk through projects that we should be aware of. They may not be
important or relevant to us right now, but they might come down the pipeline and affect
us down the road. So there's a lot of information and knowledge sharing between our
groups. And I would say that we have a pretty open dialogue about where there's a need.
Similar to the teams at Connected and Ballast, this team has a focus on open communication so
that all members are aware of what the others are working on to stimulate learning and
knowledge sharing.
Interestingly, two participants used the term “buttoned up” to describe their team, perhaps
in contrast to the volatility of the wider organization. Courtney said:
Across [Nancy’s] org, I think they've done a very good job of just hiring very talented
and buttoned up people. That has been really nice because that's not always the case in a
professional environment. And I think that people are passionate about what they do.

4
This is also a trend within the large media organizations in general, as Chapter 2 chronicled.
70
Courtney sees this professionalism as an asset to her team. However, there is some diversity
among the subgroups within this wider team. Glen described that his part of the large team is
somewhat more relaxed than others:
I'd say maybe other groups [within Nancy’s org] are a little stricter than I am, a little
more like, I don't know if that's the right way to say it, but just a little bit, maybe they're a
little more buttoned up. Maybe that's just me talking about myself more than anything.
Anyway, I think that's the same overall kind of culture within [Nancy’s] organization.
There's also within [Nancy’s] organization, and this comes down from [Nancy], very
much a sense of let's take care of ourselves.
As we shall see in the following section, this sense of taking care of each other was especially
felt during COVID remote work, when Nancy instituted several ways to facilitate mutual support
for her team members.
New Ways of Organizing
More so than at other organizations, participants at Ballast noted that they were working
longer hours and had difficulty separating work and life during the COVID pandemic. As a
result, many of the practices participants at Ballast described to me were focused on maintaining
the culture they had before, albeit remotely. Nancy explained to me how she tried to facilitate
support during those challenging times:
We've gone through a torture test of a year now, and things are still hanging together.
They're stressed, they're strained, and I think a big part of it is the remote working. We
were all co-located in a pretty small space. It facilitated, you know, mutual support and
just a feeling of camaraderie, and you cannot recreate that virtually in my opinion. You
know, I've tried to do things like keep a reasonable number of team meetings going, I
have morning coffee hours on my calendar where people can just drop into my Webex
and talk to me. I kind of try to create what was the old situation of me being in early in
the morning when people knew they could stop by my office and talk if they needed to.
Nancy has tried to recreate what had been working before the pandemic by maintaining an open
line of communication not only with her through the morning coffee hours but also with the rest
of the team through team meetings. Similarly, the team has tried to institute new policies to deal
71
with overwhelm. Grace described how they were experimenting with limiting the number of
meetings:
There's no casual conversation in the hallway or jumping over to somebody's desk with a
question. It's a, I need to set up 15 minutes and talk to you, and your calendar is like back
to back booked. That's a real challenge and something that has been addressed a couple
of times by [Nancy], myself and our group, as well as our senior leadership as well.
There has been a few times with we to try to do no meetings on Fridays, which it's still
hard to do because, I mean, Fridays. Because everything is so formalized without that in
person experience really, really hard.
Much like Neil from Connected described, the team at Ballast felt that communication was
becoming too formal, and their calendars were full of meetings that tried to replicate the in-
person experience. This feeling of calendar overwhelm forced them to find a solution to this
problem. To them, the answer lay in limiting the number of meetings, rather than adding to them,
as Parker had done at Connected.
Other leaders within this team at Ballast talked about giving their teams support, echoing
Nancy’s description of how she made herself available to the team. Glen talked about how he
provides a forum for his specific team to come together for support:
Obviously, whenever there is anything big happened, reach out to the team, maybe to say,
does anybody want to talk about anything, how's everybody feeling? And I think giving
everybody the freedom to complain, I think, helps because nobody's keeping anything
bottled up inside. And like I said before, there are things where it's like, yeah, this just
sucks. I mean, there's things where we can try to think through how can we fix this or
make this better?
Grace decided to compensate for this lack of in-person experience by overcommunicating with
her team, encouraging open lines of communication to ensure that all her team’s work was done
efficiently. She described:
I keep telling my team all the time, but I'm overcommunicating, just so you know, so it's
not that I'm checking up on you. I just need to know for myself what's happening across
the team. Otherwise I go crazy and think something's been lost or has fallen through the
cracks. So I ask the team to proactively tell me where they are at the end of the day so
that I don't have to nudge them and they don't feel like I'm checking up on them.
72
Here, we see that the lack of in-person experience again forced the leaders of this team at Ballast
to find a solution. Grace’s solution to the problem of being removed from her team was
increased virtual communication, basically, at least once at the end of the day.
Along the lines of open communication, many people at Ballast talked about using Slack,
which facilitates quicker communication. Courtney explained:
Everyone, everyone is on Slack now. I communicate with my boss. I communicate with
my senior analyst in emojis. I, and we have like an ongoing chat, and it's easy to. And we
have a, before COVID, we were having a team standing twice a week, we have moved
that to three times a week. So the closer communication has helped us in some way.
Unlike previous experiments to reduce meetings, Courtney’s group has tried to institute a few
more check-in meetings, much like Parker at Connected had done with his team. This is a form
of overcommunication to make sure that certain projects do not slip through the cracks and that
everyone still feels supported.
Hiring and Training
Much like the teams at Bullseye and Connected, the team at Ballast Communications
looks for a mixture of skills. Some examples they gave me, however, were more focused on
filling particular needs. Glen, for example, told me about how he was approaching hiring a
someone at the manager level for his team:
The primary difference being in a manager role or a more junior role, at the end of the
day, I'm looking for bright people. I'm looking for bright people that I would feel
comfortable working with that I think can pick things up. And in those instances [more
junior roles], it's an understanding of I'm more than happy for you to come in and learn.
And part of that is it's my responsibility to make sure you learn and teach. For this role, I
need somebody that already knows how to do the job. And so in that case, you know,
table stakes is your just baseline knowledge of analytics, knowledge of research
methodologies, knowledge of how to communicate, huge, that's a huge piece of it. And
how to communicate with salespeople. That's also very different.
Glen’s emphasis on communication with sales teams echoes Dusty’s comments about how he
trains his team at Bullseye. Both Glen and Dusty recognize that communicating with individuals
73
outside the research and data world is vital. Glen expects that someone at the manager level
would already have these skills, and Dusty makes sure to devote time to training his analysts to
do this with internal teams before moving to external teams.
Aside from particular skills, teams at Ballast also look for individuals who can work
within Ballast’s dynamic and fast-paced culture. Lily explained:
So organically within that process of trying to understand people's work styles, I slowly
broke everyone down and we started talking more and communicating. I think that's also
something I look for in the hiring process. It's really important, I would say, at our
company to be open to change, open to people, open to flexibility and just having a
personality that you can talk to people, you can present yourself, you can present your
work. And I think that naturally attracts a certain type of person to the role.
In the sections above, Bobby mentioned that it is important for people at Ballast to be open to
change, and Grace and Courtney described the open lines of communication on their team. Lily’s
quote here shows that she specifically looks to hire individuals who fit in those environments.
USA Media
I spoke to two members of the business intelligence group at USA Media, both of whom
worked on different subspecialties within the team. The two participants also noted that remote
work was somewhat of a disruption but did not describe it to the extent that the participants at
Ballast did. They noted a lack of discussion with other members of the organization and that
some other people were not as respectful of people’s time and work-life balance. They did,
however, acknowledge that they had learned to cope with the situation well.
Organizational Culture
The most notable aspect of USA Media’s culture—and what distinguishes it from the
other organizations—is that the two participants described its culture as emphasizing work-life
balance. Describing the transition from working in for-profit media to working at USA Media,
Peter said:
74
It took me a while to figure out what was going on and whether or not it was like a trick,
right? Because the fact that, like, nobody was yelling at each other in meetings or the fact
that I like or leaving at five o'clock or that people are like, “Oh, I got that email, but I'll
do it later, it's fine” was like such a shock to me. Like, I was like, that can't be real. I
ended up like just fell back into old habits. Like on the first day, I was still at work at like
6:30 and the aforementioned CMO came by and was like, “You don't have to stay here.
Like it's not, it's not a big deal. Go home.” Right. And just that OK was really powerful.
In other words, this participant described culture shock. Instead of the culture of overwork that
permeates many for-profit media organizations, USA Media—including its leadership, according
to this participant—encourages its employees to have a healthy work-life balance. This is
especially important to parents, as Joan USA Media told me:
I like the culture. I, you know, honestly, I have two young kids, and it's really family
friendly. And that really helps when you're a working parent, especially during a
pandemic which none of us could have predicted. But even before then, right, like it was
a place where, you don't get your work done, and if you've got to leave early to go to your
kid's thing or whatever it is, someone's sick, it's just no big deal because there are so
many people like me.
This participant explains that the emphasis on work-life balance does not come from leadership
but is also understood between coworkers, who respect each other’s time and lives outside work.
Team Culture
This sense of balance and trust carried through into the team culture as well. Peter noted
that there was also a culture of trust in his team that came from leadership. As he explained:
I was brought in by my manager at the time. She was really the first person who in all my
roles really was like, “You're an adult and I trust you to do these things. You don't need
to send your emails to me to read first before you send them out. I will give you these
tasks to do, and I trust that you will do them. Come to me if you have help. If you have
questions, I will help you. But I'm not going to come by your desk every hour and ask
what you're doing.”
As a result, this participant at USA Media felt that he was trusted to do his work as he sees fit,
much like Brenden at Bullseye and Parker at Connected trusted their teams to do their work well.
He saw this as a positive element of the culture and juxtaposed it with previous experiences in
which he did not feel he had that trust.
75
New Ways of Organizing
The culture of USA Media is family-friendly, but at the same time, in the quote cited at
the beginning of this chapter, Joan described the environment as reactive due to several industry
changes. As a result, USA Media’s environment is as fast-paced as some of the others.
Interestingly, Joan and Peter spoke less about some of the challenges and opportunities of remote
work and electronic communication than participants at other organizations did. Neither of them
mentioned using Slack, for example. However, Joan talked about the challenges of adjusting to
work-from-home and mentioned a specific frustration: email chains that had replaced
spontaneous conversations, thereby wasting time. She described a specific incident:
It became an email chain from hell where people were just responding. I couldn't keep
track of what anyone was saying. I was just like, oh, my gosh. And I actually said to
someone, “This is an example where if we'd been in the office face to face, we could
have gotten together for five minutes and avoided this one hundred email craziness that I
just deleted because I couldn't even deal with.” You know what I mean? So there's an
example where I think there have been a couple of situations like that. Where you feel
where I felt like, oh, I've really missed that in person experience.
In this instance, the necessity of electronic communication, including email, meant that
communication became less efficient. It also took up the time and mental energy of people who
were copied on that email chain. However, neither Joan nor Peter described situations that were
geared towards improving electronic communication.
Rather, Joan and Peter described ways that USA Media tried to reinforce its family-
friendly, work-life balance culture. The two had opposing views of how USA Media was
maintaining that culture during the pandemic, however. Joan thought that the pandemic had
disrupted this work-life balance:
I think people are a little now less sort of respectful of people's time than they were in the
office. And I mean, like some people try to schedule really late afternoon meetings and I
just so, like, I'm setting up a boundary. Our son gets home from school about 4:15 or so,
like, I'm not meeting after that, and things like that. And it wasn't like that for most of
2020, and then 2021 it seems like. And I'm like, well there have been more pushes for
76
more meetings, more later in the day meetings. And people have forgotten that there's
still these other things, that we still are living where we work, kind of thing. I mean, I'm
very deliberate about setting up boundaries. Otherwise, you know, you can end up
working for forever. And so in a way, having young kids helps because I can't do that.
Like, I'm like I'm done.
Joan’s words demonstrate that she felt the culture of USA Media had moved away from being
family-friendly, as she saw scheduling late afternoon as not respectful of parents with young
children. In this way, we can interpret her refusal to take these meetings as a small action to
move the culture of USA Media back towards its family-friendly standard.
Peter, on the other hand, felt that USA Media built up an even more family-friendly
culture during the pandemic. He described:
It's [the culture] potentially been even more open, whereas they very quickly were, and
my boss too, she's at the forefront of this. But was like, “Look, we're home, this is all
new, no one knows what's going on, your family comes first, life comes first, we'll figure
it out, let us know.” And that's really been, you know, there's some people that work five
to noon and then five to eight at night or some people just work at night or some people,
you know, it's, let's just call it extreme flexibility.
In Peter’s assessment, giving people the option to have flexible work hours reinforces the work-
life balance culture that USA Media had before the pandemic. Because Peter and Joan told such
different accounts of how USA Media’s culture changed during the pandemic, we see that
culture varies within an organization and even within the same team.
A New Way of Organizing: The Case of an All-Women Team
At one of the above organizations, two women participants talked about issues of gender
impacting their ability to work well together.
5
These two participants spoke about currently
working on an all-women team and how the departure of certain (as they saw it) problematic
men from the team in the past had made their team’s culture and working environment much
better, allowing them to implement new ways of organizing. In this section, I take a deep dive

