University of Southern California Dissertations and Theses

A conceptual framework of distribution intensity

(USC Thesis Other)

A conceptual framework of distribution intensity

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content A CONCEPTUAL FRAMEWORK OF
DISTRIBUTION INTENSITY
by
Walfried M. Lassar
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Business Administration)
August 1992
Copyright 1992 Walfried M. Lassar
UM I Number: DP22691
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI
Dissertation Publishing
UMI DP22691
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-4015
This dissertation, w ritten by
Walfried M . Lassar.......................................
under the direction of h .is D issertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillm ent of re
quirem ents for the degree of
D O C T O R OF PH ILOSOPH Y
Dean of G raduate Studies
Date . . . ..I.h...
DISSERTATION COMMITTEE
Chairperson
Dedication
To my father for having taught me how to overcome obstacles and never quit. And
to my mother for practicing such spirit every day.
ii
Acknowledgments
I am thankful to my wife Sharon for her never ending support in helping
me through this project and the Ph.D. program. She not only took care of and
provided for the family during my doctoral studies, but she also replaced the father
for the children during my extended absence. In addition to that, she had to put up
with me under stress. I hope in the end it will be worth all her efforts.
iii
Contents
Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii
Abstract viii
Preface ix
1 Introduction 1
2 Literature Review 8
2.1 Defining the Distribution Intensity Construct .................................. 9
2.2 Distribution Intensity and the Categories of Goods Framework .. 14
2.3 Distribution Intensity in the Industrial Goods Channel ...................23
2.4 Distribution Intensity and Firm Performance ....................................27
2.5 Distribution Intensity and Equilibrium Channel Structure ..............32
2.6 Integration and Evaluation of Existing Literature ............................ 35
3 The Framework 38
3.1 The Manufacturer Component of Distribution Intensity ............... 42
3.1.1 Product Positioning ................................................................ 42
3.1.2 Target Focus ........................................................................... 47
3.1.3 Need for Channel Control .................................................... 49
3.1.4 Product Positioning and Need for Control ........................ 53
3.1.5 Targeting Focus and Need for Control .............................. 54
3.2 The Intermediary Component of Distribution Intensity .................. 57
3.2.1 Financial Attractiveness ........................................................ 60
3.2.2 Barriers to Entry .......................................................................62
3.2.2.1 Level of Investment ....................................................63
3.2.2.2 Relationship Uncertainty ......................................... 66
3.2.2.3 Managerial Support .................................................. 68
3.2.2.4 Contractual Restrictiveness ..................................... 69
3.2.3 Moderation Effect of Financial Attractiveness ....................70
3.3 Summary ............................................................................................... 70
iv
4 Research Methodology 73
4.1 Research Setting .....................................................................................73
4.2 Unit of Analysis .....................................................................................75
4.3 Sampling Method ...................................................................................76
4.4 Data Collection Method ........................................................................80
4.5 Measurement Instrument ...................................................................... 82
4.5.1 Operationalization of Dependent Variable:
Distribution Intensity ................................................................ 83
4.5.2 Operationalization of Independent Variables:
Manufacturer Aspect .................................................................85
4.5.3 Operationalization of Independent Variables:
Intermediary Aspect ..................................................................89
5 Results 95
5.1 Descriptive Analysis of the Sample and Constructs ................... 95
5.1.1 Sample .................................................................................... 95
5.1.2 Dependent Variable: Distribution Intensity .................... 97
5.1.3 Independent Variables ......................................................... 99
5.2 Hypotheses Tests ............................................................................... 117
5.2.1 Original Framework ........................................................... 117
5.2.2 Expanded Framework ......................................................... 119
5.2.3 Hypotheses ........................................................................... 123
5.2.3.1 Manufacturer Aspect of Framework .................. 123
5.2.3.2 Intermediary Aspect of Framework .................. 128
5.3 Summary ............................................................................................ 131
6 Summary and Conclusions 133
6.1 Research Objectives ......................................................................... 133
6.2 Methodology ...................................................................................... 135
6.3 Empirical Findings ............................................................................ 136
6.4 Limitations ......................................................................................... 138
6.5 Future Research ................................................................................ 141
References 143
v
List of Tables
Table 4.1 Items Used to Measure Distribution Intensity .......................... 85
Table 4.2 Items Used to Measure Product Positioning ............................. 87
Table 4.3 Items Used to Measure Target Focus ......................................... 88
Table 4.4 Items Used to Measure Need for Channel Control ................. 89
Table 4.5 Items Used to Measure Financial Attractiveness ...................... 92
Table 4.6 Items Used to Measure Management Support .......................... 93
Table 4.7 Items Used to Measure Contractual Restrictiveness ................ 93
Table 4.8 Items Used to Measure Investment Level ................................. 94
Table 4.9 Items Used to Measure Relationship Uncertainty .................... 94
Table 5.1 Sample Characteristics:
Dependent and Independent Variables ........................................ 97
Table 5.2 Validation of Distribution Intensity Measure ........................... 98
Table 5.3 Convergent Validity of Distribution Intensity Measure .......... 99
Table 5.4 Validation of Product Positioning Measure .............................. 101
Table 5.5 Validation of Target Focus Measure ........................................... 102
Table 5.6 Validation of Need for Control Measure .................................... 103
Table 5.7 Convergent Validity of Independent Measures:
Manufacturer Aspect ....................................................................... 106
Table 5.8 Validation of Alternative Measure
of Need for Channel Control ......................................................... 107
Table 5.9 Validation of Financial Attractiveness Measure ........................ 109
Table 5.10 Convergent Validity of Financial Attractiveness Measure ..... 110
Table 5.11 Validation of Investment Level Measure .................................. 110
Table 5.12 Validation of Relationship Uncertainty Measure .......................I l l
Table 5.13 Divergent Validity of Independent Measures .............................114
Table 5.14 Divergent Validity of Independent Measures:
Multitrait Matrix .............................................................................. 115
Table 5.15 Bi-Construct Analysis of Divergent Validity ..............................115
Table 5.16 Bi-Construct Analysis of Divergent Validity ..............................116
Table 5.17 Original Framework: Regression Model Part I ......................... 120
Table 5.18 Original Framework: Regression Model Part II ........................121
Table 5.19 Expanded Framework: Regression Model Part I ...................... 125
Table 5.20 Expanded Framework: Regression Model Part II .....................126
Table 5.21 Summary of Empirical Findings ...................................................132
vi
List of Figures
Figure 3.1 Conceptual Framework of Distribution Intensity ....................... 41
Figure 5.1 Expanded Framework of Distribution Intensity ...........................122
Figure 6.1 Emperical Results ............................................................................. 137
vii
Abstract
Although distribution intensity is a key element for channel design
decisions, research interest has been limited. Several conceptual frameworks have
been advanced, yet few have focused on the construct of distribution intensity and
none has been empirically tested. This study aims to expand our understanding of
distribution intensity by looking at the construct within one product category. It
develops an integrated conceptual framework of the construct based on two major
components. The first framework component deals with manufacturer influence on
distribution intensity based on product positioning, the scope of targeted market
segments, and manufacturer need for marketing control within the channel. The
second framework component focuses on intermediaries, especially their motivation
to participate in a manufacturer’s distribution network. The framework establishes
financial attractiveness, and barriers to entry as major factors driving intermediary
participation in a distribution network. It is intended to explain variation of
distribution intensity levels across brands within one product line. Due to its
exploratory nature, the framework is primarily tested via multiple regression
estimation procedures. Estimation via simultaneous equations in a two step least
square procedure generally confirm the manufacturer aspect of the framework and
its influence on distribution intensity. Support for the intermediary component of
the framework is mixed.
Preface
The idea for this topic was generated while I was trying to establish a
dealer network for a small computer company based in California. We tried to
develop a dealer network in the United States as well as in Germany. To my
surprise, the effort to recruit and success in signing them was very different in
both markets even though we sold the same product. Further market analysis
clearly showed that our positioning relative to competition as well as the general
distribution parameters were completely different for both markets. An idea for
research was born. After I entered the Ph.D. program, three seminars helped me
refine the research idea. Dr. Frazier’s seminars in Marketing Strategy and
Channels of Distribution helped me develop the building blocks for the conceptual
framework while Dr. Sheth’s seminar in Marketing Theory convinced me that
academic research should be practically useful and made me keep in mind the
practical implications as the research developed. This study is just the start of what
I hope will evolve into some general framework of distribution structure.
I am indebted to my committee members for the insight provided in
conceptualizing this research. My chairman Dr. Gary Frazier provided much
insight in designing this study and helped me avoid the traps of empirical study. I
Dr. David Stewart gave me important guidance in analyzing the collected data and
Dr. Everett Rogers provided most helpful insights in bringing across what I had to
say. Due to their tremendous support, this study was completed in relative speedy
fashion and generated results which might make it an important piece of research
in the area of distribution structure. Last but not least, I would like to thank my
friend and cohort Avudajappan Sankaralingam for his always open ears in
discussing this study as it went along.
x
Chapter 1
Introduction
Distribution channels are typically comprised of a set of interdependent
organizations and are concerned with the flow of products or services from
producers to markets. They are designed not only to satisfy demand by supplying
products at the right place, quantity, quality and price but also to stimulate demand
through promotional activities of channel members (Stern and El-Ansary 1988,
Lusch 1979). Consequently, channels research deals with two major areas of
investigation: research in channel management and research in channel structure.
Research in channel structure deals with the question of how distribution is
generally organized to facilitate the flow of products. The aspect of distribution
structure that deals with the issue of how many retailers a firm uses to distribute its
products in a given trade area is the concept of distribution intensity (Stern and El-
Ansary 1988).
Distribution intensity is one of the core issues in marketing since the
question of how many retailers carry a product in a given trade area has a direct
impact on the availability of that product to the endconsumer and the facilitaton of
1
an exchange. Ideal distribution intensity makes a product widely enough available
to satisfy target consumers’ needs - but not exceed them. Too much market
exposure only increases marketing costs without benefits (McCarthy and Perreault
1984). Using too many retailers could also be detrimental to the product image and
its competitive position in the market place.
Take the example of Pioneer in the 1970s, which positioned itself as a
supplier of high end Hi-Fi products. Due to Pioneer’s apparent success in the
market place, it vastly expanded its distribution network, which led to short term
gains in sales and profits. Unfortunately, it also increased intra-brand competition
among dealers selling the Pioneer brand. This intra-brand competition was mainly
price-based and led to a deterioration of the channel relationship between Pioneer
and its intermediaries. Middlemen used the Pioneer name to draw traffic but
switched consumers to other more profitable brands if possible. In the process,
they used legal as well as illegal tactics, which was detrimental to Pioneer’s quality
image and its high end product perception in the market place. As a result, it is not
clear that an increase in distribution intensity for a given trade area increases long
term sales and profits.
The Pioneer example points out an additional area of significance for
distribution intensity. The focus in designing distribution channels is on developing
a strong operating network in support of the manufacturer’s goals of sales volume,
2
market share, or profits (Hardy and Magrath 1988). Stern and El-Ansary (1988)
identify the key issues in designing the channel as the degree of market coverage,
exposure, and support necessary to achieve corporate objectives. Consequently, the
impact of distribution intensity on the relationship between manufacturer and
middlemen must be considered. Given that the distribution channel is not only an
instrument to satisfy existing demand but also to stimulate demand, the crucial
question is what kind of participation in the marketing flows should be required
from each of the outlets and how many sales outlets should be established in a
given geographic area to secure such intermediary participation so that the needs of
existing, potential, and past customers may be adequately served.
A channel structure not conducive to the overall marketing effort can also
severely limit or undermine the effectiveness of all other marketing variables. In
the case of Pioneer, the dysfunctional conflict between the manufacturer and its
intermediaries severely undermined the effectiveness of all marketing programs
implemented by Pioneer. Other marketing variables like advertising, pricing, or
promotion can easily be changed when market conditions change. Changing the
number of retailers per trade area or replacing certain retailers with others or a
company-owned network is usually slow and costly (Kotler and Armstrong 1991).
Therefore, decisions on distribution intensity become key to the complete
marketing effort since they are difficult to change but at the same time should
3
establish a channel structure which supports and enhances other strategic marketing
variables (Rosenbloom 1991).
Given the importance of distribution intensity, it is surprising how little
research has been done on the issue. The research available is mostly conceptual in
nature and based on the logic that product characteristics are the primary force
behind the level of distribution intensity (Copeland 1923, Aspinwall 1958, Miracle
1965). It commonly associates three categories of goods with various levels of
distribution intensity. At one extreme, specialty goods are usually associated with
exclusive distribution, while at the other end of the spectrum convenience goods
are associated with intensive distribution. Shopping goods are in between and
associated with selective distribution (Copeland 1923, Bucklin 1963). The little
empirical research that deals with distribution intensity does not focus on the
driving factors of intensity but instead looks at how distribution intensity is related
to firm performance (Hartung and Fisher 1965, Naert and Bultez 1975, Corstjens
and Doyle 1979, Rangan et al. 1986).
A possible reason for the lack of empirical research on factors driving
distribution intensity could be that the connection between the product categories
and their inherent level of distribution intensity appears to be such common sense
that it makes empirical test uninteresting. Consider the comparison between the
distribution intensity level of cigarettes and automobiles. Cigarettes are usually
defined as a convenience good while automobiles could be classified as shopping or
specialty goods. Based on product and usage characteristics like unit value,
purchase frequency, consumption time it is easy to see why consumer buying effort
for the two products differs and why cigarettes are distributed via many more retail
outlets per trade area than cars. An empirical test verifying the influence of
product characteristics on inherent distribution intensity would not seem to be too
illuminating.
This study aims to expand our understanding of distribution intensity by
looking at the construct within one product category. It develops an integrated
conceptual framework of the construct based on two major components. The first
framework component deals with the manufacturer and its intended distribution
intensity based on product positioning strategy, the scope of targeted segments, and
manufacturer need for marketing control within the channel. High-end product
positioning, narrow scope of target segments, and high need for control are
hypothesized as factors leading to a small number of retail outlets per trade area
while the opposite would help to explain high levels of distribution intensity for a
product. The second framework component focuses on intermediaries, especially
their availability and their motivation to participate in a manufacturer’s distribution
network. The framework establishes distribution infrastructure, financial
attractiveness, and barriers to entry as major factors driving intermediary
participation in a distribution network. A well developed distribution infrastructure
5
in a trade area, financially attractive products, and low barriers to intermediary
entry are proposed to be associated with high levels of distribution intensity while
the opposite is proposed to lead to smaller numbers of retail outlets in a trade area.
The framework is tested empirically within the audio/video equipment
industry1 . The product line used to test the framework are Hi-Fi loudspeaker
systems. The marketplace shows a wide range of distribution intensity levels for
different brands of speakers. While some manufacturers seem to favor an intensive
distribution strategy, others seem to favor highly selective or exclusive distribution.
Differences in product characteristics between speaker brands are limited and the
number of speaker manufacturers with direct sales networks or vertically integrated
distribution channels is relatively small. Both industry characteristics make
speakers an excellent product example for testing the framework.
This study has several important implications. It will expand our academic
understanding of distribution intensity by closing a gap in conceptualization and
integration of the construct. Explaining variation in distribution intensity within one
product category offers the potential to shed light on underlying factors beyond the
influence of product characteristics. It also represents the first attempt at building
an integrated conceptual framework. The focus on one product category adds a
1 SIC codes for the industry include: 3651 - Household Audio & Video Equipment; 3663 -
Radio/TV Communications Equipment; and 5731 - Radio/TV/Electronic Stores.
6
new dimension for analysis and lays the foundation for potential empirical testing
of previous conceptualizations within the expanded frame of reference. From a
managerial point of view, this study and an improved understanding of distribution
intensity could help managers in their strategic decisions when designing a
distribution channel for a firm’s product line.
The remaining paper is divided into six chapters. Having introduced the
topic, the following chapter reviews relevant literature and examines existing
conceptualizations of distribution intensity as well as empirical research pertaining
to the issue. It tries to integrate previously proposed factors and uncovers existing
gaps in our knowledge. Chapter Three develops a framework of distribution
intensity and derives hypotheses for both components of the model. Chapter Four
deals with the lay-out for empirical analysis of the proposed framework and its
hypotheses, while Chapter Five presents the results of our empirical analysis.
Chapter Six concludes the paper with a discussion of implications we can draw
from this study and what it means to future research.
7
Chapter 2
Literature Review
The two major objectives for the literature review section are to (1) define
the distribution intensity construct and (2) examine what we know about it. For the
little research that has been done on the issue of distribution intensity, it is
surprising to find multiple definitions of the construct and a lack of integration
across research streams in the existing literature.
To clarify the definition of distribution intensity, I examine the existing
definitions for the construct and discuss their strengths and weaknesses. Following
that discussion, existing conceptual and empirical knowledge about the construct is
examined. The majority of research is conceptual in nature and tries to develop
factors that explain the variance in actual levels of distribution intensity for certain
products. Conceptual research in consumer goods proposes product characteristics
as the factors influencing intensity; and research in industrial goods adds other
variables like market and competitive characteristics. Some empirical research is
found in the marketing models literature as well. To round out the picture,
normative models developing channel structure equilibria are also discussed. These
8
models are found in the economics literature and focus on the communication task
of middlemen. They are based on the assumptions of competitive market structure
and the goal of minimizing communication costs.
2.1 Defining the Distribution Intensity Construct
The issue of distribution intensity — sometimes also referred to as density of
market coverage — is basically an issue of the manufacturer’s representation in the
marketplace. It therefore is usually defined as the number of sales outlets carrying
the manufacturer’s products per trading area (Copeland 1923, Aspinwall 1958,
Miracle 1965). The three generally accepted strategic choices with regard to
intensity levels are intensive, selective, and exclusive distribution.