5
Due to the sensitivity of what these participants told me, I am not naming them or their organization here.
77
into the narrative these two women told me, unpacking how the narrative illustrates their process
of sensemaking.
The first participant, who is a frontline employee, described the background to the
situation as follows:
The males on our team, they were in the leadership roles. One was promoted unfairly.
Actually, unfairly seems like a light way to say it, it was completely messed up that the
one male was promoted over a female. Honestly, that should be an H.R. nightmare….
One of the males on our team who got promoted to a manager position, he was kind of
just like in your face doing mediocre work. That's the other thing which me and my
teammates talk about all the time is the mediocre males… Looking at that male's work,
who got promoted, I was like, how could someone who produces this mediocre work and
doesn't, he doesn't add to the vibe or the culture and he's the one getting promoted.
This participant discussed how, in her eyes, certain decisions on promotion were not merit-
based; also, seeing this injustice made her upset about the way the team was being run. She
mentioned the word “mediocre” three times, which demonstrates that she had made a strong
judgment about that person’s work. She also mentioned that seeing that man’s work triggered a
form of sensemaking for her. Something about assessing his work as mediocre and then seeing
him promoted made her stop and realize that something was not right about the culture of the
team.
The other participant, who is a middle manager, added more context about the situation
before the team was all-women:
When I joined, my manager was a man. And not to say that no people can do this, but it's
just what I think the specific type of person, not very supportive, didn't ever ask how
things were going, if I was overloaded, it wasn't like solutions on how to find it really.
And it wasn't really a focus on how we grow and make you better so much. I actually at
one point was told like, “Oh, you're just not ready for that stuff,” which now I think back
on I was like, I can't believe I believed that. Like, I should have just been saying, “Excuse
you, I can do whatever.”
This participant’s comments illustrate that her previous manager did not demonstrate trust in her
abilities. However, she mentioned that she only retroactively understood these instances as
78
harmful. She said “now I think back on” those comments, meaning that she retroactively
understood that the culture of her team previously was not positive or conducive to her thriving,
rather than realizing it in the moment her former manager told her she was not ready to take on
other responsibilities.
As the pair of participants told the story, the male managers left at a certain point. This
became an opportunity for the remaining women to create a better working environment for them
all. As the two participants describe it, the team’s culture and teamwork changed for the better. It
became more supportive and efficient. The middle manager continued:
Something that I saw more of [when they left] was honestly a deeper understanding of
the team's needs…. Now because I think there's more women on the team and there's
more female managers, I've seen this shift, and I try to do this personally just because of
my past. But I've seen a huge shift towards, “How can we develop you and grow you, and
what do you want? What would you like to get out of this? Do you want a certain title?
Do you want a certain account? Do you want, you know, what is it that would make you
happy in this role or in your career? And how can we help?” Like, obviously, we're still
concerned with growing revenue and the company's goals overall. But in my opinion and
our other managers, it's really like if we don't have a good team and who's consistent,
what's the point? Because it's just going to be that constant struggle. The reason I say that
also is because, like I mentioned, there was a lot of turnover on our team before. I think
someone left probably every three months or four months. Maybe that's too soon. But
there was at least a short period of time where I think three people left in the span of six
months and that was just nuts. And so we haven't had anyone leave the team since, for
like two years, which doesn't sound that much, but we've just had people, like that's a big
deal, trust me.
This participant mentions that the leadership in the team became more person-based and built on
developing employees. It seems that there is empathy for the needs of the team members and
also more mutual support.
6
Not only that, but this participant remembered her past (referring to
the negative comments of her previous male manager) and tried to reverse that now that she is a
manager. Instead of putting her employees down, she tries to build them up, which creates a

6
What these two women participants described to me matches my own experience, as I worked on one all-women
team during my time in ad sales research.
79
better working environment. Noticing that there was less turnover in the team also allowed this
person to retroactively understand that the problem was the old culture of the team.
The frontline employee added that men’s departure allowed other women to succeed. She
explained:
Those guys left for different opportunities, and when that happened, it obviously opened
up opportunities for the females on my team who had gotten overlooked and pushed to
the side who had been on the team for so long and who are two the smartest people I
know and work really hard. When they were finally able, when they were finally given
the jobs and the roles that they deserved on my team, we have flourished. I talk to my
team all the time, like what we have accomplished as this team is all females in the last
year, given a pandemic, is absolutely mind blowing.
This participant views the work environment as much more positive following the departure of
the men because it allowed other women to emerge as important players on the team. Not only
were the team members more supportive, but according to this participant, they were more
efficient and produced better work. As she explained:
I feel like we've made things happen. In a good, positive way, the team has taken a
direction that it could have gone in earlier… I don't know why the leaders who are no
longer here didn't move it in that direction, I couldn't guess. I don't know what they were
doing with their days. But, yeah, having the team the way we are right now is just, we're
just getting shit done and it's like good, good stuff. And that's a really good feeling
because we see it. We see it. We see the ripple effect throughout the company. Our work,
you know, we have metrics around the impact of our work and so you can see the
difference. Which is great. It shows more money for [my company], which is what they
want, but I think that we all have a little bit more job satisfaction because of it.
Even though this participant did not share specifics with me, she claimed that there are internal
performance metrics demonstrating the effectiveness of this team setup. Seeing the internal
metrics again forms a sensemaking event—it made the women, including this frontline
employee—make sense of their work as efficient. Quite probably, these women would not have
told me their story had it not had a happy ending of improving their team’s cohesion and impact
across the company.

80
Conclusion
Early on in this chapter, I established that three general factors lead to cultures conducive
to creativity and innovation: (1) freedom, autonomy, and trust, (2) resources, and (3) open
communication and support. This chapter focused largely on the first and third elements,
demonstrating that participants across these organizations had cultures that were already focused
on freedom and support. However, some of these factors were challenged during the pandemic,
so the pandemic became an opportunity for people to institute new ways of organizing to
reinforce and continue to build these cultures. Participants discussed instituting new meetings,
leaders making themselves available, and everyone using electronic communication like Slack to
facilitate more open communication while everyone was geographically distributed.
One major issue with this lack of in-person conversation is that it meant an end to small,
serendipitous conversations at the office, either with their team members or members of other
teams. Multiple participants noted that the electronic communication forced by remote work
made communication more “formal.” They noted that everything had to become a meeting,
instead of a spontaneous conversation; as a result, their days became packed with meetings,
leaving them less time to do what they needed to do. This gets at the “resources” idea, as time
became a precious resource. While Ballast had experimented with fewer meetings on Fridays,
they noted that it did not always work. Instead, leaders across organizations instituted more
meetings to make themselves available. This may put a strain on people’s time resource but
likely contributes to creating open communication, collaboration, and trust among team
members. As a result, participants expressed mixed opinions about elements like meetings and
Slack, with many saying that both of these took time away from work, but others noting that they
facilitated support and teamwork.
81
Throughout this chapter, we have seen that culture is constantly being built and
maintained. It is not static, as different ways of organizing, like being forced into remote work or
hiring new team members, challenge established norms. For the former, the leaders all worked to
maintain and reinforce cultures that they had established beforehand. They seemed intent on
creating an environment that would allow people to thrive, which is very necessary given the
amount of flux described in the last chapter. For the latter, some leaders hired individuals who fit
with established cultural norms like open communication and adaptability, whereas most focused
on hiring people with specific skills, including baseline technical skills, some knowledge of how
the industry operates, and communication skills. With this all established, in the next chapter, we
turn our attention towards some of the specific challenges these research and data workers are
facing in their work, specific data problems that make up part of the data dilemma of the media
industries.

82
Chapter 4: Sensemaking Data Problems
As Chapter 2 explained, the past decade has brought many changes in the media
industries that have forced media organizations to adapt, including the rise of streaming and the
upcoming loss of data from third-party cookies. In particular, the waning dominance of
measurement firm Nielsen’s traditional television ratings have created two new challenges. On
one hand, it leaves a hole waiting to be filled by new and innovative measurement firms. On the
other hand, those who need to use data to make advertising or programming decisions in the
meantime have to devise methods or workarounds that use imperfect data.
Changes and innovations in the ratings industry and audience measurement have received
some academic attention at the macro level (Buzzard, 2002; Napoli & Andrews, 2008). These
changes have primarily focused on the technical innovations of audience measurement,
especially Nielsen’s PeopleMeter technology, a device installed in panelists’ homes that allows
Nielsen to track their viewing, which have usually been considered incremental or sustaining
technical innovations (Napoli & Andrews, 2008). Chapter 2 chronicled changes in the media
industries and showed that the media industries increasingly rely on vendors aside from Nielsen
to understand their audience.
The issue is that no data are perfect. Chapter 2 explained that measurement firms like
Nielsen and Comscore are unable to measure all viewing, which means that their data are
incomplete. Chapter 3 chronicled the ways that the four cases highlighted in this dissertation are
organizing in ways to implement creativity and innovation, which is necessary because of the
many ways that data are imperfect in the media industries. This chapter is about how media
research and data workers make sense of data problems and thereby implement solutions to
address them.
83
Sensemaking is about organizing uncertainty and ambiguity (Weick et al., 2005). As we
shall see, in many of the solutions to data problems, participants' explanations align with the idea
of simplexity, a particular articulation of sensemaking that emphasizes combining “complexity
of thought with simplicity of action” (Colville et al., 2012, p. 5). The complexity of thought
stems from the judgments research and data workers come to make about the data they use, and
the simplicity of action is necessary because these workers must be responsive to changing
environments, as the previous two chapters described. Weick (1995) argued that sensemaking is
ongoing, and these changing environments mean that sensemaking is never done. Rather,
research and data workers continue to make sense of their work and environments. While many
of the actions these workers describe taking in this chapter are not radical departures from
previous ways of working with data in the media industries, they reflect creativity of thought and
action, which keeps adapting based on changing situations.
A Typology of Data Problems
This section introduces the typology I propose based on typical data problems
participants referenced. To devise this typology, I made note of projects that participants
described to me and categorized them according to the type of data problem and the trigger that
caused them to realize there was a problem. Almost all of these project examples were
mentioned in the interviews when I asked participants to describe a project that they thought was
particularly creative and/or innovative. As a result, participants were primed to think of creativity
and innovation when describing these projects, so they talked me through specific problems and
how they went about solving those problems.
The analysis revealed six major categories of problems: Nonexistent Metrics, Missing
Data, Unorganized Data, Big Picture Data, Data Uncertainty, and New Data, which are
84
overviewed in the table below. The rest of this section explains each type with examples. I give
particular focus to the triggers of these data problems: client questions, market needs, resources,
mergers and acquisitions, and loss of data. In the next parts of this section below, I describe the
triggers within each section. The tables below describe the categories and triggers.
Table 2
Typology of Data Problems
Problem
Category Explanation Trigger
Nonexistent
Metrics
The metric we need does not exist, so we have to create
it from existing data
Client Question,
Market Need
Missing Data
The data we need do exist, but we do not have them, so
we need to acquire them from other sources
Client Question,
Market Need
Unorganized
Data
We have data, but we need to organize them to pull
insights
Market Need,
Resources
Big Picture Data
The world is changing, so we need to collect and
package data to understand it
Market Need,
Resources
Data
Uncertainty
We don't know what data to use to answer a question or
something doesn't look right in the data, so we need to
determine how to proceed
Client Question,
Market Need
New Data
We acquired or have to find new data, and we need to
figure out how to use them
M&A, Loss of Data

Table 3
Typology of Data Problem Triggers
Trigger Explanation
Client
Question
A client asked us a question we can't directly answer
Market Need
We're aware of a market trend and need to find a way to address it before clients
ask us questions
Resources We received resources to figure out a problem
M&A Our company went through an M&A, and we have new internal data
Loss of Data We lost access to certain data/tools and have to go find new ones

85
Nonexistent Metrics
The problem of nonexistent metrics arises when research and data workers are faced with
a question for which a metric does not exist, so they have to create a new metric from existing
data or collect new data. Participants at Ballast Communications and Connected Media Research
both described this problem. Bobby from Ballast explained three specific instances where he
worked to combine data in different ways to create new metrics, and his examples were the most
illustrative of this type of problem.
Bobby’s three specific examples were as follows: creating a time metric, creating a cost
per thousand (CPM) metric, and collecting demographic metrics for set-top-box video-on-
demand. The time metric is the best illustration of the process he describes. He said that his team
recognized a market need to quantify viewing across platforms (like television and digital media
platforms—revisit the discussion of cross-platform measurement in Chapter 2 for an
elaboration). Because the measurement industry has not yet devised an accepted solution,
different organizations have had to fill this market need by devising their own metrics. Bobby
described the situation as follows:
We had to quantify all our viewing happening across all these platforms. There was no
metric that we can use to add, like there was nothing, impressions would not provide it.
Reach was, we can't include, even if we get reach from these different platforms, we
couldn't add them together.
The key sensemaking moment is when they went through the available metrics and realized that
most would not provide them with the information they needed. The metric they did need—a
metric that could quantify who was watching what on all different platforms—simply did not
exist. This triggered the team to try to find a solution. They looked at the data they had and
realized what they could use to approximate the cross-platform metric they needed. Bobby
explained:
86
But what we noticed universally is minutes and time was included across all these
platforms. And I think a lot of people [at other publishers] started doing this, too. So we
realized time, minutes, viewing consumption is something that we can add from
YouTube, from our nets [networks] too, from Hulu, Roku, all these 20 different
platforms. Time, it's something everyone gets and we can get the data. So we took this
metric, we added it together, and we came up with a cohesive story in terms of like, hey,
instead of looking at ratings and reach, the best we can do is time. And this way we can
see, like, you know, 20, 10 percent of the viewing happened on YouTube, 50 percent
happened on our sites, 30 percent happened on Roku. So that was one of the creative
ways to get to a metric that was not in the marketplace.
In order to provide potential clients with an answer about cross-platform viewing, Bobby and his
team looked at what they had available and took the simplest yet most actionable solution:
adding different time metrics to create a new and unified metric for total time viewers spent
viewing content across different platforms. This is one of the clearest illustrations of simplexity
in action. As we see, the problem of Nonexistent Metrics arises when a precise metric a team
needs does not exist and is solved by combining existing data to devise an approximation.
Missing Data
The missing data issue is when research and data workers do not have access to particular
data or metrics, but they know such data or metrics exist. This differs from the Nonexistent
Metrics problem, when the metric needed does not yet exist. Oftentimes, solutions to these
problems involve recognizing a need for data and identifying another source (such as another
data vendor, like Comscore) that has what they need. Participants at Bullseye Platforms, Ballast,
and Connected talked about this.
Dusty from Bullseye and Glen from Ballast both spoke to me about needing to integrate
sales data with other data. Dusty introduced the concept by explaining a historical example he
had worked on with a grocery store: when people swiped that store’s card at checkout, Dusty’s
team received that information and linked it with information in Bullseye’s demand-side
platform. Acquiring purchase data and linking them with Bullseye’s existing data created more
87
insights for the client. Glen from Ballast told me a similar example. He described a scenario
where changes in consumer behavior as a result of COVID-19 impacted the type of data his team
needed to use. He was working on a project for a client in food delivery and realized that the data
source he had was not adequate because it did not capture online transactions. As he explained:
Digital sales for this brand quarter over quarter of each quarter was increasing like 180%,
which makes all the sense in the world when you think about the fact that everybody
wants a delivery pickup as opposed to eating inside in the past year. Putting together a
measurement solution, we went out and we found a data source that was able to capture
things like dollars spent on Door Dash and Seamless and attribute that back to general
media for that brand. So that, I think, is an example of how the changes that are
happening in the marketplace impacted what the measurement needs were, and we were
able to innovate to solve.
This participant had not previously worked with data from Door Dash and Seamless, and he saw
integrating these data into his work as an innovative solution to a problem. Such uses of data
reflect a paradigm shift for research and data work, which has historically relied on data from
Nielsen and Comscore, which are based solely on whether audiences were exposed to content.
Integrating purchase data supplements exposure data with action data.
Neil from Connected explained another example to me. This example has to do not with
needing purchase data but with needing data about advertisements so that Neil’s team could
determine which advertisements were effective. For some background, Neil explained that
everyone sees different ads when watching the same content on a digital platform like Hulu, for
example. Because of this, his team had no way of knowing which viewers saw which ads. His
team recognized that this was a hole in what they could provide the market. As he explained:
Part of the problem to solve the CTV [connected television] issue is we needed a library
of ads to match against, which we didn't have. So we worked with the partner last year to
be able to get access to that library. And then we had to be able to automatically ingest
new ads as they appear. So if [a brand] came out with the ad today, we need to be able to
detect that that ad is new and get it integrated into our library. So the first step was
getting us access to the ads, which we did. Second step was automating and detecting
when new ads appear and making sure that we're able to incorporate that into our library
so we can measure against it.
88
In this case, the missing data were logs of potential advertisements so that Connected could work
on measuring them. This problem was triggered by a market need: Connected recognized that the
way ads were being served to viewers was changing and that they were not equipped to deal with
the change, so they devised a solution. In these three cases, the missing data are only part of the
puzzle. The solution is not only in acquiring the data but in combining those data with existing
data to solve problems.
Unorganized Data
The problem with unorganized data arises when an organization realizes it has data, but
those are data not packaged in a way that makes them useful. Unlike the Missing Data problem,
in the Unorganized Data problem, no new data are needed to make the existing data useful;
rather, the existing data merely needs to be reorganized or reanalyzed. For example, Brenden
from Bullseye mentioned that one of the problems with the shift towards big data was that
suddenly everything looks statistically significant simply because such a large amount of data is
available; the challenge is understanding how to extract insights from these data, which
necessitates statistical analyses aside from traditional significance testing. Many times, the
problem of unorganized data comes from recognizing a market need, but sometimes, an
organization may also devote particular resources to solving such problems.
Peter from USA Media described a situation where he had to organize and package data
that his organization already had. He explained that USA Media identified a problem: data for
each of its local stations existed separately, and there was no easy way to aggregate and automate
that large amount of data. Instead, an analyst had to put together a report manually every month,
but even that report did not allow internal stakeholders to identify trends or make visualizations
of the data. As a result, Peter received funding to work on a project to organize data at the local
89
station level related to particular audience demographics that USA Media targets. Once he
started working on the project, however, he expanded its scope and decided to aggregate data for
all audiences, instead of the specific demographics originally intended. He started by talking to
different business intelligence platforms and worked with them to organize the data. Then he had
to make sure that the solution was working for key stakeholders within USA Media:
I gathered a group of about 12 stations to act as a pilot group to say, like, “Here's what
I'm trying to build for you. In your perfect scenario, what would be included in this
product?” Because the idea was, I can build this myself. And I think would probably be
fine, but if nobody uses it, it's a giant waste of time, so I need some buy in. I need some
people to say, “Look, yeah, this is really going to work for us, and I'd like to see this or
that.” So talking with that group while also building this out, I was able to, you know, I
met with the pilot group probably over, I don't know, it took about six, nine months to get
this up off the ground and working. We probably met about five or six times. And I had,
you know, a kind of a working demo about five or six months in, let them use it, we got a
call, walked them through it. “Here's what it can do. You know, please try and break it.
Let me know what you think.”
The result of this iteration was that Peter created a solution that was helpful for key internal
stakeholders. Instead of his analyst manually pulling reports every month that did not make it
easy to see trends or make visualizations, Peter organized the data so that others could easily
access the data they needed. As we see from Peter’s description, the unorganized data problem
often results in iterative solutions, where research and data workers will consult with other
stakeholders, create a demo or prototype, and keep iterating in consultation with the
stakeholders.
Parker at Connected described another example to me that follows the same pattern. This
example came from filling a market need, rather than from receiving specific resources. Parker
explained that Connected had been in touch with clients about how they were using Connected’s
data, and through those conversations, his group identified a hole: clients wanted to be able to
use Connected’s data to help with media planning, or deciding which ads to run where. This
became a moment for Connected to reflect on what it was doing and what it needed to do. Parker
90
explained that a key issue they identified was that ads are shown to empty rooms about 30% of
the time, which meant that many clients were wasting money on ads that no one was watching.
He did the following to address the issue:
We put a prototype together that would allow us to pull down the client and their industry
to understand where their performance was in the past. We like to say, and we've done a
ton of research on this, that our metrics are predictive of future behavior so we can use
historical data to predict future outcomes. So by doing that, we put together a simulation
in Tableau and in Excel that would allow us to toggle between different inventory
placements based on cost and viewability. So we were able to find areas that had really
low viewability at a really high cost. So the brand is spending a lot of money for basically
wasting air time. And we're able to say, well, if you were to remove those inefficiencies
and take them and place them in high efficient areas, you're going to save cost and you're
going to reach more people and you're going to get a higher viewability score so people
actually have more exposure.
In this case, finding a solution to this problem first involved identifying the problem. Then they
were able to identify that their data did have the solution, but those data were not organized in a
way that made the solution easy to see. As a result, Connected decided to organize those data
into something a client could use on their own, much like Peter created a tool that internal
stakeholders at USA Media could use on their own. As Parker explained, the trigger for
sensemaking started with those client conversations:
Just having those innovative conversations of what are we missing from our current
products, what can the clients not get to and how can we solve for that? And let's look at
all of the different information available to us. Can we come up with a new metric? Can
we come up with a new concept that, yes, maybe it's not perfect right now, but at least
gets us one step closer to connecting with the brand and connecting with that network?
Client conversations helped identify the market need, which triggered sensemaking. In this case,
the action from the sensemaking was in determining the best solution to organize data. Much like
what Peter described with USA Media, the solution lay in iterating through conversations with
stakeholders.

91
Big Picture Data
The big picture data problem occurs when there is a general trend or event that an
organization wants to understand; in order to understand it, research and data workers will have
to create reporting that often combines different data sources. Five different participants at
Ballast mentioned three different big picture projects they had worked on. In these cases, the
triggers were market needs; individuals noticed that something in the market or the world was
changing, and they did not yet have data organized to address
it.
One of these projects was a tracker related to the advertising-supported streaming
landscape, a growing segment of the market.
7
Sean explained:
The challenge with that is there's not one great syndicated data source that gives us all the
insights we need, whether that's from viewership standpoint or from spend standpoint, we
don't really have a good source that covers it all. Like a Nielsen does cover it to some
degree and other sources cover different things, but there's not one go to source that
really captures the marketplace. So because of that, we kind of have to be crafty and
cobble together information that we do have from different sources as best we can so that
it can be summarized, and the insights hopefully can be shared with key stakeholders,
specifically within sales. So what we did, so we kept getting asked for things, we were
doing things piecemeal. A big project that we worked on at some point last year was
building a report that just focuses on the ad-supported streaming landscape. That report,
again, like it's like it's us cobbling together a lot of different information and packaging it,
so that it kind of seems like it's all one source providing it, it's coming from a lot of
different places, so that can be from, like I said, Nielsen, it could come from investor
reports or just the trades.
Sean explained that some of the sources besides Nielsen include the IAB (Interactive Advertising
Bureau) and market research sources like Magna Global and eMarketer, for example. But the key
idea here is what triggered Ballast to start creating this report. As Sean explained, his team kept