Intensive distribution is used when firms place their product or service in as
many sales outlets as possible in each trade area. The opposite to intensive
distribution is exclusive distribution, which describes the practice of placing the
product or service with only one retail outlet per trade area. Located in between
these extremes, selective distribution uses a mid-range number of retail outlets per
trade area. Manufacturers select outlets subject to certain manufacturer standards
and the middleman’s ability to meet those requirements (Rosenbloom 1991).
Therefore, the actual number of outlets in selective distribution can vary and is
determined by the selection process and rigidity of requirements set by the
9
manufacturer. While this definition of distribution intensity and its
operationalization appears to be straight forward, a closer look at the literature
exposes several problems.
First, there is no universally accepted definition of distribution intensity.
Existing literature uses several different definitions for the construct. Instead of
focusing on the interface between end-customer and the supplier represented by the
number of retail outlets per trade area, some authors center their attention on the
interface between manufacturer and middlemen and look at number of retailers per
independent channel or per channel level2.
One group of authors defines distribution intensity in terms of multiple
representation per channel type. This group usually examines channel structure
issues for manufacturers who use multiple channels to reach their markets.
Examples for such multiple channels would include manufacturer-owned retail
outlets, franchised retail outlets, and OEM channels. Authors in this group define
distribution intensity as the number of outlets in a particular channel of distribution
(Corstjens and Doyle 1979, Rangan et al. 1986, Rangan 1987). Under certain
conditions this definition could have advantages over definitions based on trade
2 Normative models in economics define distribution intensity as the number o f middlemen in
the market (Balderston 1958, Baligh and Richartz 1964). These models are concerned with the
entire channel structure and not just the competitive position of the manufacturer. Since they are
entirely theoretical and do not claim to capture real world complexities, their definition of
distribution intensity is not discussed in detail.
10
areas. If different channels serve different types of customers, if no overlap
between those customer groups occurs, and if there is basically only one trade
area, distribution intensity defined as number of middlemen per channel type would
more accurately reflect the actual level of distribution intensity. In this case, the
original definition based on middlemen per trade area would overestimate the
number of middlemen available per customer group and result in a higher
theoretical intensity level than actually encountered by the end-customer in the
marketplace. Some industrial goods markets could be envisioned as having such a
channel structure. Generally though, the definition based on number of middlemen
per channel can pose a problem since customer segments between different
channels often overlap. In that case, the definition of intensity based on number of
outlets per channel is not valid anymore, since the actual intensity faced by the
customer may consist of the combined number of outlets of several channels.
Another group of authors defines distribution intensity as the number of
middlemen per channel level (Stern and El-Ansary 1988, Rosenbloom 1991, Kotler
and Armstrong 1991). The problem with that definition is that different levels of
distribution could have different levels of intensity. Which of those different levels
would be most important to the firm’s success in the marketplace? It could be
argued that the interface between end-customer and retail outlet is the most
important link in the channel chain, since it establishes most of the customer
perception of the product and its manufacturer. Therefore, the lowest channel level
11
should receive the most attention in regard to what distribution intensity is sought
by the manufacturer. If the focus is on the retail level and the total market consists
of more than one trade area, then a focus on number of middlemen per channel
level (retail level) could be misleading since it does not give any information on
the distribution of those retailers across the different trade areas. While the overall
level of intensity across trade areas could be selective, the specific level of
intensity for certain trade areas could range from exclusive to intensive.
Even for the group of authors who base their definition on trade areas, a
subtle variation exists. Some authors use the retail outlet share instead of number
of retail outlets to define the intensity construct (Hartung and Fisher 1965, Naert
and Bultez 1975). These authors examine distribution intensity for channels
characterized by brand exclusivity, which means that retail outlets can carry either
the manufacturer’s brand or competitors’ brands but not both, and conceptualize
the construct as retail outlet share per geographical area. Retail outlet share refers
to the number of outlets carrying the manufacturer’s product in relation to total
number of outlets representing either the manufacturer’s or competitive brands.
The major shortcoming of this definition is that dependence on the relative number
of outlets clouds the issue. For example, a firm having 80% of the retail outlet
share could have 4 outlets in the trade area while the competition has only 1, or it
could have 80 outlets in comparison to the competition’s 20 outlets. In both cases
the 80% outlet share could point towards intensive distribution, while analysis of
12
the absolute numbers shows that the actual intensity could range from selective to
intensive distribution.
Taking into consideration the strengths and weaknesses of all discussed
options, this paper defines distribution intensity as the number of sales outlets per
trade area. A definition based on the absolute number of retail outlets in
comparison to retail outlet share is preferred because it more accurately reflects the
actual density of market coverage. Trade area for the purpose of this study refers
to a geographical area delineated by natural boundaries which influence buyer
shopping patterns. Trade area definitions advanced in literature are usually based
on either manufacturer, retailer, or consumer considerations. For example, Stern
and El-Ansary (1988) refer to trade area as the area from which a retailer draws or
expects to draw the vast majority of his customers. My definition of trade area
avoids such a strict focus. While the consumer is primarily concerned with product
availability in his area of consideration, that area might vary in size for different
consumers or different purchase occasions. On the other hand there will be some
upper limit as to what shopping area a consumer considers for the purchase of
certain products. Trade area delineated by natural boundaries which influence
shopping patterns seems to be a sensible solution to take both aspects into account.
13
2.2 Distribution Intensity and the Categories of Goods
Framework
The concept of distribution intensity in the marketing literature is usually
anchored in the "Categories of Goods Framework," originally established by
Copeland (1923) and extended by Aspinwall (1958), Bucklin (1963), and Miracle
(1965). Copeland’s original focus was practice oriented by deriving marketing
strategy implications from consumer buying habits. The framework extensions by
Aspinwall (1958) and Bucklin (1963) were intended to improve our conceptual
understanding of theory. All approaches have similar conceptual implications for
distribution intensity though.
The term "Categories of Goods" refers to product classifications based on
consumers’ needs and purchase efforts. Distribution intensity fits into the
framework under the premise that a relationship exists between consumer purchase
effort, the resulting product categories, and the appropriate number of retail outlets
carrying a product of a specific category. Copeland (1923) develops three product
classes based on consumer buying habits:
(1) convenience goods, which are goods the consumer is familiar with,
has a frequent need of, pays a relatively low price for, and
purchases with a minimum of effort;
(2) shopping goods, described as goods for which the consumer wants to
compare price, quality, and other important product characteristics at
the time of purchase; and
14
(3) specialty goods, which are goods with a special, non-price-related
characteristic or attraction causing consumers to go through special
efforts to visit the store in which they are sold and make the
purchase without much shopping.
These product classes are associated with certain marketing strategies of the
manufacturer. Using the three product classes, Copeland develops several
implications for marketing strategy, one of them being intensity of distribution.
Copeland argues that convenience goods require intensive distribution
because of the consumer’s desire to purchase the goods with a minimum of effort.
This means goods must be available in easily accessible stores. Shopping goods do
not require as intensive distribution as convenience goods given the fact that
consumers are willing to go through some shopping effort, travel greater distance,
and delay the purchase if necessary. Therefore, selective distribution seems
appropriate for shopping goods. On the other extreme from convenience goods,
specialty goods can be distributed through exclusive distribution, given the
assumptions that specialty goods are infrequently purchased, consumers are willing
to go through special efforts to reach the store selling them, and the manufacturer
needs to find retailers capable of using aggressive marketing strategies to attract
customers to these goods.
The logical chain between categories of goods and distribution intensity can
be described as a two step link. First, specific product characteristics influence
15
consumers’ willingness to go through various levels of physical shopping effort for
a specific product or brand. And second, the level of physical shopping effort
influences the density of market coverage necessary to avoid losing sales to the
competition. The essence of Copeland’s framework is that product characteristics
influence distribution intensity via product classes. Convenience goods require
intensive distribution, shopping goods should be selectively distributed, and
specialty goods can afford exclusive distribution.
Aspinwall’s (1958) extension develops five factors purported to determine
product classes: (1) frequency of purchase, (2) consumption time, (3) search time,
(4) extent of adjustment between production and final consumer specifications, and
(5) gross margin. He rates products on the continuum of these five factors and
arrives at three product classifications that are very similar to Copeland’s product
classes.
"Red" goods are characterized by high purchase frequency, low
consumption time, low search time, low product adjustment needs for end-
customers, and low gross margin, which makes them very similar to Copeland’s
convenience goods. Aspinwall argues that "Red" goods need intensive distribution
because consumers’ high purchase frequency with low search time translates into
minimum purchase effort on the part of the consumer. Therefore, distribution
16
strategy has to stress the key factors of product exposure and purchasing
convenience.
"Yellow" goods have the exact opposite characteristics of "Red" goods and
are very similar to Copeland’s specialty goods. The author reasons that they are
best served by exclusive distribution. Consumers are willing to go through
extensive purchase efforts but demand a high level of product adjustments to meet
consumers’ specification demands. Distribution strategy therefore focuses on a few
highly committed and knowledgeable distributors, who are able to establish dealer
service that meets the high quality levels expected for such products.
"Orange" goods are located in between "yellow" and "red" goods. In regard
to frequency of purchase, consumption or search time, extent of product
adjustments for end-customers as well as gross margins, they have medium ratings.
"Orange" goods compare to Copeland’s shopping goods. Aspinwall suggests
selective distribution as the most appropriate for these goods since consumers are
willing to go through some purchase effort, but not as much as for "yellow"
goods.
Aspinwall’s classification system is very similar to and his implications for
distribution intensity are virtually the same as Copeland’s. One reason for that
similarity might be that Aspinwall’s five product characteristics are more or less
17
detailed expressions for Copeland’s more intuitive reasoning based on product
familiarity, frequency of need, and price.
Miracle’s "Product Characteristics and Marketing Strategy" (1965) is
different to the previous two conceptualizations in several aspects. Probably the
most important difference is Miracle’s conceptualization of intensity levels. While
Copeland and Aspinwall distinguish between intensive, selective and exclusive
distribution, Miracle distinguishes between intensive and selective distribution only.
He defines selective distribution as the selection of outlets according to their
capability and suitability to serve manufacturers overall marketing strategy.
Intensity of distribution is proposed to be inversely related to channel length, which
in the case of extreme selectiveness would lead to direct sales to customers instead
of going through independent channels.
Miracle proposes five product classes based on nine classification
characteristics. He recommends intensive distribution for his "Group I" products,
which are low on the dimensions of unit value, significance of each individual
purchase to consumer, purchasing time and effort, rate of technological change,
technical complexity, consumer service need, and high on the product dimensions
of purchase frequency, consumption rapidity, and extent of product usage. For the
opposite extreme, products in "Group V", he recommends highly selective
distribution. These products are characterized by high unit value, significance of
18
individual purchase, purchasing time and effort, rate of technological change,
technical complexity, consumer service need and low purchase frequency,
consumption rapidity, or extent of product usage. The product classes in between
vary to different degrees on one or more of the nine factors. His "Group I" closely
resembles Aspinwall’s "Red" goods or Copeland’s convenience goods, while his
"Group V" is similar to "Yellow" or specialty goods. His reasoning for the
intensity of distribution levels closely follows Aspinwall’s (1958) reasoning.
Miracle’s creation of five product classes adds complexity without adding new
insights, which might be the reason why his framework failed to gain major
recognition.
Summarizing the category of goods framework and its conceptualization of
distribution intensity, we see a two step link. Product characteristics influence
consumer purchase effort, which in turn determines distribution intensity. Purchase
effort is mainly the physical effort exerted by customers. In the literature, the
distinction between product classes and the question of how to define these classes
has been debated extensively (Copeland 1923, Aspinwall 1958, Holton 1958, Luck
1959, Bucklin 1963, Groeneveld 1964, Miracle 1965, Kaish 1967, Mayer et al.
1971, Holbrook and Howard 1977, Murphy and Enis 1987). While this discussion
did not focus on distribution intensity, it had a direct impact on the issue. Through
the discussion on what that basis for product categorization should be, it was
realized that objective product characteristics might not be the only basis for
19
distinction. Another basis for product categorization is the psychological aspect of
shopping. Consequently, product classes based on preference maps, and consumer
purchase risk as a major factor for their classification were introduced (Bucklin
1963, Howard and Holbrook 1977).
The important difference in categorizing products based on preference maps
is taking into account the influence of psychological factors driving the purchase
effort with their indirect impact on distribution intensity. The preference map
framework (Bucklin 1963) establishes two classes of goods: Shopping goods and
non-shopping goods. For non-shopping goods, the consumer has an established
preference map in his mind in regard to the product to be purchased. He is either
willing to accept any number of substitutes to the brand originally sought, which
would make the product a convenience good, or he accepts no substitute, which
makes it a specialty good. In the case of shopping goods, no preference map exists
prior to the need, but is developed during the actual process of shopping. The
implications for distribution strategy, and intensity in particular are important.
Based on this framework it is conceivable that two different brands of a
product, with roughly the same product characteristics, could be positioned
differently in the consumer’s mind. Consequently, it is not necessarily the profile
of product characteristics which drives the need for different levels of distribution,
but the manufacturer’s ability to establish a preference map in the consumer’s
20
mind. In establishing a preference map either accepting or refusing substitutes, the
total marketing effort by the manufacturer determines whether intensive or
selective distribution is appropriate. In addition, the strategic distribution intensity
decision for shopping goods becomes more complex. The correct distribution
intensity is basically driven by the consumer ability and effort in establishing a
personal preference map. If that process is short and ends with the acceptance of
substitutes, a manufacturer must establish a more intensive distribution network
than if the process is extensive and ends with a preference map not accepting
substitutes. As Dommermuth (1965) points out, that process is by no means
homogeneous for products in the same product class, but instead is driven by
segmentation factors like brand image, or customer demographics, etc.3
Holbrook and Howard (1977) introduced the dimension of "risk" as a
discriminating factor in their typology of product classes and added the new class
of "preference goods" to the existing categorization framework. Preference goods
are similar to convenience goods and mainly discriminated by a higher level of
social and psychological risk involved in the product purchase (Murphy and Enis
1987). In an effort to match the four product classes with distribution intensity
levels, Murphy and Enis (1987) recommend "saturation distribution" as the
appropriate distribution strategy for convenience goods and "intensive distribution"
3 To counter this added complexity in analyzing the appropriate distribution intensity,
Dommermuth (1965) proposes the use of a shopping matrix analysis.
21
for preference goods. Unfortunately, no explanation on how saturation distribution
differs from intensive market coverage is given. Since intensive distribution is
already defined as using all available outlets in an area, an increase in density of
market coverage is hardly feasible. Therefore, the term saturation distribution
seems to be redundant.
While most conceptualizations of distribution intensity are solely based on
product categories, Bucklin (1963) added retail outlet characteristics and used a
combination of product and retail outlet categories to determine appropriate
intensity levels. Based on the notion that retail outlets as well as products can be
distinguished by the same characteristics, he classified retail outlets into
convenience, shopping, and specialty stores, using the existence of preference maps
in consumers’ minds and the willingness to substitute shops as the basis for
category distinction. Following Bucklin’s classification, Marcus et al. (1975)
develop strategic marketing implications and argue that intensive distribution is
most likely for convenience stores with convenience goods or shopping goods,
because the convenience of store location overpowers product considerations. The
consumer shops in the most accessible store and purchases the most readily
available brand in that store or he goes through a selection process considering the
brands carried. On the other hand, highly selective to exclusive distribution is most
likely when specialty patronage habits regarding brand or store are involved.
22
In summary, the focus on customer needs and shopping preferences
dominates the above mentioned work, which is based on the categories of goods
framework. Consumer needs and shopping preferences are the key factors for the
classification of goods and/or retail stores and, consequently, to distribution
strategy. Surprisingly, while the theoretical concepts of classifying products by
purchase effort and product characteristics thrived, little empirical research
verifying these concepts has been published (Dommermuth 1965, Dommermuth
and Cundiff 1967, Kleimenhagen 1966, Mason and Mayer 1972). The empirical
research that is published only tries to verify the theoretical product classifications
and their relation to purchase effort, but does not deal with the link between
product categories and distribution intensity.
2.3 Distribution Intensity in the Industrial Goods Channel
While the consumer goods oriented product categorization framework is the
main building block for the distribution intensity conceptualization several other
factors influencing distribution intensity have been introduced in the industrial
goods literature (Webster 1976, Corey et al. 1989).
Webster (1976) deals with the issue of distribution intensity in a study
dedicated to the examination of the role of industrial distributors in light of the
changing marketing environment. He introduces the normative aspect of
distribution intensity by assigning to the distributors the two key responsibilities of
23
covering the market and making the product available. The first key responsibility
is to contact present or potential customers, and the second one is to make the
product available to customers as quickly as economically feasible. Based on a
field study, Webster lists several factors which determine the number of
distributors necessary to cover the market and ensure product availability. Some of
his factors are similar to previously developed variables. He postulates that
intensive distribution is necessary for products with high purchase frequency,
which basically follows Copeland (1923) and Aspinwall (1958). He also finds that
competitively undifferentiated products require firms to use intensive distribution
networks, which can be derived from Bucklin’s (1963) preference map reasoning.
In addition, Webster introduces several other, new factors positively related to
distribution intensity:
(1) total market potential and its geographic dispersion;
(2) the manufacturer’s current market share and the intensity of
competition;
(3) the amount of technical knowledge required to sell or service the
product;
(4) the degree of interruption for the customer’s production process,
caused by the lack of product availability; and
(5) whether the product is an MRO or OEM item.