7
While the advertising-supported streaming market was already growing at the time of the interviews (due to ads
being available on services like Hulu, Peacock, Amazon Freevee, Tubi, and others), expect this area to grow even
more now. At the time of writing this dissertation, Netflix had just announced that it would begin experimenting
with advertising.
92
receiving questions about ad-supported streaming through ad hoc requests; after answering
questions piece by piece, they recognized that this was a greater market need that demanded
regular reporting.
Data Uncertainty
To continue with Sean’s ad-supported streaming tracker for a moment, he specified that
he needed to make sure that the information looks like it came from one source. However, this is
not always easy. As a result, this project also illustrates the problem of Data Uncertainty, which I
define as the issue that arises when research and data workers do not know what to make of
certain data or feel that something is off with the data. Sometimes, this might also mean that
research and data workers do not know what data to use out of many options. In the ad-supported
streaming tracker example, Sean explained that he sometimes sees conflicting information. As he
explained:
eMarketer just put out a forecast that forecasts US ad revenue for [a rival’s ad-supported
streaming service]. The forecast I was using was from MoffettNathanson, another
research firm. And those two forecasts are totally different from each other. I mean,
they're not wildly different, but they're different enough where it's something to think
through. You kind of have to think through which sources were credible. So sometimes
two different sources can be wildly different. So that's part of the challenge on my end is
like sifting through all that and trying to figure out what's the right source, or do we just
average the two sources together?
Sean articulated the moment that spurred sensemaking as such: he noticed that two forecasts
were different and needed to determine which one he thought was correct. Simplexity reduces
ambiguity through moving from “what might be going on” to “what is going on” (Colville et al.,
2012). In this case, the process Sean undertook here is the bridge between “one of these sources
might be right” to “I am going to treat this source as right,” even if some uncertainty remains. I
asked Sean if he looks at the data collection methodology or something else when making these
judgements, and he explained:
93
I guess I am paying attention to the methodology. And it's sort of like an eyeball test too,
just like what's right and what doesn't. It's kind of hard to explain how I do that. I mean, I
guess I start with what I have. Like, if I have a source that I'm comfortable with, that's
sort of my base, and then if I get another source that says something different, then I dig
into the language they use and how they may have arrived at that number, and you just
start reading it, like what seems like a better methodology? But there's definitely a lot of
gray area.
In Sean’s explanation, the process is a gut check. The uncertainty in Sean’s language—“I guess,”
“sort of,” “it’s kind of hard to explain”—shows that Sean had trouble articulating exactly how he
comes to determine which data sources are correct. He illustrated a complex thought process he
uses, but ultimately, he had to come to a simple solution: one or both of the sources had to be
seen as correct, and in the latter case, he uses a simple average of the two. Weick (1995) argues
that sensemaking is “driven by plausibility rather than accuracy” (p. 17). In this example, Sean
worked through ascertaining what he determined was plausible, rather than what was actually
accurate.
There are two other examples of data uncertainty I wish to highlight. The next is an
example Brenden from Bullseye told me about data not looking quite right to him. Brenden gave
me the background to this situation by explaining that many of Bullseye’s clients would ask for
information about their audiences. Brenden said that they would commonly answer these
questions by ranking data elements related to these audiences, but after looking at this ranker, he
acknowledged something looked off. Brenden explained:
I remember looking at it and saying, this is kind of weird because it gives you things that
are contradictory. It gives you people who are old and young, people who are rich and
poor, people who love dogs and love cats, and that's not mutually exclusive, but you'd
find all these weird things where you're like, this can't be one person, right? You can't be
one, like that doesn't make sense. And so as I thought about it, I was like, well,
realistically, everybody's trying to boil this down to an audience because that's actually
one of the ways people speak about their buyers. They'll say, who are my buyers, my
audience of buyers? And I said what we've actually got here is an audience made up of
subgroups. Right? It is one audience of buyers, which is actually many smaller, distinct
audiences, all of whom have buying in common.
94
Much like the example from Sean above, the key idea here is the moment when both Sean and
Brenden realized something was off in the data they had. In this case, the moment came when
Brenden realized that the data grouped people together as though they were one type of person
when in reality they were many types of people. The solution, as Brenden explained it to me, was
not the gut check that Sean described but instead a more ordered solution. Brenden explained to
me that he saw this problem as a data science clustering problem, so his team built out an
analysis to determine the specific clusters of buyers within the data. As Brenden explained it,
such an analysis was not yet popular in advertising research. But his team did not reinvent the
wheel: rather, they found an existing data science method and applied it in a new context. Again,
this is simplexity; Brenden’s team carefully thought through what the problem was, but they
needed a solution, so they adapted an established data science technique to fit their needs.
To continue with data uncertainty, I want to highlight an example of not knowing what
data to use, which is similar yet different from the other examples. In the previous two examples,
Sean and Brenden realized that something about the data they needed to use was off, and they
tried to determine what to do. In this example, however, Adria was faced with a question for
which she faced uncertainty about what kind of data to use. As she explained, Bullseye started
selling audio ads, but she did not have any audio data. Unlike in the Missing Data examples
described above, Adria in this example did not go find a new source for audio data. The Data
Uncertainty problem is more about trying to work through problems with existing data where the
solution is not readily apparent. Rather than looking for a new vendor for audio data, Adria
found a proxy. She described her process as follows:
The big problem was I one didn't know a lot about how audio worked and how audio ads
are sold through. So I had to figure out what are all the different, and this isn't the
exciting part obviously, but the first step was figuring out what is everything you can
target. And so really, it boiled down to, well, you can kind of just focus on, say, like a
95
genre was a big one, I think, light demographic information. And I believe that was really
all we had available, so I was like, OK, there's really not much I can do here, but they're
asking me to provide some sort of direction. So finally, what I figured out was I think it'd
be a good proxy to look at.
The key issue here is a client asked a question that Adria did not know how to answer because
she knew she did not have what the client needed. That client question triggered her to think of
what she could do, so she began by thinking through what data she had: genre and demographics.
Then she went about devising a solution:
What I ended up doing was looking at what sort of music people are looking at, and I had
bucketed it into different types. So I pulled all the popular songs, all the popular artists
within a certain genre, created an audience that that client was interested in and then saw
how did that audience skew towards each bucketed group I had made, just to pull in some
sort of metric here. Also having to exclude anything that was just trending within the
normal audience just to normalize the data so I could compare it to like, is this one
actually special or is it that some celebrity just did whatever and so that's why it's
trending right now. So that was one of my things where I was like, it's kind of this weird
creative issue where you're struggling with you just don't have the right data, honestly.
Essentially, Adria described a creative workaround. While she could not deliver exactly what the
client wanted, she was able to devise an approximation using the type of audience they were
looking for and was able to determine music they liked. Adria’s solution may sound simple, but
this is exactly the point: especially in a case like this where it was a client question, Adria needed
to find the quickest solution possible. Simplexity arises in chaotic situations, and the responsive
work of responding to clients necessitates simple action. Adria’s solution was nevertheless
creative because it involved looking at what data she had and trying to determine how to use it to
solve a problem.
In the previous two examples, both Brenden and Adria used the word “weird” when
describing their problem. Brenden used it when he described realizing the data were off (“this is
kind of weird” and “you'd find all these weird things”), and Adria used it when describing the
beginning of the data problem (“it's kind of this weird creative issue”). In both cases, it was the
96
“weirdness” of the problem that triggered the sensemaking process. Brenden also used the word
“weird” when describing how big data make everything significant, which I referenced earlier. In
this way, a gut check of “weirdness” helps research and data workers name and label issues that
begin sensemaking, so that they can go back and make sense of these problems.
New Data
The problem around New Data occurs when research and data workers have a new
dataset and have to determine how to use it. There are a few possible triggers. The first, for
which I did not have an example, is when an organization might purchase a dataset and ask
individuals to ascertain how it could be useful. I remember this happening a few times at
NBCUniversal, as people would be tasked with finding creative ways to use a potential new data
vendor. Second, through mergers and acquisitions (M&A), individuals at one of the original
organizations might have access to the other organization’s data and have to find use cases for
them. Third, an organization might lose access to a certain source and have to find another
source to fill it, thus trying to determine how best to use the new source. Below, I provide one
M&A example and one loss of data example.
For the M&A example, I will highlight an example from Dusty at Bullseye. Bullseye as
an organization has grown a lot over the last several years through acquisitions. Dusty himself
started working at a company that was acquired by Bullseye. He explained that Bullseye then
completed an acquisition of yet another company, which was an opportunity for him to create a
report combining the different data:
As our company acquired [another company], we have this access to sort of very good
allocation algorithm behind us… I'm currently working on a project [that] basically
allows the advertiser to look at their linear buys, who they reach in the linear buys, who
they reach in…linear buys, desktop buys, CTV buys, and the mobile buys. So like all of
this digital different channels altogether and see how they overlap with each other.
97
Similar to Sean at Ballast trying to tie together data sources to produce the ad-supported
streaming report, Dusty at Bullseye created a new report that ties together the company’s data.
Having the new dataset allowed him to realize that combining it with existing data would
produce useful information for Bullseye.
Similarly, Joan from USA Media, who specializes in custom or primary research like
surveys, explained a situation when she lost access to a certain data tool that she would use when
running surveys. When a project in which she would have had to use that tool arose, she went
looking for another tool to use because she knew she had to do the project differently. She found
a new tool that allowed her to build even more customization into the study she needed to run.
As she explained:
I uploaded these images for the four ads or key art pieces we were testing. Super easy. It
looked great, like it resized it and everything. And so they've got a lot. The technology is
there. They're trying to make my life easier as a researcher so that then I could focus on
the analysis. The other thing I really liked is all the analysis is there, is automated as the
study goes… So for us it may sound kind of silly, but it was doing something different.
We just haven't had this chance to do something in a different way and be more flexible.
But now that we don't have this [research tool], we're open to being able to use new
people who have really cool technology.
This is slightly different from the Ballast example in that this case was a project that the
participant could have done easily with the previous data source. Instead, in the absence of the
previous source, she found a new data source and then discovered how much more she could do
with that. It made her think creatively about new potential uses for the source.
Solving Problems: How to Tell When Something is Creative or Innovative
Many of the participants whose projects I highlight above undertake the process of
sensemaking called simplexity: finding a simple solution through complex thought to organize
chaos. As a result, many of the solutions they devise may not seem all that innovative. As a
reminder, many of these projects I describe came from me asking participants about a project
98
they thought was creative and/or innovative. In each case, after they answered, I asked them to
specify whether they thought the project was creative, innovative, or both and where they saw
the line. More often than not, participants thought their projects were creative or a mix of
creativity and innovation, though a few thought they were innovative. This reflects a conflation
of the terms “creativity” and “innovation.” What ties participants’ uses of the two terms together
is a focus on something entirely new. Following this idea, this section unpacks how participants
come to see a project as creative and/or innovative—or to use other words, when does a project
become a “big deal” to an organization? Based on the participants’ responses, the answer lies in
how projects are received by clients, both external and internal. Below, I give examples of each
to illustrate the moments when participants realized their projects were well-received by clients
and how these moments become opportunities for sensemaking.
External Clients
Earlier in this chapter, I detailed how Brenden and his team at Bullseye determined that
they could answer a clients’ audience questions by using clustering analyses from data science,
thereby illustrating that audiences were made up of specific subgroups. Brenden knew that the
data science was sound, but seeing that his clients found the analyses useful made it clear just
how much of a big deal it was perceived to be. He explained:
We developed this, and it was super popular with our clients, like being able to show
them this and say your audience is actually three different audiences or it's actually five
different audiences and this is what they all look like. This was hugely eye-opening for a
lot of our customers and they started to think about things really differently, like, “Oh,
wow, if I have these three different groups for my buyers, then will then which one of
those groups is more profitable for me or which one is bigger? Or should I have three
different ad campaigns that speak to the different reasons, like everybody could be
buying my thing, but these three different groups might have three very different reasons
for why I was their choice. And so if my advertising is only speaking to one of them, then
maybe I'm missing out on the other two.” Right. So this became a very, very popular
analysis for our team to do. And we developed a very streamlined way of collecting this
99
data, doing the analysis, presenting the findings to clients and then having some next
steps.
Brenden’s description of the analysis being “hugely eye-opening” for clients shows how he
understands how valuable this analysis was to clients: they could understand their audiences in
ways they could not previously. Brenden also explained that clients kept asking his team to do
this analysis, a further mark of how well-received it was. Brenden spends over half of the above
quote talking through how he felt his clients were beginning to think differently based on this
analysis. We can say that hearing these client comments became triggers of sensemaking for
Brenden; as a result of hearing how useful it was, Brenden enacted change by working with his
team to streamline the analysis so that they could produce it more often and more easily to
clients.
Internal Clients
Two project examples from USA Media demonstrate how a project becomes important to
an organization. These examples are about how individuals within an organization recognize that
a project is valuable. I give more context about two projects I chronicle above: Joan’s use of a
new data analysis tool to use for surveys and Peter’s example of building out a tool that
aggregated and organized local station data.
In the survey example Joan described to me, she explained that USA Media had lost
access to a tool, so she had to find another tool to carry out a survey project she needed to run.
Joan said:
For us, it was a big deal. I felt like it was a moment when like okay, aha, people finally
get it, that we can try something new and we can get the data that we need to help make
the decision.
As I described in the last chapter, Joan felt a couple of times that people within USA Media did
not react well to change; for example, during the COVID-19 pandemic and remote work, people
100
kept using email instead of finding a better way to communicate electronically, which became as
she said an “email chain from hell.” In this example, however, she described a situation when
others within USA Media recognized and were open to change. Instead of insisting that surveys
were run and analyzed the way they had been with the old tool, internal stakeholders were open
to the solution; recognizing this openness made Joan realize that it would be possible to
implement new ways of gathering insights in the future.
The second example is the aggregated local station data project that Peter described,
which I explained above. As a reminder, Peter piloted the tool with some local stations, built it
out, and then released it to the rest of the organization. This is how Peter described the moment
when he realized the tool had been useful to others within USA Media:
This is the first time that [USA Media] especially Business Intelligence had rolled
anything out like this. And the fact that it works, people are using it, people are excited
and curious about it, that it provides them this data, that it allows people a level of self
service that they never had before, it allows them to, it gives them a depth of data that
they never had before, has been really beneficial. And as we continue to roll out new
additions, new additional data sets, new ways to look at data, people just continue to get
excited and find it to be extremely useful.
In Peter’s view, seeing others within USA Media actually using the tool he built made him
realize that this was a big deal for the organization. Much like Joan described above with people
understanding that USA Media could adapt and use new tools, Peter here realizes that USA
Media could embrace change, which is often difficult for a legacy organization, as Chapter 2
discussed. Peter continued to describe the realization as follows:
I set up a base camp to go along with it, as ideally the idea was, it was to build a
community internally of [the BI platform] users that could ask each other questions and
share best practices. Mostly it's just been an additional way for people to ask me things
and email me. So with that, that's fine. But the level of engagement there, right, I get a
couple of questions or a couple posts every day. But really, a month ago I finally got after
like three years of this thing being up and alive, someone finally said, “Hey, I'm new, I
just found this, it's really cool, but what do I do with it? Can you please help?” And I got
like six or seven people from the community saying, “Hey, I used it to do this or I've
101
done X with it and it's been really helpful.” And I was like, yes, finally, this is what I was
waiting for.
To Peter, the key sensemaking moment was when another user of the tool within USA Media
helped a new user, instead of relying on Peter to help all new users. While Peter likely already
understood that the project was a big deal to USA Media because of how many people had been
engaging with it, this moment showed Peter that others within the organization saw it as a big
deal because they were using it enough to be able to help others learn it. In fact, although Peter
does not use the term, what he is describing is a community of practice, a group “of people
informally bound together by shared expertise and a passion for a joint enterprise” (Wenger &
Snyder, 2000, p. 139). In this case, the group is a collection of individuals across USA Media
who access local station data via Peter’s platform. In the examples above, Peter describes the
moment when the community of practice matured in its collaboration: they were helping each
other solve problems by sharing knowledge and best practices with other members of the
community.
Conclusion
As this chapter has shown, even projects that do not seem all that creative or
innovative—such as measuring time spent viewing, combining datasets, or creating new
reports—are seen as creative or innovative within these organizations. This follows from the
provocation that innovation in audience measurement is usually thought of as incremental
technical service innovations (Napoli & Andrews, 2008), rather than something radically
innovative that would rock the boat. As Napoli and Andrews’s (2008) study illuminated, this
categorization was because radical changes in measurement changed the way that stakeholders
understood their audiences—and often not in ways they wanted to see. With that established, it is
102
somewhat risky to implement more radical solutions, especially in an environment as client-
driven as research and data work.
This chapter has illustrated how research and data workers make sense of their work as
creative and innovative, more so the former than the latter. This is the usual form of sensemaking
in Weick’s (1995) explication: it is retrospective and ongoing. More specifically, a lot of these
examples are along the lines of simplexity, which Colville et al. (2012) defined as necessary in a
dynamic, fast-paced environment. As the previous two chapters explained, the media
environment is about as dynamic, fast-paced, and (as participants said) reactive as environments
come. Retrospective sensemaking allows research and data workers to identify what does not
make sense, like all that data that looked off, and work on ways to think of solutions—or quite
literally to enact solutions. By contrast, the next chapter takes a step away from the minutiae of
this day-to-day work by examining future-oriented sensemaking to try to understand how
research and data workers imagine their future.

103
Chapter 5: Sensemaking Digital Divides
The past two chapters have described how research and data workers have reacted to the
changing media environment through new ways of organizing and new ways of using data. In
both cases, these workers were meeting a present challenge. In this chapter, I turn my attention to
how research and data workers are anticipating the future. This is part of the trend of future-
oriented sensemaking, which is often concerned with creating new plans to address a future
challenge (Gephart, Topal, & Zhang, 2011). The future-oriented sensemaking of research and
data workers invokes issues related to digital divides. As the media industries become
increasingly datafied (Arsenault, 2017), digital divides will increase. As I introduced in Chapter
2, one of the key changes on the horizon at the time of this writing is the upcoming deprecation
of the third-party cookie, which will result in a disruption to identity, or the ability to gather data
about particular media users to serve them ads. Talk of this deprecation unearthed issues related
to two digital “divides”: between players in the industry who have and do not have data, and
between the industry and consumers whose data are collected. Discussion of the latter often
depicted consumers as having but not expressing agency; visions of this change were framed as
co-opting privacy concerns into business practices that would increase digital divides. Alongside
these changes, participants also discussed the upcoming digital divide in skills, as many
anticipated how their work might change as data science becomes more prominent.
This final chapter thus takes a future-oriented approach, as participants grapple with
changes on the horizon. Participants took transitions—the clearly-defined moment of the third-
party cookie deprecation and the less clearly-defined but no doubt perceptible shift towards data
science—to prospectively make sense of their work as more digitally divided. To begin, this
chapter gives an overview of the literature on varieties of digital divides. It then turns its
104
attention towards the third-party cookie deprecation, explaining how it is a sensemaking moment
and discussing how it exposes new digital divides. The chapter concludes by discussing the
future of research and data work as it relies more on data science, perpetuating a digital divide in
skills.
Digital Divides in the Media Industries
boyd and Crawford (2012) argue that big data creates new digital divides. These can
occur at different levels. One of the key digital divides related to data use is related to data
collection, as many media users have their data mined and used in ways that they feel powerless
to prevent. However, another digital divide is between entities in the industries that have the
ability to create such power and entities that do not—and subsequently, between individuals who
understand how to use those data and individuals who do not.
Digital Divides: Between Industry and Audience
As related to media industries, data divides are normally discussed in relation to privacy
and surveillance, such as how media users have their data mined and in turn may not know how
their data are used (Andrejevic, 2014). When media organizations rely more on data to make
decisions, they run the risk of crossing ethical lines, especially since much of big data relies on
using surveillance techniques. One of boyd and Crawford’s (2012) arguments about big data is
that they are not necessarily ethical even if available. Issues of surveillance have always been
part of the business of television broadcasting due to how ratings, including Nielsen’s television
ratings, were and still are collected (Petruska & Vanderhoef, 2014). The availability of big data
from sources including the set-top box have increased the capacity and accuracy of these
surveillance techniques, and media audiences have become accustomed to trading their data and
privacy for increased personalization (Petruska & Vanderhoef, 2014). This has led to an era of
105
surveillance capitalism, in which these monitoring and personalization techniques are
commoditized (Zuboff, 2015).
In particular, this raises new ethical debates about data collection and user agency. There
exists a so-called “privacy paradox”: users claim to care about privacy but then do not implement
any privacy protective behaviors (Hargittai & Marwick, 2016). Some studies suggest that a
divide in digital skills may account for different privacy behaviors, as those with a greater
understanding of the internet and digital literacy in general are more likely to protect their
privacy online (Büchi, Just, & Latzer, 2017; Park, 2013). However, many media users feel
defenseless against protecting their data because so many online services lack transparency and
require users to agree to data collection for the services to function properly (Andrejevic, 2014).
As a result, many users genuinely do not believe they have any control over the situation, which
leads to feelings referred to as privacy cynicism and digital resignation (Draper & Turow, 2019;
Hargittai & Marwick, 2016; Hoffmann, Lutz, & Ranzini, 2016). Many researchers suggest that
corporate interests create this feeling of resignation, as it is in their interest for people to hand
over their data, thereby perpetuating an industry-audience power divide (Draper & Turow,
2019).
Digital Divides: Within Industries and Organizations
The big data divide may also relate to divides within the media organization industries
themselves. For example, big data allow for the automatic creation of news stories, which means
that some traditional journalists have lost their jobs (Carlson, 2015; Cohen, 2015). Somewhat
less dramatically, there is also increased demand for individuals trained in data science and
visualization in news organizations (Kirkpatrick, 2015; Reilly, 2017). The same phenomenon is
seen in advertising; this shift has been characterized as a shift from the “mad men” to the “math
106
men,” where the individuals who understand how to use data to make decisions rule supreme
(Auletta, 2018; Deighton, 2017). This implies a new digital divide between individuals who have
data training and those who do not—and a shift in power towards the former.
Within industries, some organizations keep data on their users for themselves and do not
share those data with other industry players; these are often called “walled gardens” because
their ecosystem and data are essentially walled off from the competition (McDermott, 2015).
These walled gardens know a lot about their users. For example, streaming services like Netflix
have granular data about their viewers, which means they are able to better-target and better-
serve these audiences, as they did in the House of Cards example (Havens, 2014; Smith &
Telang, 2016). Furthermore, larger media organizations might be able to complement traditional
sources like Nielsen by purchasing access to a wider array of syndicated data sources, ranging
from television ad measurement data from iSpot.tv to social media data from ListenFirst, for
example. As we saw in the last chapter, purchasing access to new datasets can help solve creative
problems, meaning that they would be able to leverage these data for competitive advantage.
A theoretical concern that develops from these questions of digital divides is that big data
can act as a form of capital. In fact, big data bring power to those who have them (Iliadis &
Russo, 2016). Similarly, while data are usually discussed as commodities, data are also capital
and fit alongside Bourdieu’s ideas of social and cultural capital; like those two, it is possible for
data capital to translate into economic capital (Sadowski, 2019). In particular, data allow those
people who use them to target users, to optimize practices, and to manage and predict outcomes
(Sadowski, 2019). An example is the rise of programmatic advertising, where real-time user data
helps drive automatic advertising (Kininmonth, 2021; Malthouse, Maslowska, & Franks, 2018).
These ideas have led to so-called “data capitalism,” in which user data are commoditized,
107
leading to a power redistribution towards those with the power to collect and use such data
(West, 2019). The overarching concept from all of these ideas is that whoever has data has more
power, which creates new divides between the data-rich and the data-poor.
Sensemaking Third-Party Cookie Deprecation and the Privacy Debate
In this section, I illustrate how the end of third-party cookies serves as a significant
moment in time for research and data workers to participate in sensemaking about the future in
ways that may perpetuate digital divides. This section mixes information from the two industry
conferences I attended with information from my interviews of employees at the four case study
organizations.
The End of Third-Party Cookies as a Sensemaking Event
I introduced the concept of the deprecation of the third-party cookie in Chapter 2, where I
briefly noted the amount of anxiety the upcoming changes were causing practitioners. Here, we
see that the end of third-party cookies has become a triggering event for sensemaking, even
though it remains in the future. As Weick, Sutcliffe, and Obstfeld (2005) explain, sensemaking
involves noticing that an event is different, bracketing it for further thought, and labeling it to
find a common understanding and move forward to address it. Events that are ambiguous or
uncertain are often triggers for sensemaking (Weick, 1995; Weick et al., 2005). In the case of the
third-party cookie deprecation, research and data workers have understood that there has been a
preliminary event: an announcement of upcoming change has been made, and they are trying to
understand what it will mean. Attention to events can create sensemaking (Nigam & Ocasio,
2010), and the deprecation of the third-party cookie has received a lot of attention both within
organizations and at industry-wide conferences.
108
A key idea that has emerged is that this change will lead to a loss of data for
organizations that are not Google and Apple, bolstering a digital divide within the industry,
which I will elaborate on in the next section. This was especially true of participants at Bullseye
Platforms, which as an advertising technology company is more affected by these upcoming
changes than the three other cases. Adria from Bullseye Platforms tried to understand what the
deprecation of the third-party cookie would mean for her work:
The biggest game changer and the big thing people are working on the most is really
figuring out, with the cookie going away, what are we going to do? How is it going to
affect the business? Are we going to see changes in revenue or is it one of those things
where everyone's going to easily shift to contextual or different targeting tactic? And
personally, I don't know. I don't know how it's going to roll down. Our team has been
really reading a lot of articles and talking a lot about it. But it's one of those things where
I feel like I don't have too informed of an idea about to predict where it's going to go.
And truthfully, I think a lot of us are just kind of going about our day to day just to keep
our sanity and see what happens.
As she explained, the end of the third-party cookie will change how the business operates.
Bullseye has begun a process of sensemaking about this change, as Adria mentioned that her
team has been having lots of conversations about the issue. Other participants from Bullseye
mentioned that the company has organized company-wide events to make sense of this change,
pointing towards the fact that Bullseye has labeled the cookie issue as one to address.
As part of their sensemaking process about the third-party cookie deprecation,
participants used language involving uncertainty and “end times,” demonstrating that they
ultimately imagine an unfavorable outcome. This use of language aligns with Bruskin and
Mikkelsen’s (2020) findings that future-oriented sensemaking envisions a negative future
through metaphors that invoke uncertainty, war, and The End. In Chapter 2, I included a quote
from Brenden, the head of the team at Bullseye, who used highly dramatic language like
“everyone’s going nuts” and “people are paralyzed with indecision” to describe the situation. In
a quote above, Adria said she and her team are trying “to keep our sanity.” Other participants
109
from Bullseye and USA Media used additional end times imagery: Ken from Bullseye described
the situation as the “cookie apocalypse,” which aligns with the advertising trade press calling the
event “cookiepocalypse.”
8
Dusty from Bullseye simply said, “We have this identity crisis.”
Brenden said from Bullseye at one point, “With Google doing this in Chrome, it is throwing the
whole concept of identity into upheaval.” Peter from USA Media referenced “the death of the
third-party cookie” as a major disruption. These participants used language with negative
emotions—“paralyzed,” “apocalypse,” “crisis,” “upheaval,” and “death”—to describe the
upcoming cookie problem and how they and their peers are approaching it.
These individuals have thus envisioned that the end of third-party cookies will make their
work more difficult or perhaps impossible. The upcoming deprecation of third-party cookies has
shaken them out of their routine and forced them to think about the future of their industry.
Individuals in the industry are undertaking a process of sensemaking to create a vision that will
guide what their response to this challenge will be or at least should be; they are beginning to
enact the future, a key facet of sensemaking. By envisioning the future in negative terms, these
individuals have created a future in which the answer to third-party cookie deprecation is a death
to an established way of advertising: the end of using data from third-party cookies to deploy
targeted ads. The answer they have envisioned is a return to less precise ways of targeting,
perpetuating a digital divide between them and data-rich organizations like Apple and Google.
As individuals I interviewed and individuals at the two conferences I attended explained, a
prevailing idea is that advertising targeting might migrate to “contextual”—or targeting ads
based on the context where they are placed, rather than the intended audience. For example, a

8
For examples, see Brooks (2021) in AdExchanger and Shields (2021) in Ad Week.
110
contextual ad placement would be running advertisements for skiing products in a winter sports
publication.
After Adria from Bullseye explained that she and her colleagues are just “trying to keep
our sanity” and plod along with their work, I asked her to explain a little more about how the
third-party cookie deprecation would impact her work, and her answer illustrated how she and
others imagine a future based on contextual targeting. She answered by giving me a hypothetical
example of a project for an ice cream company. She explained that currently she could use
cookie data to create an audience target of ice cream lovers, or people who engage with a lot of
ice cream content online; these ice cream lovers can then be served ice cream ads even when
they are not on ice cream websites. She then switched to a scenario in which third-party cookies
did not exist, demonstrating that she had imagined a future without cookies. Without cookie data,
the targeting tactic she described before would not be possible. As she explained:
If there's no cookie data, that means that [Bullseye] or whoever is collecting that data is
not able to track that person and profile them as an ice cream lover or at least and
probably not accurately. And so with that, it really just takes away that entire option of
targeting. So the only way to reach them is through contextual. And the only way to gain
insights on them would be through, say, like we have some panel data that we're working
with. So it's just going to really take away a lot of our data, which the less data, the less
accuracy.
Even though Adria described a future scenario, she used the present tense, as though she were
imagining the future happening now, putting herself in her hypothetical future shoes. Her answer
to the problem—a shift towards contextual advertising—is one that others in the industry have
noted. For example, two panelists at the Cynposis conference explained that contextual
advertising never fully disappeared, especially on television, and was pushed aside in favor of
cookies, so they project that it is going to make a comeback as cookies are deprecated. As
Gephart et al. (2011) explain, part of future-oriented sensemaking involves a past orientation
with “selective reactivation of past routines and habits to provide stability and order in social
111
life” (p. 280). The expectation is that past routines will still work to some degree in the future. In
this case, the past routine is a return to business practices of an era before third-party cookies
were available, meaning that organizations without robust first-party data will be at a
disadvantage.
New Digital Divides: Privacy and Walled Gardens
To be clear, this shift in data access creates two new digital divides. First is related to the
issue of organizations like Bullseye losing data: as organizations like Apple and Google block
third-party cookies, they have data and others do not, leading to a digital divide between those
who can do good targeting and those who cannot. Second, talk of cookies relates to ideas about
privacy management and the digital divide among media users. As I argue in this section, these
two debates are intertwined: a concern for user privacy has been co-opted for interests that will
increase digital divides among data-rich and data-poor.
To begin, a prevailing view among industry practitioners is that media users have agency
to control their privacy but do not exercise it. Stephen, a measurement industry veteran who
spoke to me after the ARF conference, said:
I don't think it [the end of third-party cookies] will affect consumers. My beef is that
consumers are their own worst enemies because of the end user licensing agreements, the
as they're called EULAs, the EULAs that you sign. Where you give consent without
reading it. You know, none of us read the EULAs, we just say, you know, yeah, of course
I'll do it. And the EULAs are what gives the power to trace our GPS movements or
whatever happens to be buried in the fine print.
This view of the consumer as “their own worst enemy” who signs away their rights means that
Stephen sees consumers as having some agency and are nonetheless choosing to sign their
privacy away. This sentiment ignores the fact that users who are not as digitally literate do not
understand how to exercise this agency and control their privacy (Büchi et al., 2017; Park, 2013).
A certain segment of research and data work thus views it as users’ responsibilities to advocate
112
for themselves and exercise agency to protect themselves. Skyler, the data scientist from
Connected, made a similar philosophical argument, reflecting a view that people can and should
have more agency than they do:
If we think about what the Greeks were trying to say, we're always in pursuit of the good
life. And I think that our definition of the good life right now is to be at home and to be
lazy and not have to think much. And if that's the good life that we want, that's the track
that we're going down. I don't necessarily agree with it. I think that the more that we
remove that, the more that we're going to ask people to start thinking for themselves
again and that we're going to ask people to start making their own decisions and form
their own opinions again.
Skyler’s view suggests that education is the solution. Along the same lines, some of the
conference speakers discussed privacy by talking about the need to “empower” consumers to
become better stewards of their data by educating them, reflecting a view that simply handing
consumers more information will help them make better choices.
The need to empower consumers, however, is self-serving. Part of the issue with the
deprecation of the third-party cookie is that it will lead to a loss of data, making it more difficult
for entities that are not Google or Apple to create targeted ads. While some research and data
workers are making sense of this with uncertainty and end times language, as described above,
others are making sense of the cookie and privacy debate as a data quality issue and a business
opportunity, thinking in more positive terms. One panelist at the Cynopsis conference advocated
for looking beyond cookies; he said, “cookies suck” because people often cleared them, which
made it difficult to get quality data from them. As one panelist at the ARF conference said, “Data
quality is a function of trust,” meaning that to increase data quality, the industry needs to
increase transparency and accountability. Consumer trust, then, becomes a way for the industry
to gather better data than third-party cookies could give them. Lawrence, the CEO of Connected
Media Research, argued that the deprecation of the third-party cookie will be good for consumer
agency and, crucially, business:
113
It's basically giving the consumer the power, which means that narrative or your
explanation to the consumer is becoming more important. You cannot blindly just assume
every consumer will share the data. Some consumers will share the data. Some
consumers may decide to opt out. So how to communicate with the consumer is
becoming more important, which I think is a good thing, because if a consumer decided
to opt into your program, typically you see a higher data quality. You see it easier to use
that data to help better targeting or better measurement for the ad. I think that's a good
thing.
Lawrence and others in the industry have asked themselves how they can fix the data quality
issue. Their answer is to frame the data quality issue around being in consumers’ interest; in their
model, with more education and transparency, consumers will sign away their data, creating
better quality data that allows the industry to better-target consumers. In this conceptualization,
consumers signing away their data will not be a form of privacy cynicism or digital resignation
(Draper & Turow, 2019; Hargittai & Marwick, 2016; Hoffmann et al., 2016) but rather optimism
that they can get something in return. Lawrence and others thus were able to imagine a more
positive future scenario for the advertising industry.
Others described this shift towards increased privacy and the deprecation of the third-
party cookie as more overtly self-serving for the industry. The same Cynopsis panelist who
argued that “cookies suck” expressed that he did not think most consumers care about privacy,
rather that governments and Apple made it into an important concern. Brenden, the head of the
team at Bullseye, expressed his view as follows:
The concept of identity hasn't gone away, if you're a walled garden. If you are logged into
Google, you are logged into Google. And it doesn't matter what third-party cookies there
are and it doesn't matter any of that stuff. If you are using an Apple iPhone, they know
who you are, and they have all this information available. If you are logged into
Facebook, they have all of this. So, yes, there are legitimate consumer concerns, and I
would say rather that those have been effectively co-opted by a lot of big tech and used in
ways that really are furthering their business.
In this way, the tension surrounding privacy is a classic example of a digital divide: the entities
with most of the power—in this case, big tech—will maintain the power, even while purporting
114
to make things better for consumers. At the same time, individuals in other organizations within
the industry, like Lawrence at Connected, tried to rationalize the upcoming loss of third-party
cookie data as good for consumers, leading to better data quality. In an era of data capitalism
(Sadowski, 2019; West, 2019), collecting and deploying data is imperative to keep a competitive
advantage. Companies like Apple and Google retrench into walled gardens, and other companies
have to imagine future scenarios where they can still compete.
Sensemaking the Future of Research and Data Work
In addition to the two above digital divides, there is another brewing in the industry. This
digital divide is not between the industry and consumers or between organizations within the
industry but between individuals within the industry itself. In particular, within media
organizations, data scientists and other individuals who are data literate have taken on an
increasingly prominent role. This reflects a broader trend that those with the ability to understand
numbers have more advantages in the job market (boyd & Crawford, 2012), which has already
been observed in advertising agencies, for example (Auletta, 2018; Deighton, 2017). While none
of my case studies is an advertising agency, participants, especially those at media publishers,
spoke about anticipating a shift in power towards data analysts and data scientists.
The feelings they expressed show that they are engaging in future-oriented sensemaking.
Much like the instances described above, the research and data workers here are taking a moment
to reflect on a changing industry. Whereas the quotes I cited above are specifically referencing
the deprecation of the third-party cookie and related debates about privacy, the quotes I cite
below are talking about the idea of data science, analytics, and big data becoming more
important to the industry. As I described in Chapters 2 and 3, major shifts in the media industries
have prompted a number of changes within media organizations; many media organizations are
115
making changes to better account for the increased centrality of streaming and digital media, and
people are working on organizing to change their culture.
Chapter 1 of this dissertation opened by describing the role of research and data work,
noting that it requires a balance of skills including at least some knowledge of data alongside
strong communication and storytelling skills. As data science becomes more important, some see
research and data work as becoming more about the second part of that divide: data translation
and storytelling, which are the hallmarks of research and insights work. Lily from Ballast said,
“When we already know that automation is coming, it's like, let's do that and have the thinking
going on so people can understand how to analyze the data, but not necessarily pull it.”
Therefore, some see the shift towards data science as only amplifying the necessity of the
research and insights side of the equation. Bobby from Ballast echoed that view with this to say
about data scientists:
Data science have been working on projects with, if they achieve their ultimate goal, they
would have been significantly more important than what they are. But I think what's
happening is it's so hard to achieve that goal, not because the data scientists are bad, it's
because there's so much limitations on marketplace to get to the end goal. So when we set
out to like, hey, you know what? Let's create this identity graph where we can assign
everybody that watches our networks, we have an ID on them. So we can quantify, know,
everyone that's on this thing, that's our ultimate goal. But because of data limitations,
privacy limitations, they can get to 30% of what they wanted to get. Right. So when that
happens, when the promise is not delivered… What data scientists have, they have a lot
of potential. They can promise this, but they haven't achieved it yet because of the
limitations of the marketplace. But if they could have achieved that, right, like the goal of
identifying everyone and some 30% getting to even 60, 70%, then like, wow, like, then
that would shift significantly from research or even marketers to data scientists.
As I explained in Chapter 1, Bobby argued that research and insights workers are like consultants
or teachers of data. Bobby continued:
But I think that [consulting and teaching] importance will grow when data scientists can
create a product that is very useful for sales. That hasn't happened yet. Yeah, so like a lot
of times, like we, this is what happens is like we'll go into a project with data sales,
there'll be a promise. And there's a lot of pessimism in terms of what can be achieved
because of the limitations, because of the history.… If they [data science] get to a place
116
where it's 80, 70% of what we want to do, I think our role as translators and teachers
would increase unless data scientists become a fusion of data scientists and marketers.
Due to the immense challenges of collecting data in the media industries, which were chronicled
in Chapter 2, analyzing and making use of those data are also challenging. Bobby comes from a
background as a research and insights worker, rather than one trained in data science or
analytics. As a result, we may read his comments as justifying his continued existence in the
industry; despite the excitement about big data, he sees a continued need for research and
insights.
Another participant, however, expressed concerns about data science eclipsing research
and insights. Peter, who works with the digital assets of USA Media but is not a data scientist,
expressed his anxiety about the changing workforce:
You now have an issue where you have middle management style workers such as myself
who have spent time with these tools learning them, and you have another class coming
up behind us who are all SQL Python R. And just look at a job description right now. If
someone took the title off of it, I would assume it was for a comsci major. The job
descriptions are ridiculous, and it makes me fearful that I will be unemployable in five
years. That is a real challenge. But, you know, you've got to change with the times, right?
So trying to learn SQL, trying to learn whatever I can. However, I don't. Let's be frank. I
don't even particularly like it. So does that then mean I have to change industries because
I don't want to be in R all day? I don't care about Python, really, if you get right down to
it. So there has been at conferences over the last two or three years. It's been. Do I need to
become a data scientist to have a job? If not, do I need to know at least R SQL Python?
Both Bobby and Peter have received the message that data science has become more important.
While Bobby saw that as an opportunity for research and insights work to expand its consulting
and teaching work, Peter saw it as the end of research and insights work. Peter reflected on his
job security using negative emotions, saying that he was fearful about becoming unemployable
and that he did not care for learning SQL, R, and Python. Peter’s use of language mirrors how
some participants described the deprecation of third-party cookies with uncertainty and end times
117
language. Peter’s fears reflect a growing digital divide among those who can do data science and
analytics and those who cannot.
Conclusion
This chapter has explained how changes in the media industries that will lead to new
digital divides, specifically the deprecation of third-party cookies and the growing centrality of
data science, have become moments for future-oriented sensemaking. The cookie challenge is an
example of an announcement that shocked research and data workers out of their routines and
triggered sensemaking about what this change could mean for their work. Similarly,
conversations about the importance of data science have prompted people to understand how
their work may change.
In both cases, some individuals imagined the future negatively, whereas others imagined
the future more positively. Some like Adria imagined a future in which her work would become
more challenging without third-party cookies, while Lawrence from Connected recognized the
difficulty of collecting data without third-party cookies and imagined a future in which
consumers would give away their data. Peter recognized that data science would become more
important but worried about his job security, while Bobby saw it as an opportunity for research
and insights to expand its role. As a result, we see that while individuals engaged in future-
oriented sensemaking often imagine darker futures (Bruskin & Mikkelsen, 2020), some
recognize the potential for a darker future and flip it into a more positive future.
Nevertheless, these more positive futures may still perpetuate digital divides. The
positive future Lawrence imagined is one where consumers will more willingly agree to hand
over their data. As others have described, not all consumers have the appropriate digital skills to
understand how their data are used (Büchi et al., 2017; Park, 2013); the industry’s insistence on
118
education may or may not solve this issue. Consumers may then find themselves in an
increasingly precarious position. The death of third-party cookies will mean that media
companies will need to rely more and more on consumers being willing to share their data. But
as Draper and Turow (2019) have argued, many users already feel a sense of digital
resignation—a sense of helplessness to control their data that in fact corporations actively
encourage by making privacy policies and other transparency initiatives deliberately opaque.
This sense of digital resignation may only increase in a cookieless future. As a result, an
imagined future that is positive for the industry may not be positive for consumers.