Since Webster’s article was not intended to be a study on distribution
intensity and dealt with the issue only on a side, it does not offer a supporting
24
rationale for these factors and their influence on distribution intensity. The factors
are merely listed as empirical results of the field study and not discussed in any
detail. Nevertheless, Webster’s article is important, for it not only acknowledges
customer needs and shopping preferences based on product characteristics, but also
integrates competitive and geographical market aspects in its list of factors
influencing distribution intensity.
Corey, Cespedes and Rangan (1989) look at distribution intensity from the
manufacturer’s strategic point of view. The rationale for a manufacturer deciding
to design a channel with multiple representation in a geographical area rests with
the assumption that different distributors can reach different market segments.
Unfortunately, market segments are seldom non-overlapping and multiple
representation can have the serious drawback of intense intrabrand competition. On
the other hand, a supplier who reduces representation in the area to reduce price
competition between his resellers may run the risk of also losing sales volume. The
distribution intensity decision has to strike a sensible balance between these
considerations. The authors list three major factors influencing the decision on how
many resellers a manufacturer should franchise in a local market: (1) investment by
intermediaries, (2) sales potential in trade area, and (3) customer purchasing
behavior.
25
The first factor is the level of investment by intermediaries necessary to
service the line in the market and covers two aspects: (1) the investment in
inventories required to support a given level of sales and (2) the investment in
specialized resources such as equipment and personnel. The level of investment
required by intermediaries is proposed to be negatively correlated to channel
intensity. In case of high investments, distribution intensity must be lower to attract
resellers for the product line. An additional aspect of necessary dealer investments
is the life-cycle stage of the product. In the early stages of market development,
when the returns are uncertain, dealer investments could be jeopardized if they
have direct competitors in the same trading area (p. 79). Therefore, early product
life-cycle stages are more likely to have exclusive or highly selective intensity
patterns.
The second factor influencing distribution intensity is the available business
volume in the trade area. The authors contend that there is some level of potential
revenue below which it may be difficult to attract established resellers. Therefore,
the lower a manufacturer’s market share in the area, the fewer the parts the pie can
be divided into and the fewer the resellers in the area.
The third factor is the purchasing behavior of customers. The authors list
three aspects of the purchase process that influence the decision of how many
resellers should be established in a trade area. The first aspect is whether buyers
26
engage in a search process or not. They argue that the number of retailers in an
area can be lower for products requiring a search process on the part of the buyer.
This situation is representative of shopping goods and gives a manufacturer’s
product a greater likelihood of being considered even if it is not carried by every
distributor (Copeland 1923, Aspinwall 1958, Bucklin 1963). The other two aspects
are whether the purchase is typically unbundled or not, and whether the decision
making process involves negotiations about product features, prices, and terms.
The authors argue that in the case of unbundled purchase patterns or decision
making processes involving negotiations fewer resellers per trade area are needed.
Both aspects stress the customizing aspect of the purchase process. Following
Aspinwall’s (1958) reasoning, distribution intensity is lower when end-customer-
required product adjustments are part of the distribution process. For all three
aspects one can argue that the necessity of search, an unbundled purchase pattern,
or negotiations during decision-making accommodate lower distribution intensity
levels because the customer need for knowledgeable distributors overrides his need
for convenience.
2.4 Distribution Intensity and Firm Performance
A third stream of research concerned with the intensity concept are
marketing models of sales response that include distribution intensity as a key, if
not sole, factor. The marketing models literature represents the few empirical
research projects in which distribution intensity is a central or one of the major
27
constructs. While these marketing models work with the construct of distribution
intensity, their focus is significantly different from the theoretical research
discussed so far. In the literature discussed so far, the authors try to explain factors
influencing distribution intensity, making intensity the dependent variable. The
models literature takes distribution intensity as the or one of the independent
variables explaining firm performance like profit, sales, or market share. These
models focus on an output function to be optimized, how this output function is
influenced by distribution intensity, and what levels of distribution intensity are
appropriate (Hartung and Fisher 1965, Naert and Bultez 1975, Corstjens and Doyle
1979, Lilien and Rao 1976, Rangan, Zoltners and Becker 1986, Rangan 1987,
Farris et al. 1989). By solving distribution problems confronting a single
manufacturer, these models can be classified as normative decision support models
within the managerial marketing research stream (Rangan 1987).
One of the major issues dealt with in models research is what kind of
functional relationship exists between distribution intensity and performance.
Several studies contend that an increasing number of outlets in a channel is not
simply an additive function for total sales (Hartung and Fisher 1965, Corstjens and
Doyle 1979, Rangan et al. 1986). Instead, depending on the behavior of
intermediaries, increasing distribution intensity can have either a promoting effect
leading to higher sales, or a cannibalization effect leading to diminished sales.
28
Hartung and Fisher (1965) support the notion of a promotional effect. In
their model, manufacturer market share is a function of retail outlet share and their
empirical tests show that increased retail outlet share pushes market share upwards.
The model assumes brand exclusivity, that is stores can carry either the
manufacturer’s brand or competitors’ brands but not both at the same time and
retail outlet share is defined as the number of outlets carrying the manufacturer’s
product in relation to total number of outlets representing either the manufacturer
or competitive brands.
On the other hand, a cannibalization effect could be encountered when a
high distribution intensity leads to lessened commitment on part of the
intermediaries and ultimately could lead to diminished sales (Corstjens and Doyle
1979). In such a case, intermediaries would feel that the manufacturer does not
care about the single dealer by protecting him from intrabrand competition.
Therefore, dealers would feel less committed to push the brand but instead would
try to push competing brands, for which they have less competition or even
exclusive rights in the trade area.
An interesting result in regard to the issue can be found in Farris’ et al.
(1989) research, which was done in the context of push versus pull marketing
considerations. The authors examined consumer convenience goods and observed
not only a positive relationship between market share and distribution intensity but
29
detected increasing returns in regard to market share. Instead of diminishing gains
in market share with ever higher levels of distribution Farris et al. observed the
opposite effect, which clearly supports the promotional effect of distribution
intensity.
Another major issue is the question of effectiveness. Several models assume
there is an optimal number of sales outlets in each market that realizes the market
potential (Rangan et al. 1986, Hauser 1986). Hauser’s (1986) defender model tries
to develop prescriptions for marketing action in cases when an incumbent firm is
faced with a new competitor in its market. The model takes into account all
different marketing aspects, one of them being distribution. In regard to
distribution, the model focuses on increased product availability through investment
in distribution channels. It shows that the best defensive distribution strategy would
be decreased spending on distribution, if the market size does not dramatically
increase due to competitive entry. While this result may be counter-intuitive, the
logic behind it assumes that investment in distribution should create or increase
product availability. Hauser (1986, p. 124) reasons:
[If the market size does not increase, the competitive entry has made
the market less profitable], hence, the marginal profit for marginal
retailers is less. If there is no change in the marginal cost persuading
these retailers to make the product available, it no longer makes
economic sense to invest in those marginal retailers.
30
Of course, when we look at the total cost structure of distribution, possible
economies of scale exist for an increased number of outlets in a channel, e.g. the
unit savings in transportation costs, etc. (Corstjens and Doyle 1979).
Another aspect in the models literature is noteworthy. Most models take a
look at intermediary size (Corstjens and Doyle 1979, Rangan 1987, Rangan et al.
1986). Essentially, the models argue that increased intermediary size enhances
middlemen power. At a certain size or power level the intermediaries are assumed
to lessen their cooperation efforts, thereby decreasing the manufacturer’s channel
control.
In summary, the marketing models described above are helpful in three
ways. First, they point out that the addition of intermediaries does not necessarily
yield an additive effect in regard to sales. Instead, depending on intermediary
behavior, increasing distribution intensity can have a promotional effect leading to
increased market share for the product, or a cannibalization effect leading to
decreasing sales for the product. Second, the issue of effectiveness as major aspect
of distribution intensity is brought to light. It is assumed that the existing market
potential can be realized with an optimum number of intermediaries. Third, the
models touch on control and power issues. If intermediaries get too big,
cooperation with the manufacturer can suffer because of intermediary power and
not completely congruent sets of goals between intermediary and manufacturer.
31
Therefore, in order to keep control of the channel, manufacturer’s could be
tempted to limit the size of their intermediaries.
2.5. Distribution Intensity and Equilibrium Channel Structure
Economic models focus on the design of equilibrium channel structures and
policies. They are concerned with the entire system and not just the manufacturer
and his market position. The channel equilibrium, which these models arrive at, is
based on profit considerations and assumptions of a competitive market structure
(Balderston 1958, Baligh and Richartz 1964). The basic thrust of these models is to
give insights and to explain what channels would look like under certain conditions
and in the long run. They offer no empirical validation, nor do they claim to
capture real world complexities (Rangan 1987).
Recently, some economic models have relaxed the assumptions of pure
competition and dealt with channel structure invoking game and contract theory
(Zusman and Etgar 1981), quantity discounts as channel coordination mechanism
(Jeuland and Shugan 1983), and product differentiation’s impact on vertical
integration (McGuire and Staelin 1983). Unfortunately, none of these models focus
on the aspect of distribution intensity within the channel. The two models that
explicitly try to determine the number of middlemen to be found in the market are
by Balderston (1958) and Baligh and Richartz (1964).
32
Balderston (1958) determines the number of wholesalers to be found in the
market based on the performance of the communication task and its associated
costs. He models a channel with communication links between manufacturers,
wholesalers and retailers. The wholesaler is basically the switchboard of a
communication network with no intermediary inventories. The suppliers are the
manufacturers, and the customers are the retailers and their numbers are
determined by external variables. His model arrives at the equilibrium number of
middlemen being equal to the product of manufacturers and customers divided by
the sum of manufacturers and customers4. Balderston’s model is criticized mainly
for two reasons: its assumptions5 and its restriction to develop an equilibrium for
only one layer of intermediaries.
Balderston assumes that every firm among the suppliers must be in contact
with every customer firm (retailer), and any intermediary (wholesaler) must be in
contact with all firms in the channel levels above and below. This assumption is
necessary to give the system an "adequate bargaining" characteristic. It is assumed
that any firm requires information on all possible suppliers or buyers in the channel
levels adjoining its own before it enters into an exchange (Balderston 1958 p. 157;
4 The equilibrium as a formula presents itself as follows: wc = (S * C)/(S 4- C); with: we -
equilibrium number o f intermediaries; S - number of supplier (manufacturers) in the market; C -
number o f customers (retailers) in the market.
5 For a concise review o f Balderston’s explicit and implicit assumptions, see Baligh and
Richartz (1964).
33
Baligh and Richartz 1964 p. 669). Bucklin (1966) questions this assumption and
asks under what conditions one would expect all buyers to contact all sellers in the
market. Bucklin (p. 65) states:
From an analysis of search it can be shown that these occasions are
likely to be rare. Furthermore in the centralized markets only a
single contact for each buyer and seller may be sufficient to permit
exchange. In this instance there is no opportunity for middlemen to
reduce the number of links in the channel.
Also, one could ask under what circumstances either buyers or sellers would be
expected to permit the number of links to be reduced. Bucklin lists the example of
a market for a non-differentiated product in which buyer and sellers make
approximately five contacts before agreeing to an exchange. Since market
participants show that behavior to avoid ownership risk, it seems unreasonable to
Bucklin that they would abandon this behavior to deal with a monopoly
middlemen.
The other criticized assumption is that all firms in the system are indifferent
to the identity of the firms they deal with and to the profits those firms make,
provided that their own profits don’t suffer. Baligh and Richartz (1964) find this
assumption questionable and introduce the notion of segmented markets into the
model in order to be more realistic. The effect of segmentation, that is the
existence of market structures in which buyers have a preference for some sellers
34
and vice versa, proves to decrease the equilibrium number of intermediaries in that
market. In response to Balderston’s restriction to develop an equilibrium for just
one intermediary level, Baligh and Richartz’ model develops a solution for several
intermediary levels. In addition to determining the number of intermediaries per
level they also determine the number of intermediary levels for the market to be in
equilibrium. While their model is more general than Balderston’s, it does not
introduce new insights about distribution intensity.
In summary, the economic models for distribution intensity are mainly
efficiency driven. While they give us a good perspective on how cost reduction for
the system could lead to an equilibrium number of intermediaries, their
shortcomings, especially the lack of realism, restrict their usefulness for research
in distribution intensity.
2.6 Integration and Evaluation of Existing Literature
Until now, no research has been available which specifically tries to
develop a conceptual framework for the distribution intensity construct. While
some marketing models specifically focus on distribution intensity, they are not
concerned with a conceptual framework that explains variation of distribution
intensity, but instead try to establish a relationship between intensity levels and
firm performance. Other than that, the construct has been dealt with as a side
aspect to general distribution structure or marketing mix questions. Still,
35
conceptual research lists a host of factors regarded as influencing distribution
intensity. Unfortunately, the body of work on intensity leaves us with two major
deficiencies: (1) a lack of conceptual integration of all factors proposed to
influence distribution intensity and (2) a lack of empirical research verifying these
factors in regard to actual distribution intensity in the market place.
The first deficiency is the lack of conceptual integration in regard to the
various approaches to channel intensity. The "Category of Goods Framework" and
its extensions represents important contributions to the issue of distribution
intensity. Product characteristics and categories influencing customers shopping
preferences play a major role in determining the appropriate market coverage
between the product categories of convenience, shopping, and specialty goods.
Unfortunately, market coverage between product categories is only half of the
picture. Variations in market coverage within an individual product category should
be explained too. Work in the industrial goods channel is important as it expands
the number of influential variables by introducing competitive and market factors.
The aspect of intermediary willingness to participate in and join a distribution
network is an important contribution. The marketing models dealing with
distribution intensity are important because they represent the second link in the
chain between factors influencing distribution intensity and the distribution intensity
influencing firm performance. A combination of both research streams with the
addition of other important influential variables and the empirical verification of all
36
factors would be an important first step to a normative model for distribution
intensity strategy.
The other major deficiency of existing literature is the lack of empirical
research on verification of influential factors. All previously developed conceptual
frameworks lack the empirical research necessary to test and verify their major
propositions. The present study tries to provide some remedy for both gaps. It tries
to flush out the strategic perspective of distribution intensity from the
manufacturer’s point of view and tries to develop a conceptual framework of
intensity reflecting that strategic perspective. And it also designs an empirical test
to verify the model and its propositions.
37
Chapter 3
The Framework
This study develops a framework which is intended to explain variation in
levels of distribution intensity across brands within one product line in an industry.
While the existence of general product categories like convenience, shopping, and
specialty goods is conceptually important, the differences in distribution intensity
between these product categories are quite intuitive and seem hardly interesting for
academic research. On the other hand, explaining variation across brands within
one product line offers the opportunity to shed light on influential factors beyond
product characteristics and to add in-depth understanding helpful in mastering
overall distribution strategy.
The framework is based on the premise that customers are not
homogeneous in their shopping pattern or shopping effort even if we consider only
one specific product line. Manufacturers can enhance and exploit these differences
through their marketing effort, especially product differentiation. Consider the
example of wrist watches. Wrist watches, which generally would be associated
with shopping goods and selective distribution intensity, are highly differentiated
38
products in the market place. While most of the wrist watches have similar
functional and usage characteristics, it is easy to classify various brands in different
product classes. The variation in positioning strategies ranges from convenience
good positioning and intense distribution in the case of "Timex" to specialty good
positioning and highly selective distribution in the case of "Pattek Phillipe" or
"Rolex". Efforts to move a brand away from a generic product category and
distinguish it from competitors are common and usually supplier initiated.
Therefore, this framework looks at distribution intensity from a supply side
perspective, even though it acknowledges the consumer as the focus of all
marketing efforts.
It is proposed that the level of distribution intensity is determined by an
interdependent process between manufacturer intentions and intermediary
cooperation. This is done for two reasons. First, the integration of an intermediary
perspective reflects the importance of independent retailers in today’s market
system. While choices in regard to the number of retail outlets per trade area have
to be made by manufacturers for vertically integrated as well as non-integrated
distribution networks, the discussion of intensity strategies is generally most
directly applicable to channels comprised of independently owned institutions and
agencies (Stern and El-Ansary 1988). Second, from a managerial perspective the
distinction between intended number of retail outlets carrying the brand and
realized distribution intensity is important. Only realized distribution intensity
39
reflects the actual situation a consumer is faced with when shopping for a brand
which in turn directly influences firm performance. Potential customers are not
interested in what the original intention of the manufacturer was, but only how
many and what kind of stores actually carry the brand. And the actual number of
retail outlets carrying a brand is inherently dependent on intermediary cooperation.
As a result, the proposed framework features two major components, a
manufacturer component dealing with manufacturer influence on intensity, and an
intermediary component dealing with restrictions for establishing distribution based
on intermediary availability and willingness to join the distribution network. The
framework is shown in Figure 3.1.
For the manufacturer component, the framework lists three major factors
influencing distribution intensity: (1) product positioning, (2) target segment focus,
and (3) manufacturer need for channel control. High-end product positioning, a
narrow scope in target segments pursued by the manufacturer and a high need for
control are hypothesized to lead to low distribution intensity, with the opposite
leading to a large number of retail outlets per trade area. The two factors of
product positioning and target segment focus are also posited to have an indirect
effect on intensity through their influence on need for channel control. With high-
end product positioning and a narrow scope in target segments, manufacturers are
more likely to have a high need for control.