119
Conclusion
This dissertation has looked under the surface of media industries metrics by examining
media research and data work, an important yet understudied aspect of media work. While
previous studies have examined the ratings and audience measurement industries to understand
how data about media audiences are collected, few if any have examined what happens next.
Because of the immense changes that have occurred across media industries in recent years,
particularly the rise of streaming, media industries have a dilemma: most within the industries
realize more than ever that the data they have are imperfect, yet they have to use those data.
How, then, do individuals actually decide which data to use and how to use those data?
This dissertation has sought to investigate this broad question through the two key lenses
of sensemaking and culture and by following four organizations as case studies. First, the study
established what media research and data work is and how it fits within the broader television,
digital media, and advertising industries. Second, it described the state of rapid change these
industries are undergoing and how the COVID-19 pandemic accelerated this change. Third, it
discussed how these organizations have instituted new ways of organizing during COVID-19 to
build and reinforce cultures that promote creative and innovative work. Fourth, it overviewed
various types of problems that research and data workers face. Finally, it took a look around the
corner to see how these workers imagine the future of the industry and of their profession. The
findings of this dissertation have implications for understanding research and data work in the
post-Nielsen, post-cookie, and post-pandemic futures.
One of this study’s key contributions is defining research and data work and explaining
its role in the industry. This is a contribution to media industry studies, which have for many
years engaged with studies of ratings and audience measurement (e.g., Buzzard, 2012; Webster,
120
Phalen, & Lichty, 2014). However, research literature has not looked at this segment of the
industry through the lens of media work, which has generally been more interested in creative
fields like production, journalism, and video game production (Deuze, 2007). In this study, I
show that research and data work is creative, even if it does not involve creating media content,
in much the same way Roussel’s (2017) study of Hollywood talent agents argued that their work
is creative. Research and data work involves finding patterns in data that allows the creation of
stories and packaging data in ways that make them interesting to clients.
This dissertation’s exploration of research and data work also contributes to the field of
organizational communication through the key conceptual frameworks of organizational culture
and sensemaking. Chapter 1 unpacked how research and data workers describe themselves and
their work, which emphasizes needing to be detail-oriented and having a balance of skills. As
Chapter 3 showed, a key facet of their team cultures was that they were collaborative, which is a
critical element of culture that stimulates creativity and innovation. Even though some
participants thought that their organizations’ cultures were siloed, many participants stressed that
their team cultures were similar to their organizational cultures and indicated that the two
intertwine and impact each other. Additionally, research and data workers undertake
sensemaking when handling data issues and industrial change. Research and data workers are at
the forefront of change because they collect or work with data that are concerned with changing
media consumption patterns. As Chapter 2 discussed, technology continues to adapt and
audiences continue to evolve in their media use, so research and data workers must remain
nimble. It is imperative that they be in environments that are adaptable and open to collaboration
for them to do their work: to try to understand and sift through data.
121
Through their decision-making processes about data, research and data workers effect
change in the industries. They notice holes in multiple arrays and assemblages of data, and then
ascertain how to fill those holes, in the process, making judgments about what data to use and
how to use them, as Chapters 4 and 5 explained. These judgments topple certain datasets;
witness, for example, what has happened with the reputation of Nielsen data, which I chronicled
in the Introduction and in Chapters 1 and 2. Research and data workers have had to devise
workarounds and supplements to Nielsen data, often using creative thinking. However, their
actions may not always seem radical. Often, they partake in a process of simplexity (Colville et
al., 2012) because their work has to be responsive to clients, both internal and external. It is this
responsiveness that makes a lot of research and data work quick and creative. There is not as
much time or industrial impetus to be truly innovative. While this dissertation set out to
investigate innovation in audience measurement, the story underneath that, which this
dissertation told instead, is about how people use data and what that data use means for a
changing industry.
I conducted my research at the dawn of the post-pandemic, post-Nielsen, and post-cookie
futures, all of which are key changes. Each of these challenges the status quo and forces research
and data workers to adapt. More specifically, the post-pandemic future will force organizations
to come up with ways that continue to allow collaborative work, and the post-Nielsen and post-
cookie futures will force research and data workers to devise even more creative and innovative
data solutions, even as digital divides are perpetuated.
During the pandemic, research and data workers reported feeling overworked, had more
meetings, and had more Slack and email, but had less interaction with their team. As a result,
many organizations took the opportunity of the pandemic to reinforce collaboration and
122
camaraderie. Such strategies included holding virtual team building events, open virtual “office
hours” for people who need support, and large team meetings to facilitate group feedback on
projects. Other managers proactively reached out to team members proactively to check in on
their wellbeing and their projects.
At the same time, the waning dominance of Nielsen and the upcoming death of third-
party cookies have shaken the foundations of research and data work. As Chapter 4 discussed,
research and data workers often devise creative yet actionable solutions to solve data problems.
The future, however, will force these workers to think outside the box even more to come up
with ways that are not embedded in the Nielsen or cookie past. Traditional Nielsen metrics
measure who is watching what, yet the media environment has become much more complex; it is
not only necessary to know who is watching what, but where they are watching, how much they
are paying attention, and what actions they take during or after watching. Similarly, the death of
third-party cookies represents a fundamental shift in online advertising; no longer will third-party
tracking data be used to serve consumers targeted ads. As a result, there is increasing demand for
new data streams and individuals who can analyze data. At the same time, these changes
challenge the relationship between the industry and consumers, as a (potentially inauthentic)
concern for user privacy has co-opted industry practice. This project has sought to illuminate
these concerns.
This study has a few limitations, however. First, the number and mix of organizations
that agreed to participate in this study could have been stronger. The number of organizations,
while common and adequate, is smaller than anticipated, likely due to the pandemic, and the
organizations differ on features that would have been interesting to compare. Between the media
publishers, there is one legacy conglomerate and one legacy broadcaster; both produce linear and
123
digital offerings but are markedly different in their structures and goals. Between the data
vendors, there is one advertising technology organization and one television measurement
organization, and while both are focused on collecting and analyzing data, they are still operating
in different parts of the industry. As a result, it was difficult for me to describe a broad story arc
of the industry, but perhaps that may be premature given the rapidly changing nature of metrics.
This study is also missing any substantial engagement with advertising agencies, aside from one
conference attendee interview. In addition, because of the nature of the teams I spoke to, the two
publishers spoke more about addressing the needs of internal clients, and the two data
organizations spoke more about addressing the needs of external clients, because of the expertise
of the particular individuals willing to be interviewed in a rather secretive industry. Future
research should attempt a more balanced mix of organizations and a wider variety of professional
teams within them.
Furthermore, because of the COVID-19 pandemic and various time and space limitations,
I was unable to do participant observation, follow up interviews, or any in-person research,
which all limit the depth and breadth of the data. Also, because all of my interviews were
conducted during the same relative time period, I am unable to make any “cause and effect” or at
least before and after judgments, which might have been particularly fruitful for Chapter 3’s
investigations on new ways of organizing. Future research could use a longitudinal design to
establish these effects. Future research could also interview both internal and external clients of
research and data workers to understand whether the outcomes of their projects, like those
discussed in Chapter 4, are actually seen as creative and/or innovative. Both of these would
contribute to theory surrounding the relationships among organizational culture, creativity, and
innovation.
124
In addition to those methodological suggestions, future research could also examine
research and data work through different theoretical frames. In this study, I focused on the lenses
of sensemaking and organizational culture. One key theoretical frame that could inform future
work in this space is Giddens’s (1984) structuration theory, which argues that agents and
structures co-constitute their environments. This would inform a similar study from two angles.
First would be a deeper understanding of the structuring of media industries. Structuration would
show how research and data workers operate within certain organizational and industrial
structures of opportunities and constraints to co-constitute their environments, but this was
hampered by the lack of participant observation. For example, in Chapter 2, I focused on
lingering legacy thinking and how it may thwart innovation. With the right participating
organizations, future research could study how research and data work is still beholden to
outdated ways of thinking (the rules and resources drawn upon in praxis and their habitual
thought processes), and how these structures recursively limit the potential of this work.
The second way structuration could impact a similar study would be through
understanding “audiencemaking” (Ettema & Whitney, 1994). Webster (2011, 2014) has argued
that structuration theory helps explain audience attention through the industry and the audience
co-constituting each other. This duality is seen in media through measures, like ratings and other
audience data. The media industries are so interested in what audiences want, which affords
audiences some kind of power, but the industry can only know what the audience wants through
measuring it (Webster, 2011). This creates an abstracted audience known as the “institutionally
effective audience” (Ettema & Whitney, 1994). Traditional measurement, however, necessitated
that panelists participate in the ratings collection, a further instance of this duality and the
reflexive nature of the data. Because this audience is industrially-created, Turow and Draper
125
(2014) argue that it is vital to understand how the industry understands its audience because such
an analysis can unearth biases, much like other scholarly research which stresses the need to
understand the humans behind algorithms (Napoli, 2014a, 2014b; Noble, 2018).
Chapter 5 of this dissertation, which centered on digital divides, illustrated the tension
between the industry and the audience with regards to privacy: many within the industry do not
think highly of audiences, choosing to see them as complicit in their own exploitation. Chapter 2,
however, showed that others within the industry think audiences can effect change in the
industry by changing their behavior and forcing the industry to follow suit. This tension in
research and data workers’ view of the audience—on one side, the audience is exploited, and on
the other, it is powerful—remains underdeveloped in this project and deserves more attention
once organizations are again willing to accept researchers in their midst.
This study has examined research and data work from a handful of different angles, but
as described just now, other viable avenues remain. As I have shown, understanding research and
data workers is one important vantage point into understanding how the media industries resolve
their data dilemmas, as they have a boots on the ground view of the ways that market
information regimes within the industry are developing. As the industries continue to evolve, this
work will continue to serve a role in those industries. But research and data workers cannot
afford to rest on their laurels and must continue to adapt to change. As machine learning,
automation, and artificial intelligence become more prominent, more and more new metrics and
analytical capabilities will flood the market, further pushing legacy practices to the margins. As a
result, research and data work as described in this dissertation may be threatened. Many of the
findings in this dissertation unearthed that people within the industry feel a sense of dread and
panic about the future. For workers at data vendors, it is a feeling of needing to jockey for data
126
dominance to be a valued source of information to the industry. For workers at publishers, it is a
feeling of needing to evaluate and use the plethora of available vendors to best steward their
properties. The data dilemma will no doubt increase in the coming years, and research and data
workers must grow with it.

127
References
Alvesson, M., & Sveningsson, S. (2008). Changing organizational culture: Cultural change
work in progress. Milton Park: Routledge.
Amabile, T. M. (1998). How to kill creativity. Harvard Business Review, 76(5), 77–87.
Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing the work
environment for creativity. Academy of Management Journal, 39(5), 1154–1184.
https://doi.org/10.2307/256995
Anand, A. N., & Peterson, R. A. (2000). When market information constitutes fields:
Sensemaking of markets in the commercial music industry. Organization Science, 11(3),
270–284.
Andrejevic, M. (2014). The big data divide. International Journal of Communication, 8, 1673–
1689.
Anderson, N., Potočnik, K., & Zhou, J. (2014). Innovation and creativity in organizations: A
state-of-the-science review, prospective commentary, and guiding framework. Journal of
Management, 40(5), 1297–1333. https://doi.org/10.1177/0149206314527128
Ang, I. (1991). Desperately seeking the audience. New York: Routledge.
Angle, H. L. (1989). Psychology and organizational innovation. In A. H. Van De Ven, H. L.
Angle, & M. S. Poole (Eds.), Research on the Management of Innovation: The Minnesota
Studies (pp. 135–170). New York: Harper & Row.
Arsenault, A. H. (2017). The datafication of media: Big data and the media industries.
International Journal of Media & Cultural Politics, 13(1&2), 7–24.
https://doi.org/10.1386/macp.13.1-2.7

128
Auletta, K. (2018). How the Math Men overthrew the Mad Men. Retrieved April 14, 2022, from
https://www.newyorker.com/news/annals-of-communications/how-the-math-men-
overthrew-the-mad-men
Bloor, G., & Dawson, P. (1994). Understanding professional culture in organizational context.
Organization Studies, 15(2), 275–295. https://doi.org/10.2307/3319775
Bourdon, J., & Méadel, C. (2014). Deconstructing the ratings machine: An introduction. In J.
Bourdon & C. Méadel (Eds.), Television Audiences Across the World: Deconstructing the
Ratings Machine (pp. 1–30). New York: Palgrave Macmillan.
boyd, d., & Crawford, K. (2012). Critical questions for big data: Provocations for a cultural,
technological, and scholarly phenomenon. Information, Communication & Society, 15(5),
662–679. https://doi.org/10.1080/1369118X.2012.678878
Brooks, M. (2021). Meet the four horsemen of mobile’s answer to the cookiepocalypse.
Retrieved April 20, 2022, from https://www.adexchanger.com/data-driven-thinking/meet-
the-four-horsemen-of-mobiles-answer-to-the-cookiepocalypse/
Brown, A. D., Colville, I., & Pye, A. (2015). Making sense of sensemaking in organization
studies. Organization Studies, 36(2), 265–277.
https://doi.org/10.1177/0170840614559259
Bruskin, S., & Mikkelsen, E. N. (2020). Anticipating the end: Exploring future-oriented
sensemaking of change through metaphors. Journal of Organizational Change
Management, 33(7), 1401–1415. https://doi.org/10.1108/JOCM-11-2019-0342
Büchi, M., Just, N., & Latzer, M. (2017). Caring is not enough: The importance of Internet skills
for online privacy protection. Information Communication and Society, 20(8), 1261–
1278. https://doi.org/10.1080/1369118X.2016.1229001
129
Buzzard, K. (2002). The Peoplemeter Wars: A case study of technological innovation and
diffusion in the ratings industry. The Journal of Media Economics, 15(4), 273–291.
https://doi.org/10.1207/S15327736ME1504
Buzzard, K. (2012). Tracking the audience: The ratings industry from analog to digital. New
York: Routledge.
Caldwell, J. T. (2008). Production culture: Industrial reflexivity and critical practice in film and
television. Durham: Duke University Press.
Carlson, M. (2006). Tapping into TiVo: Digital video recorders and the transition from schedules
to surveillance in television. New Media and Society, 8(1), 97–115.
https://doi.org/10.1177/1461444806059877
Carlson, M. (2015). The robotic reporter: Automated journalism and the redefinition of labor,
compositional forms, and journalistic authority. Digital Journalism, 3(3), 416–431.
https://doi.org/10.1080/21670811.2014.976412
Chan-olmsted, S. M., & Ha, L. S. (2003). Internet business models for broadcasters: How
television stations perceive and integrate the internet. Journal of Broadcasting &
Electronic Media, 47(4), 597–616. https://doi.org/10.1207/s15506878jobem4704
Child, J. (2015). Managing organizational change. In Organization: Contemporary Principles
and Practice (2nd ed., pp. 333–369). Chichester: John Wiley & Sons, Inc.
Christensen, C. M., & Overdorf, M. (2000). Meeting the challenge of disruptive change.
Harvard Business Review, 78(2), 66–76.
Cohen, N. S. (2015). From pink slips to pink slime: Transforming media labor in a digital age.
The Communication Review, 18(2), 98–122.
https://doi.org/10.1080/10714421.2015.1031996
130
Colville, I., Brown, A. D., & Pye, A. (2012). Simplexity: Sensemaking, organizing and
storytelling for our time. Human Relations, 65(1), 5–15.
https://doi.org/10.1177/0018726711425617
Crossan, M. M., & Apaydin, M. (2010). A multi-dimensional framework of organizational
innovation: A systematic review of the literature. Journal of Management Studies, 47(6),
1154–1191. https://doi.org/10.1111/j.1467-6486.2009.00880.x
Curtin, M., Holt, J., & Sanson, K. (2014). Introduction: Making of a revolution. In M. Curtin, J.
Holt, & K. Sanson (Eds.), Distribution Revolution (pp. 1–17). Oakland, CA: University
of California Press.
D’Addario, D. (2018). The streaming era owes thanks to ‘House of Cards’ (Column). Retrieved
March 3, 2022, from https://variety.com/2018/tv/columns/house-of-cards-season-6-
streaming-era-1203017193/
Damanpour, F., Walker, R. M., & Avellaneda, C. N. (2009). Combinative effects of innovation
types and organizational performance: A longitudinal study of service organizations.
Journal of Management Studies, 46(4), 650–675. https://doi.org/10.1111/j.1467-
6486.2008.00814.x
Danneels, E. (2004). Disruptive technology reconsidered: A critique and research agenda. The
Journal of Product Innovation Management, 21(4), 246–258.
Dawson, P. (2003). Understanding organizational change: The contemporary experience of
people at work. London: SAGE Publications, Inc.
Deighton, J. (2017). Rethinking the profession formerly known as advertising: How data science
is disrupting the work of agencies. Journal of Advertising Research, 57(4), 357–361.
https://doi.org/10.2501/JAR-2017-045
131
Deuze, M. (2007). Media work. Cambridge: Polity Press.
Deuze, M., & Steward, B. (2011). Managing media work. In M. Deuze (Ed.), Managing Media
Work (pp. 1–10). Thousand Oaks, CA: SAGE Publications, Inc.
Draper, N. A., & Turow, J. (2019). The corporate cultivation of digital resignation. New Media
and Society, 21(8), 1824–1839. https://doi.org/10.1177/1461444819833
Drazin, R., Glynn, M. A., & Kazanjian, R. K. (1999). Multilevel theorizing about creativity in
organizations: A sensemaking perspective. The Academy of Management Review, 24(2),
286–307.
Eisenberg, E. M., & Riley, P. (2001). Organizational culture. In F. M. Jablin & L. L. Putnam
(Eds.), The New Handbook of Organizational Communication: Advances in Theory,
Research, and Methods (pp. 291–322). Thousand Oaks, CA: SAGE Publications, Inc.
Ellis, J. (2000). Scheduling: The last creative act in television? Media, Culture & Society, 22(1),
25–38.
Epstein, E. J. (1973). News from nowhere: Television and the news. New York: Vintage Books.
Ettema, J. S., & Whitney, D. C. (1994). The money arrow: An introduction to audiencemaking.
In J. S. Ettema & D. C. Whitney (Eds.), Audiencemaking: How the Media Create the
Audience (pp. 1–18). Thousand Oaks, CA: SAGE Publications, Inc.
Evans, S. K. (2018). Making sense of innovation: Process, product, and storytelling innovation in
public service broadcasting organizations. Journalism Studies, 19(1), 4–24.
https://doi.org/10.1080/1461670X.2016.1154446
Gans, H. J. (1979). Deciding what's news: A study of CBS Evening News, NBC Nightly News,
Newsweek, and Time. New York: Pantheon Books.

132
Gephart, R. P., Topal, C., & Zhang, Z. (2011). Future-oriented sensemaking: Temporalities and
institutional legitimation. In T. Hernes & S. Maitlis (Eds.), Process, Sensemaking, and
Organizing (pp. 275–312). Oxford: Oxford University Press.
https://doi.org/10.1093/acprof:oso/9780199594566.003.0013
Giddens, A. (1984). The constitution of society: Outline of the theory of structuration. Berkeley
and Los Angeles, CA: University of California Press.
Gitlin, T. (2000). Inside prime time: With a new introduction. Berkeley and Los Angeles, CA:
University of California Press.
Greene, A. (2013). How ‘Lilyhammer’ changed the TV world. Retrieved March 3, 2022, from
https://www.rollingstone.com/music/music-news/how-lilyhammer-changed-the-tv-world-
203682/
Hargittai, E., & Marwick, A. (2016). “What can I really do?” Explaining the privacy paradox
with online apathy. International Journal of Communication, 10, 3737–3757.
Hartley, J. (1987). Invisible fictions: Television audiences, paedocracy, pleasure. Textual
Practice, 1(2), 121–138. https://doi.org/10.1080/09502368708582010
Hartley, J. (2005). Creative industries. In J. Hartley (Ed.), Creative Industries (pp. 1–40).
London: Wiley-Blackwell.
Havens, T. (2014). Media programming in an era of big data. Media Industries, 1(2), 5–9.
Havens, T., Lotz, A. D., & Tinic, S. (2009). Critical media industry studies: A research approach.
Communication, Culture & Critique, 2(2), 234–253. https://doi.org/10.1111/j.1753-
9137.2009.01037.x

133
Hercher, J. (2021). Google grants a third-party cookie reprieve, delays deprecation by two years.
Retrieved March 3, 2022, from https://www.adexchanger.com/online-advertising/google-
grants-a-third-party-cookie-reprieve-delays-deprecation-by-two-years/
Hesmondhalgh, D. (2019). The cultural industries approach: Distinctive features of the culture-
producing businesses. In The Cultural Industries (pp. 27–44). London: SAGE
Publications, Inc.
Hesmondhalgh, D., & Lobato, R. (2019). Television device ecologies, prominence and
datafication: The neglected importance of the set-top box. Media, Culture & Society,
41(7), 958–974. https://doi.org/10.1177/0163443719857615
Hill, R. C., & Levenhagen, M. (1995). Metaphors and mental models: Sensemaking and
sensegiving in innovative and entrepreneurial activities. Journal of Management, 21(6),
1057–1074. https://doi.org/10.1177/014920639502100603
Hoffmann, C. P., Lutz, C., & Ranzini, G. (2016). Privacy cynicism: A new approach to the
privacy paradox. Cyberpsychology, 10(4). https://doi.org/10.5817/CP2016-4-7
Hofstede, G. (1998). Identifying organizational subcultures: An empirical approach. Journal of
Management Studies, 35(1), 1–12. https://doi.org/10.1111/1467-6486.00081
Hofstede, G., Neuijen, B., Ohayv, D. D., & Sanders, G. (1990). Measuring organizational
cultures: A qualitative and quantitative study across twenty cases. Administrative Science
Quarterly, 35(2), 286–316.
Holt, J., & Perren, A. (2009). Introduction: Does the world really need one more field of study?
In J. Holt & A. Perren (Eds.), Media Industries: History, Theory, and Method (pp. 1–16).
Malden, MA: Wiley-Blackwell.

134
Iliadis, A., & Russo, F. (2016). Critical data studies: An introduction. Big Data & Society, 3(1),
1–7. https://doi.org/10.1177/2053951716674238
Järventie-Thesleff, R., Moisander, J., & Villi, M. (2014). The strategic challenge of continuous
change in multi-platform media organizations — A strategy- as-practice perspective.
International Journal on Media Management, 16(3–4), 123–138.
https://doi.org/10.1080/14241277.2014.919920
Jassawalla, A. R., & Sashittal, H. C. (2002). Cultures that support product-innovation processes.
Academy of Management Executive, 16(3), 42–54.
Kanter, R. M. (1983). The change masters. New York: Simon and Schuster.
Kelly, J. P. (2019). Television by the numbers: The challenges of audience measurement in the
age of Big Data. Convergence: The International Journal of Research into New Media
Technologies, 25(1), 113–132. https://doi.org/10.1177/1354856517700854
Kininmonth, S. (2021). “Anything that can be traded, will be traded”: The contests to automate
and financialize advertising futures markets. Television and New Media, 1–17.
https://doi.org/10.1177/15274764211008504
Kirkpatrick, K. (2015). Putting the data science into journalism. Communications of the ACM,
58(5), 15–17. https://doi.org/10.1145/2742484
Kosterich, A. (2016). Reconfiguring the “hits”: The new portrait of television program success in
an era of big data. International Journal on Media Management, 18(1), 43–58.
https://doi.org/10.1080/14241277.2016.1166431
Krumsvik, A. H., Milan, S., Ní Bhroin, N., & Storsul, T. (2019). Making (sense of) media
innovations. In M. Deuze & M. Prenger (Eds.), Making Media: Production, Practices,
and Professions (pp. 193–206). Amsterdam: Amsterdam University Press.
135
Küng, L. (2011). Managing strategy and maximizing innovation in media organizations. In M.
Deuze (Ed.), Managing Media Work (pp. 43–56). Thousand Oaks, CA: SAGE
Publications, Inc.
Landau, D., & Drori, I. (2008). Narratives as sensemaking accounts: The case of an R&D
laboratory. Journal of Organizational Change Management, 21(6), 701–720.
https://doi.org/10.1108/09534810810915736
Lewis, L. K. (2014). Organizational change and innovation. In L. L. Putnam & D. K. Mumby
(Eds.), The SAGE Handbook of Organizational Communication: Advances in Theory,
Research, and Methods (3rd ed., pp. 503–524). Thousand Oaks, CA: SAGE Publications,
Inc.
Litman, B. R. (1983). US TV networks’ response to new technology. Telecommunications
Policy, 7(2), 163–177.
Maitlis, S., & Christianson, M. (2014). Sensemaking in organizations: Taking stock and moving
forward. Academy of Management Annals, 8(1), 57–125.
https://doi.org/10.1080/19416520.2014.873177
Maitlis, S., & Sonenshein, S. (2010). Sensemaking in crisis and change: Inspiration and insights
from Weick (1988). Journal of Management Studies, 47(3), 551–580.
https://doi.org/10.1111/j.1467-6486.2010.00908.x
Malmelin, N., & Villi, M. (2017). Media work in change: Understanding the role of media
professionals in times of digital transformation and convergence. Sociology Compass,
11(7), 1–9. https://doi.org/10.1111/soc4.12494