40
Figure 3.1: A Conceptual Framework of Distribution Intensity
Product
Positioning
Target
Focus
H 5: (+) H 4: (+)
Need for
Control
H 3: (-) H 2: (-)
Distribution
Intensity
H 6: (+)
H 8: (-)
Financial
Attractiveness
Barriers to
Intermediary
Participation
H7c: (-)
H7a: (+)
Investment
Level
Management
Support
H7b;/(+) H7d\(+)
Relationship
Uncertainty
Contractual
Restrictiveness
Manufacturer
Aspect
Intermediary
Aspect
The intermediary component includes the constructs of financial
attractiveness and barriers to entry, both of which represent middleman motivation
to join a distribution network. Barriers to middleman entry are conceptualized as
(1) specialized investments, (2) managerial support for middlemen, (3) relationship
uncertainty between middleman and manufacturer, and (4) contractual
restrictiveness between the channel partners. Financially attractive brands and low
barriers to intermediary entry drive up intermediary participation in a distribution
network and are proposed to be associated with high levels of distribution intensity.
The opposite is proposed to lead to smaller numbers of retail outlets in a trade
area.
3.1 The Manufacturer Component of Distribution Intensity
The following section introduces the three major constructs proposed to
influence distribution intensity from a manufacturer’s perspective. The section first
elaborates on the direct effect of product positioning, targeting focus, and need for
control on distribution intensity. It then introduces the indirect effects by explaining
the influence of product positioning and targeting focus on need for control.
3.1.1 Product Positioning
Product positioning refers to a manufacturer’s efforts to select a set of
product benefits that management wants to convey to a defined target segment to
42
meet its needs (Assael 1985). The target segment is defined as a group of
prospective customers with similar needs (Assael 1985). The actual product
position depends on how potential customers perceive the product’s characteristics
and features in comparison to their needs and in comparison to competitive
products. Product positioning, therefore, is based on the manufacturer’s ability to
convey a set of product benefits to a defined target segment (Kotler 1989).
Positioning strategy involves finding an attribute space for the firm ’s brand within
the range of competitive brands or products and, then, placing the brand in that
space.
Approaches to product positioning vary and products can be positioned on a
multitude of dimensions (Assael 1985, Aaker and Shansby 1982). At the same time
manufacturers should strive for a brand concept coherent with any product
positioning. In a paper dealing with brand concept management, Park, Jaworski
and Mclnnes (1986) argue that only a coordinated marketing effort which
communicates a coherent message to potential consumers will be successful in the
long run. Consequently, the brand should be positioned within the boundaries of a
predetermined brand concept6, and communication activities must be coordinated
with other marketing mix activities including distribution parameters to achieve a
coherent consumer perception of the brand.
6 Park et al. (1986) define brand concept as the brand meaning derived from basic consumer
needs. The basic consumer needs influencing the selection of a brand concept are functional,
symbolic, or experiential needs.
43
The present framework focuses on the price and quality relationship as the
major product positioning dimension. Price-quality product positioning is
conceptualized within the extremes of value-driven versus prestige-driven
positioning. In the case of "value" positioning, the manufacturer focuses on low
price as the differentiation criterion versus competitors. On the other end of the
spectrum, "prestige" positioning features high priced products with apparently
higher quality. The manufacturer focuses on performance and prestige image
aspects to differentiate his product from the competition. Products in between the
two extremes can be influenced by either positioning strategy. Manufacturers of
these "middle-class" products differentiate their brands with respect to one or the
other extreme. They could either focus on (1) higher quality than the value
products but still a low price or (2) only a small quality disadvantage to the
prestige products but a substantially lower price.
The proposed use of the product positioning concept shares some
similarities with cost-leadership or differentiation strategies listed by Porter (1980).
Cost-leadership strategies usually offer a standardized product at low per-unit cost
for price sensitive buyers and tend to favor price-driven competition.
Differentiation strategies, on the other hand, try to offer unique product or service
features for price insensitive buyers and are more performance driven. They are
located at the upper end of the price/quality dimension of product positioning.
44
Cost-leadership strategies are assumed to be price driven, with the intent to
offer a lower price than the competition (Porter 1980). One of the features of such
strategy is the focus on market share or a large number of units sold. The
competitive advantage is the price differential versus competition, and it can be
gained by means of economies of scale. To reach a large number of units sales, the
manufacturer has to make it easy for consumers to buy the product. In addition to
extensive market coverage the manufacturer can reduce the shopping effort by
increasing his representation in a trade area. A manufacturer would increase the
number of retailers carrying his brand as long as the marginal contribution of each
additional retailer is larger than the manufacturer’s marginal distribution costs for
the added retailer (Hauser 1986, Hardy and Magrath 1988).
Another effect of increasing multiple representation in a trade area is the
increased level of intra-brand competition, which helps keep the retail price of the
brand down and improves the competitive position vis-a-vis other "value" brands.
Higher distribution intensity with more retailers carrying the same brand in a given
trade area will most likely lead to more competition between retailers carrying that
brand. For "value" brands focusing on price differentiation instead of performance
features, the ensuing competition will be price driven. Even though a side effect of
such price competition could be that dealers lower their service quality due to
profit pressure, such dealer quality reduction is not detrimental to the
manufacturer’s strategy. Since customers are price driven, the loss of service will
45
not significantly diminish the increased sales potential due to lower prices for that
brand. Therefore, using high distribution intensity for cost-leadership strategies
may pay off by initiating or keeping alive intrabrand competition between brand
carrying retailers.
On the other hand, brands at the high end of the positioning spectrum are
usually less volume driven. These "prestige" brands generally register fewer unit
sales but higher profit margins per unit. Manufacturer’s following such strategy
tend to limit the number of their retailers per trade area in order to keep the
prestigious image of the brand. The intent is to avoid price driven intra-brand
competition which could hurt the image of the brand. Intrabrand competition could
be detrimental to a congruent and high-end brand concept because it is easier to
reduce the retail price than to increase dealer quality features, both from a time
frame for implementation as well as from an investment point of view. Therefore,
price-based intrabrand competition leading to erosion of dealer service quality due
to profit pressures is undesirable for brands sold through differentiation strategies.
Lower distribution intensity with fewer retailers carrying the same brand in a given
trade area may limit the competitive pressures on the retail network for that brand
and make it easier for the manufacturer to keep up a congruent brand concept or
image with its high-end price/quality product positioning. From a brand
management perspective, a less intensive distribution could also help create or
sustain a "snob-appeal", because the product is not easily available but customers
46
have seek out the sales outlet for the product (Park et al. 1988). Consequently, the
first hypothesis can be stated as:
Hypotheses 1:
A product position at the high end of the price/quality spectrum
will exhibit low distribution intensity while products positioned at
the low end will exhibit high distribution intensity.
3.1.2 Target Focus
The second construct influencing distribution intensity focuses on the
market segments the manufacturer intends to serve. A market segment is usually
defined as a group of consumers with homogeneous needs and/or purchasing
behavior (Doyle and Saunders 1985, Porter 1985). The derivation of market
segments is based on the assumption that the heterogeneous demand function for a
product class can be disaggregated into distinct groups of homogeneous demand
functions (Dickson and Ginter 1987). Target focus for the purpose of this research
deals with the question of whether a manufacturer pursues a broad scope target
segments or whether its focus is rather narrow. Two dimensions of targeting focus
are examined: (1) the number of unique market segments simultaneously served by
the firm with the same brand and (2) the number of prospective consumers or sales
potential in each segment served.
In their discussion on the economics of market segmentation Bonoma and
Shapiro (1984) argue that market segmentation is an expensive process and costs
47
rise very quickly when manufacturers serve additional unique segments. Products
that meet the needs of one segment might be largely inappropriate for the needs of
another. Marketing programs face the same dilemma. As a result, incremental
market strategy and implementation requirements of added business can strain
personnel, budgets, and dilute the brand image. The benefits of serving diverse
segments in terms of additional sales volume, scale economies, and marginal
profits must compensate for the additional costs. With this economic background to
target segment specificity, we can look at its impact on distribution intensity.
Targeting strategy is proposed to be more focused when fewer target
segments are served. Manufacturers pursuing a larger number of diverse segments
have to increase their presence in the market place since their customer segments
might differ substantially on needs. The manufacturer here has to use different
kinds of distributors to serve the different market segments. Even though all
consumers may be located in the same trade area, they would not necessarily
frequent similar outlets. Therefore, the spectrum of retail outlets necessary to hit
all segments is broader, which increases multiple representation in the same area
(Hlavacek and McCuistion 1983).
Targeting strategy is also proposed to be more focused when the number of
potential customers in the served segment is small. The case of a manufacturer
who serves more populous target segments is similar to the above situation. Even
48
though segments are usually characterized by a group of prospective customers
with similar needs, the manufacturer here has to deal with a larger variation in
shopping patterns or needs based on the larger number of customers in the
segment. Again, the firm needs to increase its multiple representation to completely
cover the segment. On the other hand, manufacturers with very specific and
narrowly focused target market segments may not have to increase their physical
presence if they are able to more precisely establish the shopping pattern of their
relevant segment or segments. As a result, the second hypotheses can be stated as
follows:
Hypotheses 2:
With increasing focus in a brand’s targeting strategy, the
intensity of distribution for the brand will decrease.
3.1.3 Need for Channel Control
Any decision on channel structure also includes the consideration of how to
manage the channel once it is established. Channels of distribution management
involves the adjustment of product availability, local promotion, final buyer price,
and quality maintenance. If independent middlemen are involved in the channel,
the issue of control arises. Since these middlemen are independent businesses, they
are subject to limited control of the manufacturer (Rosenbloom 1978). Referring to
that situation, Little (1970) points out that:
49
Because firms [comprising the marketing channel] are loosely
arranged, the advantages of central direction are in large measure
missing. The absence of single ownership, or close contractual
agreements, means that the benefits of a formal [superior-
subordinatejbase are not realized. The reward and penalty system is
not as precise and is less easily affected. Similarly, overall planning
for the entire system is uncoordinated and the perspective necessary
to maximize total system effort is diffused. Less recognition of
common goals by various member firms in the channel, as compared
to a formally structured organization is also probable.
Due to potentially divergent interests between manufacturer and
intermediary, control procedures must be in place to insure compliance with the
desired marketing mix (Bucklin 1973). Control is defined as the extent to which
one party actually influences the strategic and operating decisions of another party
in the channel (Skinner and Guiltinan 1985). Generally, the rationale for
manufacturers to control their channel stems from three sources.
First, inadequately trained middlemen could be detrimental to the perceived
brand concept especially in the case of high end product positioning. Bucklin
(1973) points out, that when left to their own devices, inexperienced middlemen
could end up making decisions which are not in the best interest of the
manufacturer or the intermediary itself. Such decisions could reduce or completely
eliminate the potential for success of manufacturers’ marketing efforts.
50
Second, heterogeneous decision making among dealers in the interest of
individual profit maximization could also pose detrimental effects to the coherent
brand concept (Bucklin 1973). A brand concept on the other hand calls for a
coordinated marketing effort which may require intermediaries to deviate from
their individually most profitable routes. Therefore, manufacturers with the need
for coordinated marketing strategy have a need for channel control.
Third, as mentioned in the strategic part of the model, intra-systemic
competition can have detrimental effects on the competitive product position
(Bucklin 1973). Based on the assumption that middleman interest in a brand is
directly influenced by the unit gross margin, intra-brand competition and its
decrease in gross margins can cause intermediaries to lose their willingness to
promote the brand or maintain quality. This lack of interest in turn could lead to a
deterioration in the competitive product position.
Cespedes (1988) expands on that point. He argues that an increasing
number of intermediaries will lead to a loss of supplier control over the actual flow
of its products in the channel and the way the product is presented to customers.
His argument is built on the premise of limited resources available to the
manufacturer. As the number of intermediaries increases, so do the opportunities
for transshipment, different levels of maintenance and repair services by various
intermediaries, different stocking levels, and overlapping sales efforts.
51
Additionally, the manufacturer’s ability to influence prices tends to decrease with
an increasing number of middlemen. A manufacturer would have to exert increased
efforts to keep control of all these aspects. Such increased efforts would take away
resources from core functions. In effect, the question becomes a question of where
to allocate resources. With an increased distribution network and potentially higher
sales resources usually go into production.
A combination of Bucklin’s (1973) and Cespedes’ (1988) line of argument
leads to the conclusion that firms with a high need for control over the marketing
mix in the channel of distribution tend to use highly selective to exclusive
distribution. Firms with a low need for control over their marketing mix will tend
to use intensive distribution. As a result, hypothesis number three can be stated as:
Hypothesis 3:
Manufacturers’ need for channel control is inversely related to
the level of distribution intensity.
In addition to the directly influencing distribution intensity, need for channel
control itself is proposed to be a function of product positioning strategy and target
segment specificity. Both constructs are proposed to influence the manufacturer
need for control in his pursuit of a congruent and consistent brand concept.
52
3.1.4 Product Positioning and Need for Control
For manufacturers following a cost-leadership strategy and "value”-based
product positioning, the need for control over the marketing mix and the
distribution channel may be less than for manufacturers following a differentiation
strategy and "prestige" product positioning.
In the case of cost-leadership and low price/quality product positioning,
price is a major decision factor for potential consumers. In the context of brand-
concept management, the message to be communicated by the supplier would be
value for the money. Suppliers are not dependent on value adding services through
their dealer network. The manufacturer is mainly concerned with bringing the
product to market as efficiently as possible. Usually such efficiency issue are in the
domain of the supplier. Therefore, the manufacturer does not have the need to
influence its dealers, instead the manufacturer can rely on competitive market
forces as they keep retail prices low.
"Prestige" brands being marketed through differentiation strategies are
located at the higher end of the price-quality positioning dimension. It is more
likely that their competitive positioning is based on image, product performance,
and service features rather than on price. Products sold through differentiation
strategies usually carry higher profit margins since they are geared to less price-
sensitive consumers. High-end products usually require higher levels of dealer
53
expertise, service quality or higher overall dealer image since these are the areas of
consumer sensitivity. Therefore, manufacturers rely more heavily on the value
adding activities of dealers. Suppliers have to establish coordinated and congruent
marketing in regard to performance and service features to communicate a
congruent brand image and attract potential customers. Consequently, it is
important for the manufacturer to retain marketing control within the channel of
distribution in order to achieve and assure brand consistency in terms of dealer
support, service expertise and product performance. The influence of product
positioning on manufacturer need for control can be summarized in a hypothesis as
follows:
Hypothesis 4:
Product positioning in regard to the price/quality spectrum is
positively related to manufacturer need for control in the
channel.
3.1.5 Target Focus and Need for Control
Manufacturers dealing with specific target segments have to be aware of a
different danger. Their focus must be on a close match between retailer
characteristics and customer requirements based on the firm’s marketing strategy.
The trap for them is the failure to enlist close-to-perfeet retailer support. If a
manufacturer with a specific target segment loses part of the segment due to
suboptimal dealer support, it hurts his business more severely than if a
manufacturer with many or broader target segments loses parts of one segment or
54
an entire segment. Therefore, specific target segments will require more selective
practices in choosing retailers which in turn will lead to less intensive distribution
levels.
Manufacturers with specific and narrow target segments are more likely to
follow focus strategies as defined by Porter (1980). Under focus strategies, a
manufacturer offers products that fulfill the needs of a particular buyer segment in
an industry and follows either a cost-leadership or differentiation pattern. The main
difference between cost-leadership and differentiation strategies is their narrow
market scope, centering on industry niches as opposed to a broad, industry-wide
scope. In regard to my model, manufacturers following a focus strategy would
offer only specialized products compared to other manufacturers, who offer a full-
line of products for their brands and use either cost-leadership or differentiation
strategies. Focus strategies are an excellent example for narrow target segments
due to a limited product range offered by the manufacturer. In that case, keeping
these selected segments by means of qualitatively superior dealer support is
important for the focus strategy to succeed.
Another line of reasoning again draws from Rueckert et al.’s (1985)
framework, especially the structural component of specialization. Dealers’
specialization can enhance adaptiveness of the whole distribution channel when the
tasks performed are more non-routine and the task environment is somewhat
55
complex or unstable. Manufacturers following a focus strategy and having highly
specific and narrow target segments must adapt to changing customer needs and
demands to survive in a market niche.
Consequently, for manufacturers with a narrow scope in their target
segments the need for marketing control in the channel would be high. Specific
market segments need to be catered to in a more controlled and coordinated
fashion, since the partial loss of those segments is more serious than for broad
target segments. Given the proposed inverse correlation between need for control
and distribution intensity, the ultimate consequence would be low distribution
intensity. The fewer the number of unique segments served by a manufacturer and
its brand and/or the smaller the number of prospective customers for those target
segments, the lower the distribution intensity the manufacturer will use.
Summarizing this point, the following hypothesis is developed:
Hypothesis 5:
With increasing focus in a brand’s targeting strategy, the
manufacturer need for channel control rises.
After having developed the hypotheses for the manufacturer strategy aspect
of the framework, the following section focuses on the second aspect of the model:
implementation.
56
3.2 The Intermediary Component of Distribution Intensity
The previous section focused on manufacturer considerations in regard to
distribution intensity. The following section looks at the implementation aspect of
any such intensity considerations and focuses on intermediary issues. In order to
explain the importance of integrating middlemen and their motivation to join a
distribution network into the conceptual framework, a comparison between the two
extremes of fully integrated versus non-integrated distribution is helpful.
If a manufacturer relies on a fully integrated distribution network, the
question of distribution intensity becomes a question of financial resources. The
choice of distribution intensity is driven by the product positioning and the
availability of financial means necessary to establish the appropriate number of
outlets per trade area. In this case, in which the manufacturer has complete control
over the distribution channel, implementation is a question of developing and
dispersing financial resources to open as many stores per trade area as intended.
If, on the other hand, the manufacturer distributes his products through non
integrated channels, the availability and willingness of middlemen to join the
distribution network becomes one of the driving forces of distribution intensity.
Not only is it likely to influence the successful implementation of any structural
channel decision, it may also influence the original choice by the manufacturer.
57
Several authors model the implementation of distribution intensity strategy
as a multi-step process (Rosenbloom 1991, Shipley 1984, Hlavacek and McCuiston
1983, Pegram 1965). Steps included in that process deal with finding prospective
channel members, applying selection criteria to determine the suitability of
prospective channel members, and securing the prospective channel members as
actual channel members (Rosenbloom 1991).
In the case of intensive distribution strategy, the selection process is less
important, since manufacturers usually place their products in every logical outlet.
The focus here is to blanket the market and make the product universally available
which leads to less discrimination in the selection of retailers other than maybe
satisfactory credit history (Pegram 1965). For selective to exclusive intensity
strategies the selection process is highly important, since it represents the juncture
at which the manufacturer has the greatest opportunity for channel control over
how the product is marketed in the channel, thereby establishing the foundation for
successfully marketing his product (Pegram 1965). This paper focuses on variation
in distribution intensity for selectively distributed products and, therefore, deals in
detail with issues connected to the selection process.
The three steps listed above point to two underlying dimensions of
implementing distribution intensity strategy. Finding prospective channel members
and determining their suitability for a specific distribution network deals with the
58
underlying dimension of intermediary availability. Securing the prospective channel
members as actual members is an issue of motivating those middlemen to
participate in the manufacturer designed channel. In the context of Hi-Fi speakers,
the manufacturers generally deal with the same set of intermediaries across trade
areas. Consequently, the implementation aspect of my conceptual framework for
distribution intensity focusses on intermediary motivation to join a distribution
network at the manufacturer-designed intensity level.
The motivation to participate in the distribution network rests mainly on the
financial attractiveness of such an affiliation and the barriers to entry these
middlemen face when considering participation. Barriers to entry for the purpose
of this study are conceptualized as (1) the level of investment necessary by the
middleman to be able to carry the product, (2) managerial support by the
manufacturer, (3) relationship uncertainty between middleman and manufacturer,
and (4) contractual restrictiveness faced by the intermediary.
While financial attractiveness has a direct bearing on the availability of
intermediaries, entry barriers are proposed to be mediated by financial
attractiveness. In essence, the model states that barriers of entry can be
compensated for if financial attractiveness is at a high enough level to overcome
initial obstacles. The following three sections examine the two motivational factors
and their interaction.
59
3.2.1 Financial Attractiveness
One of the main forces driving a middleman’s motivation to participate in
any given distribution network is the financial attractiveness of being affiliated with
the network. Financial attractiveness refers to the economic gains an intermediary
can expect from selling the manufacturer’s product. The larger the financial gains
an average individual intermediary can expect, the higher the probability for a
certain dealer to be motivated to carry that brand. On an accumulated basis that
means a larger number of dealers will be interested in carrying the particular
brand.
The rationale behind the relationship between financial attractiveness and
intermediary availability is based on micro-economic theory. The larger the
financial incentive is for distributing a certain product, the harder it is for the
intermediary to find alternative activities or investments yielding comparable
economic gains. With the number of alternatives reduced, the number of
middlemen deciding to carry the product given the level of financial gain increases.
This trend might continue until the distribution intensity in a given trade area is
high enough to either lead the manufacturer to reduce the financial incentives for
middlemen or to lead to intra-brand competition in the distribution channel with the
same effect. The effect in both cases is a stabilization of dealers joining in the
distribution of the product, since other economic alternatives might become viable.
60
Financial attractiveness is usually conceptualized as a combination of gross
margins and sales potential (Corey et al. 1989), with higher gross margins and
higher sales potential leading to higher financial attractiveness. In the case of
uneven levels, higher sales potential compensates for lower margins and vice
versa.
Gross margin as one of the aspects of financial attractiveness influences
intermediary willingness to join a distribution channel. In Bucklin’s (1973) "Theory
of Channel Control," the author proposes a model of channel control based on the
two constructs of (1) tolerance and (2) payoff function. He relates both functions to
the profits earned by intermediaries and how that relationship influences supplier
authority. Within that framework, Bucklin posits gross margin as a major factor
influencing middlemen’s willingness to join a distribution network. Higher gross
margins are proposed to attract more middlemen to a distribution channel.
The other aspect of financial attractiveness and its influence on intermediary
participation is the potential sales volume in a trade area. Sales volume is proposed
to be positively related to financial attractiveness. With higher sales volume
potential for the brand in question, more intermediaries are willing to join the
channel network (Corey et al. 1989). A large projected sales volume for the brand
means a larger sales pie to be divided among retailers. Financial attractiveness
rises because retailers could benefit from economies of scale in their selling effort,
61
and their risk of failing is lower than if they carry brands with only limited sales
potential.
In summary, increasing the financial attractiveness of carrying the products
of one manufacturer will increase the number of intermediaries motivated to join
his distribution network, and will, therefore, increase distribution intensity for the
product in a given trade area. As a result, the two aspects of financial
attractiveness can be integrated to develop the following hypothesis:
Hypothesis 6:
Financial attractiveness of a manufacturer’s brand is positively
related to distribution intensity.
3.2.2 Barriers to Entry
The second factor having a main influence on middleman motivation to
participate in any given distribution network are the barriers to entry affiliated with
the distribution network in question. Barriers to entry are defined as the situational
circumstances and requirements which prevent a prospective intermediary from
carrying the product or even from establishing a business relationship with the
manufacturer. It is proposed that high barriers to middleman entry will lower
middleman motivation to join a distribution channel and vice versa.
62
Middlemen who are unable to overcome initial barriers to entry will look
for alternative business opportunities and drop from the pool of participants in the
channel. Overcoming initial barriers to entry would mean a higher effort and
commitment on the part of the intermediary, which increases his risk for the
business venture (Porter 1980). Higher barriers to middleman entry will cause a
larger number of middlemen to look for alternative forms of business
opportunities.
Entry barriers are conceptualized by four constructs: (1) level of
investment, (2) relationship uncertainty, (3) managerial support, and (4) contractual
restrictiveness. All four constructs have been dealt with in existing literature and
are proposed to impact intermediary motivation.
3.2.2.1 Level of Investment
Level o f investment is defined as the capital necessary to establish the
capability to retail the specific product. Increasing investment levels are proposed
to increase barriers to entry. This paper focuses on two dimensions of investment:
(1) capital expenditure in human factors like expertise or training and (2)
investment in inventory and/or service facilities.
Corey et al. (1989) suggest that level of investment is directly related to the
willingness of intermediaries to join a given distribution channel. Investments are
63
necessary to meet the required level of expertise necessary to sell the product, as
well as the level of inventory required to serve the sales potential in the trade area.
The level of reseller investments necessary to service the product line is inversely
related to middlemen’s willingness to join a channel. Larger investments for a
given product or brand constitute a higher barrier to entry, because the middleman
must have larger total profits to arrive at the same return on investment as an
intermediary with lower levels of investment. The need for larger profits increases
the level of business risk for the intermediary, which in turn drives up the effective
barrier to his entry and reduces his motivation to join the channel.
The restrictions of limited capital availability in regard to investments in
inventory and service facilities are straight-forward. But dealer expertise, which
requires intermediaries to invest in training, adds another twist to the capital
restrictions. In addition to the need for capital, training also takes time and effort.
The fact that expertise must be acquired over time if it is not already available also
represents a barrier to entry. Dealers must believe they will gain a competitive
edge, if they go through the effort to acquire expertise. This, in effect, is a barrier
to entry that decreases the availability of interested intermediaries and tends to
lower the distribution intensity for that brand. As a result, higher levels of
expertise required of the dealer mean higher investment in training people,
showroom interior, laboratory, repair facility, etc., which in turn leads to fewer
dealers being interested in making themselves available to join the distribution
64
channel for that brand. Under the assumption that alternative routes to make money
are available, dealers first will seek easier routes to profit.
In connection with the level of investment, the risk of adequate returns on
investment must be mentioned. Corey et al. (1989) give the example of the early
stage in product life-cycle, where the intermediaries run a higher risk in securing
an adequate rate of return on their investments because of an unproven product.
But risk also comes in other shapes and forms. Another example for intermediary
risk includes the representation of small or unknown brands with low consumer
recognition and financially weak manufacturers. In this case, the retailer again runs
a higher risk of securing an adequate return on investment. Higher risks are
associated with lower motivation for joining the channel and must be compensated
for through the mediating influence of financial attractiveness. One option for
compensating the dealer for the higher risk is the profit margin achieved by
restricting intra-channel price competition, which usually means less intensive
distribution per trade area.
In summary, the relationship between investment level and barriers to entry
is proposed to be positive, with higher investment requirements increasing the
barriers to entry which in turn leads to lower intermediary motivation to join the
channel of distribution.
65
Hypothesis 7a:
The required level of investment on the part of the retailer will
be inversely related to distribution intensity.
3.2.2.2. Relationship Uncertainty
Relationship uncertainty for the purpose of this research is defined as the
intermediary perception of the likelihood that the relationship with the
manufacturer would continue on a long-term basis. It is proposed to be positively
related to barriers to entry, with higher uncertainty leading to a higher barrier to
entry.
Empirical research has shown that the assurance of a continual business
relationship with the manufacturer can be an important means to motivate
middlemen (Sibley and Teas 1979). This need for a longterm relationship assurance
is most likely based on their need for an ongoing stream of income and business
survival (Shipley 1984).
Anderson and Weitz (1989) developed a model on determinants of
continuity in a conventional industrial channel dyad. Even though their model
focuses on the manufacturer aspect of the issue, some of the advantages derived
from a high continuity expectation on the part of the intermediary can be applied to
my model. Anderson and Weitz point out a higher likelihood of cooperative
behavior by sales agents if these agents can expect to be around for the payoff of
66
their upfront investments in such cooperative action. This point is in line with
Shipley’s (1984) point of view that motivation is driven by need for financial
payback and survival. Most important, though, is Anderson and W eitz’ assertion
that switching costs between manufacturer and retailer are rising in markets in
which concentration tendencies can be found. Therefore, both partners have a
mutual interest in long-term stability of the relationship. If such long-term stability
cannot be expected, the barriers to entering into such a relationship are rising.
Another stream of reasoning for rising barriers to entry with increasing
uncertainty is based on economic theory. The higher the uncertainty of continuity
the higher the probability of no return on investment for the intermediary, which in
turn drives up his opportunity costs. With the overall return decreasing, his
barriers to entry are rising since more business alternatives become viable for the
middleman and the number of intermediaries who have no better economic
alternatives to choose from is decreasing. As a result, the following hypothesis can
be stated:
Hypothesis 7b:
Uncertainty about the future prospects of a business relationship
will be inversely related to distribution intensity.
67
3.2.2.3 Managerial Support
Managerial support is defined as the actions taken by the manufacturer to
reduce middleman problems in day-to-day operations. Aspects of middleman
support could include promotional material, telephone help-lines, accounting,
credit, and billing assistance and other actions geared towards relieving the
intermediary of operational pressure.
Studies among intermediaries have shown that intermediaries can be
motivated by manufacturer’s willingness and ability to provide adequate support to
middlemen (Shipley 1984, Bobrow 1976, Sibley and Teas 1979, Webster 1975).
These support services reduce business risk and, therefore, are likely to decrease
perceived barriers to entering the channel of distribution. In addition, Rosenbloom
(1978) argues that actions like these increase middleman motivation by showing
manufacturer commitment to the intermediary. Such commitment by the
manufacturer decreases the middleman’s feeling of risk which could help lowering
the perceived barriers to entry. The hypothesis following such rationale is:
Hypothesis 7c:
The level of manufacturer sponsored management support will
be positively related to distribution intensity.
68
3.2.2.4 Contractual Restrictiveness
Contractual restrictiveness refers to the degree to which specifications in the
formal agreement between manufacturer and intermediary restrict the dealer’s
freedom of managerial choice. The present framework proposes that contractual
restrictions are raising the barriers to entry which in turn is detrimental to the
willingness of independent intermediaries to join a particular distribution network.
The more restrictions a manufacturer stipulates the fewer middlemen will be
interested in joining the network. In a product market the intermediary might have
different goals from the manufacturer and, therefore, is apprehensive of any
restrictions on his managerial freedom. The interrelationship between contractual
limitations and financial attractiveness can be seen in Bucklin’s (1973) "Theory of
Channel Control." With an increase in profits or payoff distributed to the
middleman, his tolerance of manufacturer control and restrictions increases.
Contractual restrictions can come in various forms like the number of
requirements the dealer has to meet to fulfill manufacturer’s expectations; the
number of areas in which the manufacturer stipulates restrictions (accounting,
sales, display and store appearance, etc.); strictness of termination clauses and
others. The more restrictive the relationship between distributor and manufacturer
is, the higher the effective barriers to entry. This leads to fewer intermediaries
being interested in joining a distribution network. As a result of the aspects of
contractual restrictions, the following hypothesis can be developed:
69
Hypothesis 7d:
Contractual restrictiveness will be inversely related to
distribution intensity.
3.2.3 Moderation Effect of Financial Attractiveness
The impact of barriers to entry may be moderated by financial
attractiveness. A high level of financial incentives for carrying a brand may tempt
a middleman to stretch and to try overcoming entry barriers beyond his original
abilities, simply due to the probability of high returns. Support institutions, like
banks, might raise their support ceilings based on the same rationale and a
probability of high returns. In effect, they help the intermediary to lower initial
barriers of entry which leads to the following hypothesis:
Hypothesis 8:
High financial attractiveness of the brand will moderate the
influence of entry barriers on distribution intensity.
3.3. Summary
The conceptual framework of distribution intensity depicts an interactive
two-step process. In a first step, the manufacturer decision on intensity has to
reflect product positioning, target segment and control considerations. In a second
step, intermediary motivation to participate in the given channel network has to be
considered. This two component model has the advantage of explicitly integrating
aspects of both partners in a distribution relationship.
70
From a manufacturer’s point of view it could be argued that a higher
availability of capable and willing intermediaries in a trade area does nothing else
but give the manufacturer an opportunity to fully implement its original plan in
regard to the number of retail outlets used. On the other hand, a low level of
middleman availability causes the manufacturer to adjust its goals for market
representation downward. This argument underestimates the importance of
intermediaries as a market force. The phenomenon of grey markets with
unauthorized dealers is a case in point.
The pressure on intermediaries to be competitive in a highly developed
markets or highly attractive financial terms for a brand can lead to the development
of unauthorized distribution channels. In these cases, authorized dealers have to
compete with unauthorized dealers, who get their product from sources other than
the manufacturer. Consequently, brand representation or distribution intensity will
be higher than originally designed by the manufacturer.
An example for such a situation was the distribution of IBM PC’s in their
prime time. In addition to authorized dealers, many "grey market" dealers popped
up and by competing on price instead of service or quality, eroded profit margins
in the distribution channel, effectively forcing IBM to revamp its dealer network.
This example shows that intensity structure has to be a reaction to the market and
competitive structure, and the manufacturer has to have the resources or ability to
71
entice the appropriate kind and number of intermediaries to join his distribution
network.
Only if all manufacturer and intermediary aspects are properly integrated
into a coherent strategy for distribution intensity will the actual distribution
intensity in the market place come close to being optimally effective in supporting
all other marketing variables.
72
Chapter 4
Research Methodology
The primary objective of this study is to develop a better understanding of
distribution intensity from a channel perspective. In pursuit of this objective the
developed framework examines two components of distribution intensity: (1) a
manufacturer component based on product positioning, targeting and need for
channel control, and (2) an intermediary component based on financial
attractiveness and barriers of entry. This section describes the proposed research
method for testing the conceptual framework empirically.
4.1 Research Setting
Prior to developing the empirical test design, three issues must be
addressed: (1) the market setting for the test, (2) the channel level at which
intensity is observed, and (3) the product line for which the model is tested.
The first major issue is whether the model should be tested in an industrial
goods or consumer goods setting. The latter setting is chosen because we expect to
encounter a wider variation of distribution intensity levels in the consumer goods
73
market. There are many more potential target segments in the consumer goods
market than there are target segments in the industrial goods market. Matching the
variety of consumer segments distribution is similarly varied. A wider spread of
intensity levels across brands in a trade area will allow us to explore some of the
intricacies of intensity strategies developed by manufacturers and provide richer
insights into the phenomenon thereby potentially increasing the explanatory power
of my model. While it is expected that the model is generalizable across both
markets, consumer goods markets provide the better setting to test the model.
Second, what channel level should be analyzed? Should this study focus on
the retail level only or should it include all existing levels of channel structure. As
pointed out before, the intensity level observed by the end-consumer should be the
relevant level for this study. Considering previous conceptualizations of distribution
intensity, the logical basis for the need of different intensity levels has been the
consumer shopping patterns (Copeland 1923, Aspinwall 1956, Bucklin 1963).
Consumer shopping usually takes place at the retail level. Therefore, the empirical
analysis of the present study will focus on the retail level and the number of retail
outlets within a trade area.
The product line chosen for this study is Hi-Fi speakers in the consumer
electronics industry. Two primary reasons exist for this decision. First, distribution
for that product form is predominantly handled by a network of independent
74
retailers, even though some exceptions exist7. Second, the speaker industry is
basically a mature market with low industry dynamics. Speaker technology has not
changed much for the past 20 years. Product positioning by manufacturers and the
established distribution networks have been quite stable. Consequently the
established distribution networks can be presumed to be close to equilibrium and
the data collected should be shielded from noise based on environmental dynamics.
4.2. Unit of Analysis
The unit of analysis for this study is the individual Hi-Fi speaker brand.
Observing individual brands instead of manufacturers is preferred for two reasons.
First, some manufacturers established multiple brands in the market place with
each brand positioned independently. In addition, manufacturers go through
considerable effort to establish individual brand name recognition among
consumers. Therefore, from a manufacturers point of view individual brand focus
is necessary. Secondly, consumers usually refer to brand names not manufacturers
when considering a speaker purchase. Consumers of Hi-Fi electronics shop for
products by Fisher or Panasonic, for example. Only rarely are they aware that both
brands are made by the same manufacturer.
7 Tandy Corporation distributes their brands o f electronic equipment including speakers through
company owned or franchised "Radio Shack" stores. Sears and some other department store
companies distribute their store brands like "LXI" exclusively through their department store
outlets.
75
Manufacturers use multiple brands to appeal to different market segments
without running the risk of diluting brand focus. Being positioned differently and
usually independently, these brands might be required to be marketed through
different distribution networks with various levels of distribution intensity.
4.3. Sampling Method
The framework presented in Chapter Three looks at manufacturer as well as
the intermediary aspects of distribution intensity. An empirical study based on the
framework ideally would include both sides in the sample. Therefore, this study
originally planned to integrate both aspects and provide for a full perspective by
drawing a manufacturer as well as dealer sample. Manufacturers should be
included because they are assumed to initially drive the process of distribution
strategy by way of their decisions in regard to product characteristics, targeting
strategy, pricing, and selection of distribution channels. The inclusion of retailers
is based on the assumption that their willingness to participate in the manufacturer
intended channel set-up alters the actual form of the channel.
However, in initial interviews and a pre-test it became clear that gathering
data from manufacturers only was appropriate. Manufacturers seem to have a much
clearer picture of general distribution issues than retailers. With the exception of
national retail chains, dealers appear to be less knowledgeable about distribution
issues concerning the Hi-Fi speaker brands they carry and the general competitive
76
or distribution picture in the speaker industry. In addition, the researcher
encountered an absolute unwillingness of some retailers to share any such
information if it was available. While manufacturers are willing to share sensitive
information once they are convinced of the potential value this study can add to our
knowledge of distribution, retailers are very protective of any information which
could be construed as a proprietary. The Pre-test showed an unacceptably low
percentage of dealer participation. Out of 65 inquiries only 10 questionnaires were
returned (15% return rate) and only 5 (8% useful returns) had enough information
to be analyzed. Generally, the information that was provided proved to be
incomplete and several respondents commented on the questionnaires that they
simply could not answer our questions due to lack of information.
During initial in-depth interviews with manufacturers, it became clear that
most firms seemed to have a good level of knowledge on what is happening on the
retail level. Therefore, the present study tries to generate all pertinent information
from manufacturers.
The number of Hi-Fi speaker brands marketed domestically via independent
retail outlets is limited and the cooperation of suppliers in this small industrial
community was a major concern for the research. Initial interviews indicated that
gaining manufacturer cooperation for this study would be a challenge. Given the
fact that this study relies on suppliers, extensive efforts were made to ensure
77
participation in our research. In the process of developing the sample and
preparing data collection, a three step approach was used.
1. Step - Sample Preparation:
Due to the fact that the number of suppliers is small, the study tried to
generate a complete population sample. The electronics industry has an industry
association which assists members and other interested parties in gathering
information about the industry. Initial contacts to the association showed that a list
of Hi-Fi speaker manufacturers did not exist but had to be generated from existing
membership files. In addition to the costs for such service, the list would have
been incomplete because not all manufacturer’s are members of the association.
Therefore, the researcher decided to obtain a conference catalogue of the
Consumer Electronics Show (CES), which is the prime show case and trade
meeting for the industry. This list was used to generate the sample of
manufacturers for the study. The CES is designed to introduce products and
manufacturers to the marketplace and is open to the trade only8. As a result,
manufacturers who are utilizing retail networks are present or represented by their
distributors. This makes the CES generated manufacturer list an almost perfect fit
for this study. Missing manufacturers were added by going through trade
8 Starting with the latest CES in May 1992 in Chicago, the industry opted to allow public access
for one weekend.
78
magazines and looking for advertisements or articles as well as by talking to
retailers.
2. Step - Participation Commitment:
Pretests showed initial manufacturer resistance to answer questions in
regard to distribution issues when the researcher introduced himself and the study
via telephone. The reason for resistance was rooted in the fear that this study is a
competitive inquiry or sponsored by competitors. Therefore, we deemed it
mandatory to visit as many manufacturers as possible in order to introduce the
study on a personal basis. In one on one talks facing each other it is much easier
to alleviate problems of distrust and installing confidence in the sole academic
purpose of this research.
In order to meet as many firms as possible, the researcher attended CES in
January 1992 in Las Vegas, Nevada. Due to time limitations (the conference lasts
only four days), and the large number of firms represented (more than 200), not all
speaker manufacturers could be contacted. Altogether, about 40 interviews were
conducted. In these interviews the researcher introduced himself and the project.
The manufacturers were made aware of the importance of their participation and
assured of complete confidentiality regarding their responses. In all cases, a verbal
commitment to respond to the forthcoming questionnaire was pursued. The
79
majority of contacts made led to a positive manufacturer attitude towards the
research project and a commitment to participate.
3. Step - Identification of Individual Respondents:
During the personal interviews at the conference suitable individuals who
are most familiar with marketing and distribution issues were identified for each
firm. If the interviewed manufacturer representative was not the person who would
respond to the forthcoming questionnaire, the name, title and address of the
potential respondent was gathered. In some of these cases, the interviewee agreed
to forward the questionnaire and enforce response. For firms not covered by
interviews, individuals were identified by choosing the marketing directors or
marketing vice presidents listed in the conference catalogue or through subsequent
telephone conversations with firms.
4.4. Data Collection Method
The study uses primary data collection via questionnaires as its tool to
generate information for analysis. The questionnaire developed had three versions.
One version focused on home speakers, while the other versions were tailored for
automotive or other speakers respectively. Except for differences in wording to
anchor responses to speaker types, the home-/car-speaker questionnaires were
identical. The third questionnaire version for other speakers also differed in its
order of two blocks of questions. The change in ordering was used to facilitate the
80
respondent’s decision to focus on one type of speaker in his/her responses to the
questions.
The questionnaires were completed by a key respondent most closely
associated with marketing strategy or distribution channel issues. The use of key
informants has been criticized in the marketing literature (Phillips, 1981), but the
specific nature of solicited responses justified the use of this approach. The
information necessary for our analysis relates to the narrow range of the firm ’s
activities and also requires some knowledge of the competitive environment.
Therefore, we needed to find the individual who not only knows the marketing and
distribution strategy for the firm but also how it compares to its competitors and is
likely to provide unbiased in-depth data necessary for our analysis (John and Reve
1982). The personal interviews at CES as well as by telephone helped identify the
key informants.
To ensure the highest possible return rate for our inquiry, the study used a
four step approach for soliciting responses. First, the identified key respondents
received an announcement letter introducing the study, its potential value to
academic research and the importance of firm participation. The introductory
letter announced the arrival of the questionnaire which was scheduled for
approximately five days after receiving the first letter. As a second step, the first
wave of our questionnaire was send out. The questionnaire was accompanied by a
81
cover-letter urging respondents to participate and reminding them of the value the
study will add to academic as well as professional knowledge. In addition to that,
we included a pre-paid and addressed return envelope. The third step included
calling firms who were expected but did not respond to the mailing within two
weeks in order to inquire about problems and trying to solicit their participation. In
the fourth and final step a second wave of questionnaires was sent out. The second
questionnaire was accompanied by a cover-letter and pre-paid business return
envelope. In addition to that, we included a brochure, which introduced the new
distribution management program at the university. The goal was to show the
immediate contributions this study makes to our academic institution and
underscore the study’s importance. In all four phases, we assured participating
firms of keeping data confidential and invited their questions by listing telephone
numbers of the participating researchers. Companies did make use of the numbers
and ninety percent of all fielded phone calls led to subsequent participation in the
survey.
4.5 Measurement Instrument
The measurement instrument was developed in three stages. A review of
existing literature was used to specify the domain of constructs and to evaluate
existing scales for the framework’s constructs. In addition to previously used
measures, alternative scales were developed. In a second stage, interviews with
retail store managers and manufacturer’s representatives were used to evaluate the
82
content validity of the measures and check wording in the preliminary
questionnaire. The last step was a small scale pre-test with data collected from
approximately 65 respondents on the retail level. A pre-test on the manufacturer
level was not feasible due to the small sample population. The pre-test was used
to identify potential problems in scales and data collection procedure. Poorly
worded or overlapping measures were identified and a response rate estimate was
obtained. The goal of this exercise was to refine the measurement instrument and
develop strategies to overcome data collection problems. As indicated before, the
extremely low response rate for usable questionnaires of 8 percent (5 out of 65
questionnaires) caused us to switch to manufacturers as the source for information
and led to the three-step approach of data collection.
4.5.1 Operationalization of Dependent Variable:
Distribution Intensity
Even though this is the first empirical study testing a conceptual framework
for distribution intensity, a number of previous studies have conceptualized the
construct of distribution intensity as the number of retail outlets carrying a brand in
a given trade area (Hartung and Fisher 1965, Cespedes 1989). Usually, this
conceptualization includes a comparison between the firm ’s number of retail outlets
and the total number of possible retail outlets in the trade area or a direct
comparison to competitors. In addition to the number of outlets used, one could
conceptualize distribution intensity as the level of selectivity a firm displays when
83
signing up dealers (Bagozzi 1986, Hardy and Magrath 1988). Manufacturers being
highly selective in their choice of retailers are associated with low distribution
intensity while firms which are not selective are associated with highly intensive
distribution.
Based on pre-test results, this study uses both conceptualizations for the
construct of distribution intensity. Table 4.1 presents the items used to measure
distribution intensity. Items DI2 and DI6 are comparisons between the number of
outlets carrying the brand versus its competition while items D ll and DI3-DI5
inquire about the policy in signing up dealers. For the present study the
comparison of the firm’s number of retail outlets per trade area versus the total
number of retail outlets per trade area was disregarded. Since responses are
averaged over the entire domestic market it is not feasible to arrive at a meaningful
total number of retail outlets per trade area. The items used to measure the
construct include attitude statements on perceived distribution intensity as well as
estimates on numbers of retailers carrying the brand in an average trade area.
Attitude statements are on a five-point scale anchored at the low end by strongly
disagree (1) and at the high end by strongly agree (5).
84
Table 4.1
Items Used to Measure Distribution Intensity
I t e m S c a l e
Attitude Measures: Number of Outlets
D ll We try to keep the number o f retailer
carrying our brand in each trade area to a
minimum.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
DI2 In comparison to all other speaker brands
in the market, our brand tends to be
available through more retail outlets per
trade area.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
Attitude Measures: Selection Strategy
DI3 We require retailers to meet strict
manufacturer imposed standards before we
let them sell our brand.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
DI4 W e are highly selective in our choice of
retailers who carry our brand in each trade
area.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
DI5 Only a select few retail outlets per trade
area are allowed to carry our brand.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
Direct Measures: Distribution Intensity
DI6 Compared to your competition, how would
you describe your brand’s distribution
pattern for each trade area?
Exclusive 1 2 3 4 5 Intensive
4.5.2 Operationalization of Independent Variables:
Manufacturer Aspects
The product positioning construct was defined on the price/quality
dimension. The operationalization of the construct included items for both
dimensions. Attitude statements included two items on brand image and
performance (PP1 and PP2) and one item on retailer positioning (PP3). The item
on retailer positioning was included to reflect the need for congruence between
85
product and distribution in positioning a brand. Direct measures included a direct
question in regard to product positioning between high and low end as well as
questions on product price, performance, quality, and prestige. Table 4.2 lists the
items.
Target focus was originally operationalized in two dimensions of (1)
number of target segments and (2) sales potential per segment. The pre-test showed
that the second dimension was confusing since manufacturers have already
difficulty determining the number of target segments they pursue. Therefore, the
items included in the questionnaire stayed clear of asking for sales potential per
target segment. All items were attitude statements on a 5-point scale anchored by
strongly disagree or strongly agree. Table 4.3 shows a list of included items.
The construct of need for channel control was split into two sets of
questions. In the first set we asked for attitude statements in regard to the
similarity of dealer efforts in selling the brand (CC1 and CC2). This was basically
an indirect way of asking about the latent need for control in the distribution
channel. In the second set of questions we asked directly about the level of
influence a manufacturer tries to exert on its intermediaries. See Table 4.4 for all
items included in the questionnaire.
86
Table 4.2
Items Used to Measure Product Positioning
I t e m S c a l e
Attitude Measures
PP1 Our brand carries a very prestigious image
in the marketplace.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
PP2 Our brand is known for its high
performance.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
PP3 Our distribution structure emphasizes high
end retailers.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
How do you position your brand on the
following product characteristics?
PP4 Retail Price Low End 1 2 3 4 5 High End
PP5 Prestige/Image of Brand Low End 1 2 3 4 5 High End
PP6 Speaker Technology Low End 1 2 3 4 5 High End
PP7 Performance Low End 1 2 3 4 5 High End
PP8 Overall Product Quality Low End 1 2 3 4 5 High End
Comparison to Competition
PP9 In comparison to all other speaker brands
available in the domestic market, where is
your brand positioned?
At the At the
low end 1 2 3 4 5 high end
87
Table 4.3
Items Used to Measure Target Focus
I t em S c a l e
Attitude Statements: Brand
TF1 By design, our brand has a small number o f
potential customers.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
TF2 By design, our speakers appeal to a narrow
spectrum of consumers only.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
TF3 We use a niche strategy for marketing our
brand.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
TF4 The marketing strategy for our brand is not
focused on specific consumer segments.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
TF5 Our speaker brand can be best described as
a brand o f mass appeal.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
Attitude Statements: Retailers
TF6 Most o f our retail accounts could be
characterized as mass merchandising
retailers.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
88
Table 4.4
Items Used to Measure Need for Channel Control
I t e m S c a l e
Attitude Statements
CC1 If we would not coordinate the marketing
activities o f our retailers, they would vary a
great deal in how they sell our brand.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
CC2 We try hard to keep all retailer initiated
marketing activities similar across our retail
network.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
Please indicate the extent to which you as the
manufacturer directly or indirectly (i.e. through
your distributors) exert effort to coordinate the
following retailer activities. Effort to Influence
CC3 Implementation o f Sales Promotions Low 1 2 3 4 5 High
CC4 Use o f Promotional Aids (e.g. Brochures,
Posters)
Low 1 2 3 4 5 High
CC5 Use o f Product Displays (e.g. Sales Racks etc.) Low 1 2 3 4 5 High
CC6 Dealer Sales Presentations to Consumers Low 1 2 3 4 5 High
CC7 Store Appearance (e.g. Layout, Shelf-Design
etc.)
Low 1 2 3 4 5 High
CC8 Dealer Inventory Levels Low 1 2 3 4 5 High
4.5.3 Operationalization of Independent Variables:
Intermediary Aspects
Financial attractiveness was originally conceptualized in two dimensions of
(1) sales volume or potential and (2) profit margins for the brand. Items included
in the questionnaire reflected the conceptualization. Table 4.5 displays the items
used to measure financial attractiveness. Item FA1 asked for an attitude statement
89
in regard to the sales volume on the individual retailer level, while items FA2,
FA3 and FA6-FA8 look at the overall domestic sales potential for the brand. The
second dimension of profit margins was covered by items FA4 and FA5 which
inquire about the suggested and actual retail margins in comparison to competing
brands. Except for items FA7 and FA8 which directly asked for sales volume and
market share respectively, all other items are measured on a five-point scale.
Management support and contractual restrictiveness are both measured via
indices. The construct of management support is construed as an index consisting
of ten potential support activities. Table 4.6 lists all items for the index. The
items are measured on a six point scale. Activities not provided by the
manufacturer are scored as a zero while a scale between 1 (not important) and 5
(very important) reflects the importance of individual support activities for the
manufacturer dealer relationship. Contractual restrictiveness was conceptualized as
the number of functions being regulated in the contractual relationship between
channel partners. Table 4.7 lists the 7 items included in the measure. All items are
scored on a yes/no scale. For manufacturers who do not have a contract with their
dealers the index is scored as a zero.
Items measuring the investment level construct include three attitude
statements. They are displayed in Table 4.8. Two items measure time and money a
dealer has to invest in a brand to be able to represent the brand. The third item
90
measures the investment in inventory to be kept for that speaker brand. All items
are stated as comparisons between the manufacturer’s brand and the competition.
The last construct to measured is relationship uncertainty. Three attitude
statements were included and are displayed in Table 4.9. These items measure
manufacturer attitude towards keeping existing channel partners on a long-term
basis as well as manufacturer promptness in terminating a relationship.
91
Table 4.5
Items Used to Measure Financial Attractiveness
I t em S c a l e
Attitude Statement
FA1 Compared to all other speaker brands, our
brand tends to generate the largest sales
volume in speakers for retailers who carry
it.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
Relative to your competition, the sales
volume for your brand in the domestic
market...
FA2 calculated in US $ is ... Low 1 2 3 4 5 High
FA3 calculated in units is ... Low 1 2 3 4 5 High
Relative to all other home speaker brands
in the market, how does your brand compare
on the following issues?
FA4 The suggested retail margin for our brand
is ...
Lower 1 2 3 4 5 Higher
FA5 The actual retail margin for our brand is Lower 1 2 3 4 5 Higher
FA6 The sales potential for our brand is ... Lower 1 2 3 4 5 Higher
Direct Measures
FA7 What was the approximate sales volume (in
US $) of your brand in 1991? Please mark
the appropriate category.
less than 1 million US $
1 million - 5 million US $
5 million - 20 million US $
20 million - 50 million US $
50 million - 100 million US $
100 million - 200 million US $ ___
more than 200 million US $ ___
FA8 What is your brand’s estimated market share
in the loudspeaker market? Percent
92
Table 4.6
Items Used to Measure Management Support
Item Scale
MS Which of the following assistances do you
provide to your retailers and how important
is each in contributing to dealer satisfaction
in your judgement? Provided
Not
Important
Very
Important
a. Promotional Allowances Yes/No 1 2 3 4 5
b. Promotional Material Yes/No 1 2 3 4 5
c. Product Displays Yes/No 1 2 3 4 5
d. Inventory Management Yes/No 1 2 3 4 5
e. Accounting Support Yes/No 1 2 3 4 5
f. Inventory Financing Yes/No 1 2 3 4 5
g-
Dealer Hotline (800-Number) Yes/No 1 2 3 4 5
h. Consumer Hotline (800-Number) Yes/No 1 2 3 4 5
i. Over-the-Counter Warranty Policy Yes/No 1 2 3 4 5
j-
Other: Yes/No 1 2 3 4 5
Table 4.7
Items Used to Measure Contractual Restrictiveness
Item Scale
CR Which of the following issues are addressed in your
contractual agreement with retailers?
a. Product Displays for Brand Yes / No
b. Inventory Levels for Brand Yes / No
c. Brand Sales Goals for Store Yes / No
d. Brand Sales Promotions Yes / No
e. Product Pricing for Brand Yes / No
f. Quantity Price Discounts for Dealer Yes / No
g-
Contract Termination Yes / No
93
Table 4.8
Items Used to Measure Investment Level
It e m S c a l e
Relative to all other speaker brands in the
market, how does your brand compare on
brand specific dealers investments?
IL1 The amount o f money dealers have to spent
on training new sales people in order to
handle our brand is ...
Much Much
Lower 1 2 3 4 5 Higher
IL2 The amount o f time new sales people have to
spent on training in order to handle our brand
is ...
Much Much
Lower 1 2 3 4 5 Higher
IL3 The level o f inventory needed to be
adequately stocked for our brand is ...
Much Much
Lower 1 2 3 4 5 Higher
Table 4.9
Items Used to Measure Relationship Uncertainty
I t e m S c a l e
RU1 We consider our relationship with retailers
to be a longterm alliance.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
RU2 Retailers who carry our brand are expected
to be selling our products for a long time.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
RU3 If retailers don’t meet our expectations, we
terminate our business relationship with
them promptly.
Strongly Strongly
Disagree 1 2 3 4 5 Agree
94
Chapter 5
Results
This chapter is divided into two sections. The first section contains a
descriptive analysis of the sample and the measures used to test the framework and
its hypotheses. It includes means, standard deviations, frequency distributions and
correlations for the constructs. In addition to the descriptive analysis, it presents
results of construct reliability and validity tests. The second section deals with
testing the research hypotheses and presents the results. The study uses linear
regression models to analyze the framework. Due to the exploratory nature of this
research we forego the opportunity to employ advanced analysis techniques like
LISREL but instead focus on multiple regression models as the appropriate tools
for testing the hypotheses.
5.1 Descriptive Analysis of the Sample and Constructs
5.1.1 Sample
The total sample consisted of 219 Hi-Fi Speaker brands. Some
manufacturers market several brands in the United States. If these brands were
produced and marketed by semi-independent divisions of the firm, they were listed
95
and contacted individually. An example here is Harman/Kardon, which has
several brands in the market (Harman/Kardon, Pyle, EPI). For other multiple
brand firms with singular marketing functions, only one name was listed and one
contact made. In these cases, the pre-test showed the distribution structure
between brands to be very similar. A total of 85 usable questionnaires were
returned, which is a 38.8 percent return rate. Of the 85 brands analyzed, 58
(68%) dealt with home speakers, 22 (26%) dealt with automotive speakers, and 5
(6%) focused on specialty speakers. During the first wave, 68 (80%)
questionnaires were returned while the second wave yielded 17 (20%) responses.
An overview of the characteristics of the sample is presented in Table 5.1.
The average distribution intensity for the sample was 2.12 indicating a somewhat
less intensive distribution structure. Product positioning has a sample mean of 3.91
with a standard deviation of .85, which indicates high end product positioning for
the brands included in the sample. Target focus has a mean of 2.17 with a standard
deviation of .94 and need for channel control averages 2.99 with a standard
deviation of .91. Both means indicate a relatively centered distribution within the
sample. On a scale from 0 to 50, the index for managerial support has an average
of 17.42. Finally, the average score for the index of contractual restrictiveness
is 1.96. On the scale of 0 to 7, this indicates a generally low contractual
restrictiveness for the sample. Financial attractiveness, investment level, and
96
relationship uncertainty are all on the lower end with means of 1.73, 2.88, and
1.48 respectively.
Table 5.1
Sample Characteristics: Dependent and Independent Variables
Construct0 Measure Mean Standard
Deviation
n
Distribution Intensity INTENSE 2.12 .99 83
Product Positioning POSITION 3.91 .85 82
Target Focus FOCUS 2.17 .94 80
Need for Channel Control CONTROL 2.99 .91 84
Financial Attractiveness FINANCE 1.73 .71 80
Managerial Support0 SUPPORT 17.42 8.53 85
Contractual Restrictiveness3 ) CONTRACT 1.96 2.11 85
Investment Level INVEST 2.88 .91 85
Relationship Uncertainty UNCERTTY 1.48 .61 83
0 Measurement scales: Minimum = 1; Maximum = 5
2 ) Index measure: Minimum = 0; Maximum — 7
3 ) Index measure: Minimum = 0; Maximum = 50
5.1.2 Dependent Variable: Distribution Intensity
The dependent variable of distribution intensity was operationalized by six
items as described in the previous chapter (Table 4.1). Of these six items in the
questionnaire four were used to compute distribution intensity (INTENSE) based
on factor analysis and correlation results. While all six items appeared to have
face validity, the items dropped did not contribute to the reliability of the scales.
97
The four items retained after factor analysis are presented in Table 5.2.
Intercorrelation between the four items are positive and significant. Cronbach’s
Coefficient Alpha for the standardized variables is .81, which is well above the
suggested .60 cutoff level suggested for basic marketing research (cf. Churchill,
1979).
Table 5.2
Validation of Distribution Intensity Measure (INTENSE)
D I2 : In com parison to a l l o th e r speaker brands in th e m arke t, our brand tends to
be a v a ila b le through more r e t a il o u tle ts per tra d e a re a .
D I3 : W e re q u ire r e t a ile r s to meet s t r i c t m an ufacture r imposed s ta n d a rd s b e fo re
we le t them s e ll our brand.
D I4 : W e a re h ig h ly s e le c tiv e in our ch o ic e o f r e t a ile r s who c a r r y our brand in
each tra d e are a.
D I5 : O nly a s e le c t few r e t a il o u tle ts per tra d e area a re a llo w e d to c a r r y our
brand.
Cronbach C o e ffic ie n t Alpha
f o r R A W v a ria b le s : 0.804012
fo r STANDARDIZED v a ria b le s : 0.805086
Pearson C o rre la tio n C o e ffic ie n ts / Prob > ]R| under Ho: Rho=0 / Number o f O b se rva tio n s
DI3 DI4 DI5 DI2
DI3 1.00000 0.71289 0.51669 0.36294
0.0 0.0001 0.0001 0.0006
85 85 83 85
DI4 1.00000 0.64523 0.32454
0 .0 0.0001 0.0024
85 83 85
DI5 1.00000 0.48585
0.0 0.0001
83 83
DI2 1.00000
0 .0
85
98
Convergent validity of the construct measure was checked by relating the
measure to an alternative item (DI6) directly related to distribution intensity and
independent from the measure (see Table 4.1). Convergent validity refers to the
extent of agreement between two maximally independent measures of the same
construct. If the construct measure is related in a positive significant fashion to the
alternative measures, a claim for convergent validity can be made. Table 5.3
presents the correlation between the construct measure and its maximally different
alternative. The correlation is positive and significant at the p <0.0001 level.
Therefore a reasonable claim of convergent validity can be made.
Table 5.3
Convergent Validity of Distribution Intensity Measure
D I6: Compared to you r c o m p e titio n ,
d is t r ib u t io n p a tte rn fo r each
how would you d e s c rib e you r b ra n d 's
tra d e area?
Pearson C o rre la tio n C o e ffic ie n ts / Prob > {RI under Ho: Rho=0 / Number o f O b s e rv a tio n s
INTENSE DI6
INTENSE 1.00000 0.64925
0 .0 0.0001
83 80
DI6 1.00000
0 .0
82
5.1.3 Independent Variables
The independent variables can be split into manufacturer aspects and
intermediary aspects. The independent variables related to manufacturer aspects,
99
Product Positioning (POSITION), Target Focus (FOCUS) and Need for Channel
Control (CONTROL) were measured through multiple items on 5 point scales.
Nine items were included in the questionnaire to measure product
positioning (Table 4.2). Six of these items were retained to compute product
positioning (POSITION), based on factor analysis and correlation results.
Although all nine items seemed to have face validity, not all contributed to the
reliability of the scales. The six items retained after factor analysis and their
intercorrelations are presented in Table 5.4. Intercorrelations between the six
items are positive and significant at p <0.0001. Cronbach’s Coefficient Alpha for
the standardized items is .90, well above the cutoff level of .60. The other items
did not contribute to the reliability of the scale and were dropped.
Similarly, the questionnaire used six items to measure target focus (Table
4.3). Based on factor analysis and correlations, four items were included in the
construct measure of target focus (FOCUS). The four items and their
intercorrelations are presented in Table 5.5. Although all items appeared to have
face validity, two did not contribute to the reliability of the construct measure.
Once again, the correlations for the included items are all positive and significant
at the p < 0.02 level. Cronbach’s Coefficient Alpha is .72, again well above the
recommended cutoff point.
100
Table 5.4
Validation of Product Positioning Measure
PP1: Our brand c a r r ie s a v e ry p re s tig io u s image in th e m a rke tp la ce .
PP2: Our brand is known fo r it s h ig h perform ance.
PP3: Our d is t r ib u t io n s tr u c tu r e emphasizes h ig h end r e t a ile r s .
PP4: R e ta il P ric e
PP5: P re s tig e /Im a g e o f Brand
PP9: In com parison to a l l o th e r speaker brands a v a ila b le in the
where is you r brand p o s itio n e d ?
dom estic m arke t.
Cronbach C o e ffic ie n t Alpha
f o r R AW v a ria b le s : 0.902783
f o r STANDARDIZED v a ria b le s : 0.902830
Pearson C o rre la tio n C o e ffic ie n ts / Prob > jR { under Ho: Rho=0 / Number o f O b se rva tio n s
PP9 PP5 PP1 PP2 PP4 PP3
PP9 1.00000 0.65942 0.59136
0 .0 0.0001 0.0001
82 82 82
0.68238
0.0001
82
0.74914
0.0001
82
0.64216
0.0001
82
PP5 1.00000 0.72625
0 .0 0.0001
85 85
0.71933
0.0001
85
0.54052
0.0001
85
0.57911
0.0001
85
PP1 1.00000
0 .0
85
0.69645
0.0001
85
0.47911
0.0001
85
0.45926
0.0001
85
PP2 1.00000
0.0
85
0.45501
0.0001
85
0.52971
0.0001
85
PP4 1.00000
0 .0
85
0.60511
0.0001
85
PP3 1.00000
0 .0
85
101
Table 5.5
Validation of Target Focus Measure
TF3: W e use a n ic h e s tra te g y f o r m arke ting our brand.
TF4: The m a rk e tin g s tra te g y fo r our brand is not focused on s p e c ific consumer
segm ents.
TF5: Our speaker brand can be best d e s c rib e d as a brand o f mass a p p e a I.
TF6: Most o f our r e t a i l accounts c o u ld be c h a ra c te riz e d as mass
r e t a ile r s .
m erchandi s i ng
Cronbach C o e ffic ie n t A lpha
f o r R A W v a ria b le s : 0.723102
f o r STANDARDIZED v a ria b le s : 0.722550
Pearson C o rre la tio n C o e ffic ie n ts / Prob > |R| under Ho: Rho=0 / Number o f O b se rva tio n s
TF3 TF4 TF5 TF6
TF3 1.00000 0.36858 0.42347
0 .0 0.0007 0.0001
82 81 82
0.36207
0.0009
81
TF4 1.00000 0.25697
0 .0 0.0190
83 83
0.37699
0.0005
81
TF5 1.00000
0 .0
85
0.57790
0.0001
83
TF6 1.00000
0 .0
83
The third independent construct of need for control was measured by eight
items. Correlation and factor analysis resulted in retaining four items to measure
the control construct (CONTROL). These four items and their intercorrelations
are presented in Table 5.6. All other items did not contribute to the reliability of
the scale and were dropped. As before, the correlations between the four items are
102
positive and significant, although to a somewhat lesser degree of p < .04.
Cronbach’s Coefficient Alpha is .66. Even though these results are weaker than
those obtained for product positioning and target focus, coefficient alpha is still
above the recommended .60 cutoff point.
Table 5.6
Validation of Need for Control Measure
CC1: I f we would n o t c o o rd in a te th e m arke tin g a c t iv it ie s o f our r e t a ile r s , th e y
w ould v a ry a g re a t deal in how th e y s e ll our brand.
Please indicate the extent to which you as th e m an u fa ctu re r directly or indirectly
( i . e . throu gh you r d is tr ib u to r s ) e x e rt effort to c o o rd in a te th e fo llo w in g
r e t a ile r a c t iv it ie s .
CC3: Im ple m en ta tion o f Sales P rom otions
CC4: Use o f P rom o tion al A id s ( e .g . B rochures, P o ste rs)
CC5: Use o f P roduct D is p la y s ( e .g . Sales Racks e tc .)
Cronbach C o e ffic ie n t Alpha
f o r R AW v a ria b le s : 0.663246
f o r STANDARDIZED v a ria b le s : 0.663700
Pearson C o rre la tio n C o e ffic ie n ts / Prob > J R | under Ho: Rho=0 / Number o f Obse
CC3 C C S CC4 CC1
CC3 1.00000 0.31858 0.32223 0.30460
0 .0 0.0030 0.0026 0.0048
85 85 85 84
CC5 1.00000 0.43747 0.22787
0.0 0.0001 0.0371
85 85 84
CC4 1.00000 0.37153
0 .0 0.0005
85 84
CC1 1.00000
0 .0
84
103
All of the above three independent measures were checked for convergent
validity and the results are displayed in Table 5.7. Convergent validity for the
product positioning construct was checked by relating the measure to the
alternative item of overall product quality (PP8). The alternative item covers one
of the two dimensions of product positioning as conceptualized in Chapter three
and is maximally different from the measure. The correlation between measure and
alternative item is positive and significant at the p <0.0001 level. Therefore a
reasonable claim of convergent validity can be made. Similarly, convergent validity
for target focus was checked, relating the construct measure to an alternative item
looking at the spectrum of pursued consumers. Again, the correlation is positive
and significant at the p <0.0001 level.
The convergent validity check for the third independent variable of need for
channel control differed slightly from the previous checks. Due to the character of
the measurement instrument, the alternative variable for convergent validity is not
a single variable. It is a combination of three items, CC6, CC7 and CC8,
integrated into one alternative measure for the construct (see Table 4.4).
Correlation and factor analysis verifies that the three included items contribute
highly to the reliability of the scale. Intercorrelations and the items are presented in
Table 5.7. Correlations between the items are positive and significant and
Cronbach’s Coefficient Alpha is .75. The alternative measure is chosen because it
represents the same concept of need for control as the construct measurement. In
104
contrast to the measurement which covers variables of combined effort between
manufacturer and dealer, the alternative measure focuses on dealer initiated
variables. As before, the correlation between the two alternative measures is
positive and significant at the p <0.0012 level. While slightly weaker than for the
other two independent variables, the construct measure of need for control can also
be assumed to have convergent validity.
The independent variables related to intermediary aspects of the framework
include Financial Attractiveness (FINANCE), Investment Level (INVEST),
Relationship Uncertainty (UNCERTTY), Contractual Restrictiveness
(CONTRACT), and Managerial Support for dealers (SUPPORT). While
relationship uncertainty and investment level were entirely measured through
multiple items on a 5 point scale, financial attractiveness also included a 7 point
scale for the brands actual sales volume in 1991 and a direct measure of market
share. The constructs of contractual restrictiveness and management support were
designed as indices with yes/no answers. For the managerial support construct the
yes/no scale was extended to also include a 5 point scale to determine the level of
manufacturer provided services.
105
Table 5.7
Convergent Validity of Independent Measures: Manufacturer Aspect
Product Positioning Measure:
PP8: How do you position your brand on the following product
characteristics? Overall Product Quality
POSITION PP8
POSITION 1.00000 0.60243
0.0 0.0001
82 85
PP8 1.00000
0.0
85
Target Focus Measure:
TF2: By design, our speakers appeal to a narrow spectrum of
consumers only.
FOCUS TF2
FOCUS 1.00000 0.57962
0 .0 0.0001
80 80
TF2 1.00000
0 .0
85
Need for Channel Control:
CONTROL NFC_1
1.00000 0.34707
0 .0 0.0012
84 84
1.00000
0.0
85
Pearson C o rre la tio n C o e ffic ie n ts / Prob > j R j under Ho: Rho=0 / Number o f O b se rva tio n s
CONTROL
NFC ALT
106
Table 5.8
Validation of Alternative Measure for Need for Channel Control
Please indicate th e extent to which you as th e m a n u fa ctu re r directly or indirectly
( i . e . throu gh your d is tr ib u to r s ) e x e rt effort to c o o rd in a te the fo llo w in g
r e t a ile r a c t iv it ie s .
CC6: D e a le r S ales P re s e n ta tio n s to Consumers
CC7: S to re Appearance ( e .g . Layout, S h e lf-D e s ig n e t c . )
CC8: D e ale r In v e n to ry Levels
Cronbach C o e ffic ie n t Alpha
fo r R A W v a ria b le s : 0.749728
fo r STANDARDIZED v a ria b le s : 0.749728
Pearson C o rre la tio n C o e ffic ie n ts / Prob > |R| under Ho: Rho=0 / N = 85
CC6 CC7 CC8
CC6 1.00000 0.63779 0.38762
0.0 .0001 0.0002
CC7 1.00000
0 .0
0.47350
.0001
CC8 0.47350
0.0001
1.00000
0.0
Of the eight items measuring financial attractiveness in the questionnaires
four were retained to measure the construct (FINANCE), based on correlation and
factor analysis results. These four items and their intercorrelations are presented
in Table 5.9. The other items did not contribute to the reliability of the scale and
were dropped. As before, the correlations between the four items are positive and
107
significant and Cronbach’s Coefficient Alpha is .84, well above the .60 cutoff
point.
Surprisingly, the two items measuring the profitability had to be dropped
from consideration. Two reasons could help explain their failure to contribute to
the explanation of financial attractiveness. First, for some brands the diversity of
profit margins within the retailer network may have prohibited meaningful
responses. Secondly, the limited number of firms in the market may cause the
industry to settle for generally accepted standard retail margins. Consequently, the
second dimension of financial attractiveness which focused on profitability will not
be represented in the construct measurement.
Item FA8 which inquires about the market share for the brand was used as
the alternative measure for checking convergent validity. The measure was deemed
appropriate, because it focuses on the dimension of overall sales potential for the
brand and is maximally different from the items included in the construct measure.
The correlation between the alternative measures is positive and significant at the p
<0.0006 level and a reasonable claim of convergent validity can be made. The
results are presented in Table 5.9.
The questionnaire contained only three items for measuring the dealer
investment level for carrying the brand. As a result of factor analysis and
108
correlations, two of these measures were retained to measure the construct
(INVEST). The third item did not contribute to the reliability of the scale. As
displayed in Table 5.11, the correlation between the remaining two items was
positive and significant at the p < .0001 level with a .88 reliability coefficient.
Table 5.9
Validation o f Financial Attractiveness M easure
FA1: Compared to a l l o th e r speaker brands, our brand tends to g e n e ra te th e
la rg e s t s a le s volume in speakers fo r r e t a ile r s who c a rry i t .
R e la tiv e to all other home speaker brands in th e m arket, how does
compare on th e fo llo w in g issues?
you r brand
FA2: c a lc u la te d in US $ is . . .
FA3: c a lc u la te d in units is . . .
FA7: What was th e approxim ate s a le s volume ( in US $ ) o f your brand in
P lease mark th e a p p ro p ria te c a te g o ry .
1991?
Cronbach C o e ffic ie n t Alpha
fo r R A W v a ria b le s : 0.845544
fo r STANDARDIZED v a ria b le s : 0.847076
Pearson C o rre la tio n C o e ffic ie n ts / Prob > jRj under Ho: Rho=0 / Number o f O b s e rv a tio n s
FA1 FA2 FA3 FA7
FA1 1.00000 0.60649 0.56194
0 .0 0.0001 0.0001
85 80 80
0.33239
0.0021
83
FA2 1.00000 0.86569
0 .0 0.0001
80 80
0.48932
0.0001
80
FA3 1.00000
0 .0
80
0.62823
0.0001
80
FA7 1.00000
0 .0
83
109
Table 5.10
Convergent Validity of Financial Attractiveness Measure
FA8: What is you r b ra n d 's e s tim a te d m arket share in th e loudspeaker m arket?
Pearson C o rre la tio n C o e ffic ie n ts / Prob > jR| under Ho: Rho=0 / Number o f O b se rva tio n s
FINANCE FA8
FINANCE 1.00000 0.43443
0.0 0.0006
80 59
FA8 1.00000
0 .0
59
Table 5.11
Validation of Investment Level Measure
I L 1: The amount o f m on ey d e a le rs have to spent on t r a in in g new s a le s people in
o rd e r to handle our brand is . . .
IL 2 : The amount o f tim e new s a le s people have to spent on t r a in in g in o rd e r to
ha nd le our brand is . . .
Cronbach C o e ffic ie n t A lpha
f o r R A W v a ria b le s : 0.875791
f o r STANDARDIZED v a ria b le s : 0.875791
Pearson C o rre la tio n C o e ffic ie n ts / Prob > ] R j under Ho: Rho=0 / N = 85
I LI IL2
IL1 1.00000 0.77903
0.0 0.0001
IL2 1.00000
0 .0
110
Similarly, the construct of relationship uncertainty was measured by four
items in the questionnaire. Again, only two items were retained as a result of
correlation and factor analysis. The other two items did not contribute to the
reliability of the scale and were dropped. Table 5.12 shows the correlation of the
two remaining items, which was positive and significant at the p < .0001 level
and a reliability coefficient of .69, which although being somewhat weak still
meets the cutoff requirement.
Table 5.12
Validation of Relationship Uncertainty Measure
RU1: W e c o n s id e r our r e la tio n s h ip w ith r e t a ile r s to be a lo ng te rm a llia n c e .
RU2: R e ta ile r s who c a rry our brand a re expected to be s e llin g ou r p ro d u c ts f o r a
long tim e .
Cronbach C o e ffic ie n t Alpha
fo r R A W v a ria b le s : 0.691033
fo r STANDARDIZED v a ria b le s : 0.691033
Pearson C o rre la tio n C o e ffic ie n ts / Prob > |R | under Ho: Rho=0 / N = 83
RU 1 RU2
RU1 1.00000
0 .0
0.52792
0.0001
RU2 1.00000
0 .0
For the remaining two constructs of contractual restrictiveness
(CONTRACT) and managerial support (SUPPORT) indices were formed, which
111
were designed to show the extent of restrictions or support in the relationship
between manufacturer and intermediary.
While Cronbach’s Coefficient Alpha establishes reliability for the
independent constructs, and this was confirmed by factor analysis, another
necessary requirement of academic research is to demonstrate that the measures
also possess divergent validity. Gerbing and Anderson (1988) argue that all
possible scales should be used in one single analysis so that consistency can be
checked. The question is whether the measures reflect essentially distinct
underlying constructs.
In order to check for divergent validity, factor analysis was performed on
all constructs included in the measurement of independent variables except for the
indices. The factor analysis used all scale items included in the six remaining
constructs. The results of Varimax Rotation are displayed in Table 5.13. The
rotated factor pattern indicates that a six factor solution fits the data best. The first
six factors have eigenvalues greater than one, and the cumulative variance
explained by these six factors is above 72%. In addition, the hypothesized factor
loadings after a varimax rotation are generally above .50 with only three variables
below that level. Furthermore, all but six variables have loadings below .40 on
inappropriate factors.
112
One reason for the exceptions could be the large number of variables and
constructs bundled into one analysis step. Another reason of course would be a
lack of discriminatory power between constructs. Therefore, a multi-trait matrix
of all independent constructs was designed. The diagonal included Cronbach’s
Coefficient Alpha for each construct while the off-diagonal cells showed the
correlations between constructs. Table 5.14 shows the results. All reliability
coefficients are higher than any off-diagonal correlation, suggesting that the
correlation within a construct is higher than the correlations between constructs.
Some of the correlations between constructs exceed 40%, though. Most of the
exceptions indicated in the factor analysis of all construct variables are part of
these construct correlations.
Consequently a bi-construct factor analysis was conducted, containing all
variables pertaining to the constructs in question. Table 5.15 shows the factor
analysis results for the product positioning construct (POSITION) versus target
focus (FOCUS) or investment level (INVEST). In both cases, it shows a two
factor solution as best fitting the data. The hypothesized factor loadings after
varimax are above .50 for the positioning versus focus analysis with all but one
variable loading significantly lower on the inappropriate factor. For the
comparison between positioning and investment level, loadings exceed .70 with all
variables loading significantly lower on inappropriate factors. The factor
113
eigenvalues for both comparisons exceed 2.5 and the cumulative explained variance
exceeds 65% and 70% respectively.
Table 5.13
Divergent Validity of Independent Measures
I tem FACTOR1 FACT0R2 FACT0R3 FACT0R4 FACT0R5 FACT0R6
PP9 0.84123 -0.04305 -0.04995 0.10712 0.19743 0.08990
PP5 0.83885 -0.07311 0.07332 -0.24653 0.08806 0.02685
PP1 0.80033 0.16580 -0.05205 -0.16515 0.14190 -0.01505
PP2 0.83718 -0.00927 0.15341 -0.11511 0.22785 -0.00219
PP4 0.72256 0.06819 -0.33951 0.22963 0.13917 0.14085
PP3 0.71953 -0.17971 -0.18400 0.03581 -0.02506 0.40461
TF3 0.56279 -0.05872 0.14087 -0.11997 -0.05846 0.33351
TF4 0.15817 -0.07300 0.08312 -0.16306 0.00161 0.82752
TF5 0.55616 -0.34966 -0.41688 0.17490 0.25908 0.26571
TF6 0.37104 -0.32933 -0.20711 0.01137 -0.04100 0.57164
CC3 -0.21643 0.04235 0.48569 -0.15137 0.45195 0.10150
CC5 -0.08852 0.39368 0.58580 0.20174 0.06899 0.03854
CC4 0.08679 0.10145 0.75184 0.15435 0.20228 0.09538
C C1 -0.00544 0.11630 0.74336 -0.17514 -0.00610 -0.15 71 7
FA1 0.21082 0.64055 0.35576 -0.11644 -0.24488 -0.34321
FA2 0.14090 0.90743 0.06122 0.03066 0.08750 -0.11755
FA3 -0.06242 0.91332 0.14043 0.10947 0.06206 -0.07223
FA7 -0.42370 0.70325 0.14068 -0.18282 0.15342 0.01325
IL1 0.40284 0.06960 0.20120 -0.02235 0.76116 0.00677
IL2 0.41166 0.09549 0.04991 -0.13536 0.80600 -0.08 70 9
RU1 -0.10460 0.05028 -0.04481 0.83127 -0.04655 -0.02 25 9
RU2 -0.04879 -0.05233 0.06144 0.85897 -0.09516 -0.13 03 8
E i genvaIue 6.1026 4.0420 1.9579 1.5557 1.2158 1.0321
% of Variance 0.2774 0.4611 0.5501 0.6208 0.6761 0.7230
114
Table 5.14
Divergent Validity of Independent Measures: Multitrait Matrix
POSITION FOCUS CONTROL FINANCE INVEST UNCERTTY CONTRACT SUPPORT
POSITION 0 .9 0
FOCUS 0.61 0 .7 2
CONTROL -0 .1 0 -0.21 0 .6 6
FINANCE -0.11 -0 .3 9 0.43 0 .8 5
INVEST 0.4 6 0.29 0.17 0.08 0 .8 8
UNCERTTY -0 .1 3 -0 .1 2 0.08 -0 .0 4 -0 .1 4 0 .6 9
CONTRACT 0.11 -0 .0 9 0.26 0.25 0.19 -0 .2 2 N/A
SUPPORT -0 .2 8 -0 .4 0 0.44 0.34 -0 .0 6 0 .0 6 0.11 N/A
Table 5.15
Bi-Construct Analysis of Divergent Validity
POSITION versus FOCUS POSITION versus INVEST
R o tated F actor P a tte rn R otated F a c to r P a tte rn
FACTOR1 FACT0R2 FACT0R1 FACT0R2
PP9 0 .8 1 7 5 8 0.31232 PP9 0 .8 5 4 5 2 0.23415
PP5 0 .8 5 7 8 3 0.16232 PP5 0 .8 1 3 2 0 0.24936
PP1 0 .8 5 8 3 0 0.05430 PP1 0 .7 3 6 1 0 0.30521
PP2 0 .8 4 8 2 9 0.19145 PP2 0 .7 3 6 7 9 0.42501
PP4 0 .6 3 8 6 5 0.42617 PP4 0 .7 7 0 0 1 0.14092
PP3 0 .5 3 3 5 6 0.65669 PP3 0 .8 0 9 2 7 0.00688
TF3 0.39033 0 .5 1 4 7 8 I L 1 0.18292 0 .9 1 1 6 9
TF4 -0.00894 0 .6 8 8 4 1 IL2 0.20898 0 .9 2 2 4 8
TF5 0.43072 0 .6 6 4 0 6
TF6 0.11930 0 .8 2 6 1 1
E igenvalues 3.905328 2.638701
E igenvalues 3.801168 2.092839
C u m ulative P ercent o f
V a ria n ce E x p la in e d : 65.44%
C um ulative P ercent o f
V arian ce E x p la in e d : 73.68%
115
Similarly, the factor analysis between need for control (CONTROL) and
financial attractiveness (FINANCE) constructs shows a two factor solution best
fitting the data with eigenvalues above 1.4 and explained variance greater than
60%. Table 5.16 shows that hypothesized factor loadings after varimax are greater
than .50 with all variables loading significantly lower on inappropriate factors.
Overall, the results of the initial six construct factor analysis in combination
with the detailed analysis show consistency and support the notion of divergent
validity between the six constructs.
Table 5.16
Bi-Construct Analysis of Divergent Validity
CONTROL versus FINANCE
R otated F actor P a tte rn
FACTOR1 FACT0R2
CC3 0.03140 0 .7 1 5 4 9
CC5 0.43405 0 .5 4 1 6 3
CC4 0.12230 0 .7 6 1 8 2
CC1 0.17904 0 .6 8 9 6 8
FA1 0 .7 3 4 6 9 0.21503
FA2 0 .9 1 3 8 0 0.03552
FA3 0 .9 2 3 0 2 0.11930
FA7 0 .6 8 4 5 1 0.22967
E i genvaIues 2.931707 1.975798
C um ulative P ercent o f
V arian ce E x p la in e d : 6 1 .3 4 %
116
5.2 Hypotheses Tests
This section presents the results of the tests performed to confirm the seven
hypotheses developed in Chapter III. The study is exploratory in nature and
therefore does not lend itself to highly specified model testing. On the other hand,
the framework represents a fully recursive set of two simultaneous equations with
(1) INTENSE = f (POSITION, FOCUS, CONTROL, FINANCE, INVEST,
CONTRACT, SUPPORT, UNCERTTY, INTERACT) and
(2) CONTROL = g (POSITION, FOCUS).
Therefore, the framework and its hypotheses is tested with a two-step least squares
linear regression model in a system of linear equations (SYSLIN 2SLS) as it is the
most appropriate tool for checking the hypothesized relationships.
5.2.1 Original Framework
The SYSLIN procedure with two-stage least square estimation was run on
the model as shown in Figure 3.1. Tables 5.17 and 5.18 present the two
regression models of (1) CONTROL on POSITION and FOCUS and (2) INTENSE
on all independent constructs. Results for the first regression model, presented in
Table 5.17, indicate that the model is not significant (F = 1.961, p < . 14). The
model explains less than 3% of the variance (Adjusted R-square = .0256). Results
for the overall model, presented in Table 5.18, show that the overall model is
significant (F=20.159, p < .0001) and explains almost 68% of the variance
117
(Adjusted R-square = .6774). On the other hand, the analysis also shows that the
model is mis-specified. The CONTROL variable is estimated to be a linear
combination of all other independent variables. The solutions for all parameters in
the model are biased and the statistics can be misleading.
In order to arrive at an unbiased estimate for the overall model and its
parameters, it was necessary to find a model specification which is full rank when
analyzed by two-stage least square estimation. Further analysis by simply
regressing CONTROL on POSITION or FOCUS and correlating the residuals
showed an extremely high correlation coefficient of more than 98 %, which
indicated that an important interaction term between POSITION and FOCUS
exists. Conceptually, such an interaction term makes sense because products at the
extreme high end (e.g. $ 10,000 per speaker) will have only a very narrow
segment of potential customers, which means the manufacturer must have an
extremely specific target focus. In such cases manufacturer’s need for control over
the channel must be extremely high, since one lost sale in a segment with only a
small number of customers can hurt the firm tremendously. Since the introduction
of an interaction term between product positioning and target focus makes
conceptual sense and is important for correct model specification, subsequent
analysis procedures were run on the expanded model only. The new model is
shown in Figure 5.1 and has the following form:
118
(1) INTENSE = f (POSITION, FOCUS, CONTROL, FINANCE,
INVEST, CONTRACT, SUPPORT, UNCERTTY,
INTERACT) and
(2) CONTROL = g (POSITION, FOCUS, POS*FOC).
5.2.2 Expanded Framework
Like before, the SYSLIN procedure with two-stage least square estimation
was run on the new expanded model as shown in Figure 5.1. Tables 5.19 and
5.20 present the two regression models of (1) CONTROL on POSITION FOCUS
and POS*FOC and (2) INTENSE on POSITION, FOCUS, CONTROL,
FINANCE, INVEST, CONTRACT, SUPPORT, UNCERTTY, INTERACT.
Results for the first regression model, presented in Table 5.19, show that
the model is significant (F=4.151, p<.0091). The model explains about