136
Malmelin, N., & Virta, S. (2017). Managing for serendipity: Exploring the organizational
prerequisites for emergent Creativity. International Journal on Media Management,
19(3), 222–239. https://doi.org/10.1080/14241277.2017.1308947
Malthouse, E. C., Maslowska, E., & Franks, J. U. (2018). Understanding programmatic TV
advertising. International Journal of Advertising, 37(5), 769–784.
https://doi.org/10.1080/02650487.2018.1461733
Martin-Rios, C. (2016). Sensemaking of organizational innovation and change in public research
organizations. International Journal of Organizational Analysis, 24(3), 516–531.
https://doi.org/10.1108/IJOA-07-2014-0784
Marshak, R. J. (2002). Changing the language of change: How new contexts and concepts are
challenging the ways we think and talk about organizational change. Strategic Change,
11(5), 279–286.
McDermott, J. (2015). Welcome to the ‘walled garden’ era of ad tech. Retrieved April 15, 2022,
from https://digiday.com/media/walled-garden-era-ad-tech/
McLean, L. D. (2005). Organizational culture’s influence on creativity and innovation: A review
of the literature and implications for human resource development. Advances in
Developing Human Resources, 7(2), 226–246.
https://doi.org/10.1177/1523422305274528
Meziani, N., & Cabantous, L. (2020). Acting intuition into sense: How film crews make sense
with embodied ways of knowing. Journal of Management Studies, 57(7), 1384–1419.
https://doi.org/10.1111/joms.12619

137
Miles, M. B., Huberman, A. M., & Saldaña, J. (2020). Fundamentals of qualitative data analysis.
In Qualitative Data Analysis: A Methods Sourcebook (4th ed., pp. 61–99). Thousand
Oaks, CA: SAGE Publications, Inc.
Napoli, P. M. (2003). Audience economics: Media institutions and the audience marketplace.
New York: Columbia University Press.
Napoli, P. M. (2011). Audience evolution: New technologies and the transformation of media
audiences. New York: Columbia University Press.
Napoli, P. M. (2014a). Automated media: An institutional theory perspective on algorithmic
media production and consumption. Communication Theory, 24(3), 340–360.
https://doi.org/10.1111/comt.12039
Napoli, P. M. (2014b). On automation in media industries: Integrating algorithmic media
production into media industries scholarship. Media Industries, 1(1), 33–38.
Napoli, P. M., & Andrews, K. (2008). Managing innovation in audience measurement: The US
case studies of BookScan and the Local People Meter. In C. Dal Zotto & H. Van
Kranenburg (Eds.), Management and Innovation in the Media Industry (pp. 263–286).
Cheltenham, UK: Edward Elger.
Nielsen. (n.d.). Total Content Ratings. Retrieved May 6, 2018, from
http://www.nielsen.com/us/en/solutions/capabilities/total-content-ratings.html
Nielsen. (2020). Cross-media currency becomes a reality with Nielsen ONE. Retrieved January
4, 2021, from https://www.nielsen.com/us/en/press-releases/2020/cross-media-currency-
becomes-reality-with-nielsen-one/

138
Nigam, A., & Ocasio, W. (2010). Event attention, environmental sensemaking, and change in
institutional logics: An inductive analysis of the effects of public attention to Clinton’s
health care reform initiative. Organization Science, 21(4), 823–841.
https://doi.org/10.1287/orsc.1090.0490
Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. New
York: NYU Press.
Oakley, K., & O’Connor, J. (2015). The cultural industries: An introduction. In K. Oakley & J.
O’Connor (Eds.), The Routledge Companion to the Cultural Industries (pp. 1–32). New
York: Routledge.
Park, Y. J. (2013). Digital literacy and privacy behavior online. Communication Research, 40(2),
215–236. https://doi.org/10.1177/0093650211418338
Patriotta, G., & Gruber, D. A. (2015). Newsmaking and sensemaking: Navigating temporal
transitions between planned and unexpected events. Organization Science, 26(6), 1574–
1592. https://doi.org/10.1287/orsc.2015.1005
Petruska, K., & Vanderhoef, J. (2014). TV that watches you: Data collection and the connected
living room. Spectator, 34(2), 33–42.
Phalen, P. (2018). Writing Hollywood: The work and professional culture of television writers.
New York: Routledge.
Poggi, J. (2016). NBCU says Nielsen’s Total Content Ratings aren’t ready for primetime.
Retrieved April 29, 2019, from https://adage.com/article/media/nbcu-nielsen-s-total-
content-ratings-ready-primetime/307179

139
Poggi, J. (2017). Nielsen delays wide release of Total Content Ratings, doesn’t set a new target
date. Retrieved April 29, 2019, from https://adage.com/article/media/nielsen-s-total-
content-ratings-syndicated-data-hold/307735
Prescott, M. E. (2014). Big data and competitive advantage at Nielsen. Management Decision,
52(3), 573–601. https://doi.org/10.1108/MD-09-2013-0437
Reilly, S. (2017). The need to help journalists with data and information visualization. IEEE
Computer Graphics & Applications, 37(2), 8–10.
https://doi.org/10.1126/science.283.5405.1134
Roussel, V. (2017). Representing talent: Hollywood agents and the making of movies. Chicago:
The University of Chicago Press.
Sadowski, J. (2019). When data is capital: Datafication, accumulation, and extraction. Big Data
& Society, 6(1), 1–12. https://doi.org/10.1177/2053951718820549
Schiff, A. (2022). Watch out, Nielsen ONE, Comscore is busily building a unified cross-platform
measurement offering of its own. Retrieved January 12, 2022, from
https://www.adexchanger.com/platforms/watch-out-nielsen-one-comscore-is-busily-
building-a-unified-cross-platform-measurement-offering-of-its-own/
Schumpeter, J. A. (1950). Capitalism, socialism and democracy. London: Unwin University
Books.
Schwartz, D. (2018). Embedded in the crowd: Creative freelancers, crowdsourced work, and
occupational community. Work and Occupations, 45(3), 247–282.
https://doi.org/10.1177/0730888418762263

140
Shields, R. (2021). Key questions need to be answered after Google delays the cookiepocalypse.
Retrieved April 20, 2022, from https://www.adweek.com/programmatic/key-questions-
need-to-be-answered-after-google-delays-the-cookiepocalypse/
Skaggs, R. (2019). Socializing rejection and failure in artistic occupational communities. Work
and Occupations, 46(2), 149–175. https://doi.org/10.1177/0730888418796546
Smith, M. D., & Telang, R. (2016). Streaming, sharing, stealing: Big data and the future of
entertainment. Cambridge, Massachusetts: The MIT Press.
Somech, A., & Drach-Zahavy, A. (2013). Translating team creativity to innovation
implementation: The role of team composition and climate for innovation. Journal of
Management, 39(3), 684–708. https://doi.org/10.1177/0149206310394187
Steinberg, B. (2021a). Nielsen’s national ratings accreditation suspended by Media Rating
Council. Retrieved September 1, 2021, from https://variety.com/2021/tv/news/nielsen-
national-ratings-suspension-media-rating-council-1235053949/
Steinberg, B. (2021b). With Nielsen under scrutiny, NBCU wants to build new measurement
system. Retrieved January 12, 2022, from https://variety.com/2021/tv/news/nbcuniversal-
nielsen-measurement-tv-advertising-1235046711/
Steinberg, B. (2022). WarnerMedia Taps Comscore, iSpot.TV and VideoAmp in audience-
measurement bid. Retrieved January 12, 2022, from
https://variety.com/2022/tv/news/warnermedia-tv-ratings-measurment-nielsen-
1235147807/

141
Stone, M. L. (2014). Big data for media. Reuters Institute for the Study of Journalism. Retrieved
from https://ora.ox.ac.uk/objects/uuid:0b03bcfe-5c1d-41aa-97d6-
838b8ea3921a/download_file?file_format=pdf&safe_filename=Big%2BData%2Bfor%2
BMedia&type_of_work=Report
Turow, J., & Draper, N. (2014). Industry conceptions of audience in the digital space: A research
agenda. Cultural Studies, 28(4), 643–656. https://doi.org/10.1080/09502386.2014.888929
Uricchio, W. (2004). Television’s next generation: Technology/interface culture/flow. In L.
Spigel & J. Olsson (Eds.), Television after TV: Essays on a Medium in Transition (pp.
163–182). Durham: Duke University Press.
van Dijck, J. (2014). Datafication, dataism and dataveillance. Surveillance & Society, 12(2),
197–208.
Van Maanen, J., & Barley, S. R. (1984). Occupational communities: Culture and control in
organizations. Research in Organizational Behavior, 6, 287–365.
Webster, J. G. (2011). The duality of media: A structurational theory of public attention.
Communication Theory, 21(1), 43–66. https://doi.org/10.1111/j.1468-2885.2010.01375.x
Webster, J. G. (2014). The marketplace of attention: How audiences take shape in a digital age.
Cambridge, Massachusetts: MIT Press.
Webster, J. G., Phalen, P. F., & Lichty, L. W. (2014). Ratings analysis: Audience measurement
and analytics (4th ed.). New York: Routledge.
Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: SAGE Publications,
Inc.

142
Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (2005). Organizing and the process of
sensemaking. Organization Science, 16(4), 409–421.
https://doi.org/10.1287/orsc.1050.0133
Wenger, E. C., & Snyder, W. M. (2000). Communities of practice: The organizational frontier.
Harvard Business Review, 78(1), 139–145.
West, M. A., & Sacramento, C. (2012). Creativity and innovation: The role of team and
organizational climate. In M. D. Mumford (Ed.), Handbook of Organizational Creativity
(pp. 359–385). Elsevier Inc. https://doi.org/10.1016/B978-0-12-374714-3.00015-X
West, S. M. (2019). Data capitalism: Redefining the logics of surveillance and privacy. Business
and Society, 58(1), 20–41. https://doi.org/10.1177/0007650317718185
Weststar, J. (2015). Understanding video game developers as an occupational community.
Information Communication and Society, 18(10), 1238–1252.
https://doi.org/10.1080/1369118X.2015.1036094
Wyatt, J. (2014). Market research in the media industries: On the strategic relationship between
client and supplier. In D. Johnson, D. Kompare, & A. Santo (Eds.), Making Media Work:
Cultures of Management in the Entertainment Industries (pp. 254–274). New York: NYU
Press.
Zuboff, S. (2015). Big other: Surveillance capitalism and the prospects of an information
civilization. Journal of Information Technology, 30(1), 75–89.
https://doi.org/10.1057/jit.2015.5

143
Appendix: Interview Guide

Establishing Questions
● To begin, can you briefly describe your work background and some of the research work
you’ve been involved in?
● What is your current position? How would you describe the work you do?
● Tell me about a typical day at your job.

Disruption & Challenges
● What do you see as the biggest disruptor in the current media climate?
o What do you think is the biggest change it has brought about?
o How has your day-to-day work changed as a result of this disruption?
o What about your organization?
● What do you see as the biggest challenge facing media researchers today?
o How are you and your team working to address it?

Organizations and Teams
● Tell me about your team. How do you work on new projects?
● Have you ever pitched a project idea to your managers? How did it go?
● Tell me about a project you worked on that was really creative and/or innovative. How
did it come to be?
o Do you think the project was more creative or innovative? Why?
o How would you describe creativity and innovation?
● What words would you use to describe your organization’s culture?
o What about your team’s culture?

Big Data
● How do you deal with new streams of data?
● How is “big data” part of your job?
o When did this start?
o What (if any) challenges did this introduce?

Audiences
● How would you characterize changes in the audience landscape in the years you’ve been
working in research?
● How would you describe your organizations’ target audience?
● How does your work understand and engage your organization’s target audience?

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Staying ahead of the digital tsunami: strategy, innovation and change in public media organizations

PDF

Comic books incorporated: industrial strategy and the legitimation of lowbrow media

PDF

Data-driven culture in the consumer goods industry

PDF

A seat at the table: navigating identity, power, and safety in work meetings

PDF

Translating data protection into practice: exploring gaps in data privacy and cybersecurity within organizations

PDF

Media reinvented: the transformation of news in a networked society

PDF

These are their stories: two decades of Showrunner production, content, and context in Law & Order: SVU

PDF

Building a unicorn: management of innovation, collaboration, and change in a Silicon Valley start-up

PDF

Rebels with a cause: Youth, social movements, and media

PDF

Sunsetting: platform closure and the construction of digital cultural loss

PDF

Making business news in the digital age

PDF

Carnival in the birdcage: a study of the film and media industries' reforms in China after 1978 and its soft power

PDF

The new news: vision, structure, and the digital myth in online journalism

PDF

Cosmopolitan logics and limits: networked discourse, affect, and identity in responses to the Syrian refugee crisis

PDF

Virtually human? Negotiation of (non)humanness and agency in the sociotechnical assemblage of virtual influencers

PDF

New theoretical and research directions for foresight scenario work: narratives, sensemaking, and networks

PDF

The missing link: the role of organizational culture in technology change management communication

PDF

Upvoting the news: breaking news aggregation, crowd collaboration, and algorithm-driven attention on reddit.com

PDF

Competing logics and organizational failure: the structuring of California prison health care

PDF

Asian American media activism: past, present, and digital futures

Asset Metadata

Creator Jonckheere, Natalie (author)

Core Title The data dilemma: sensemaking and cultures of research in the media industries

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Degree Conferral Date 2022-08

Publication Date 07/22/2024

Defense Date 05/16/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag creativity and innovation,data,media industries,OAI-PMH Harvest,organizational change,organizational culture

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Riley, Patricia (committee chair), Williams, Dmitri (committee chair), Jenkins, Henry (committee member)

Creator Email nejonckheere@gmail.com,njonckhe@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111373969

Unique identifier UC111373969

Legacy Identifier etd-Jonckheere-10911

Document Type Dissertation

Format application/pdf (imt)

Rights Jonckheere, Natalie

Type texts

Source 20220722-usctheses-batch-961 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu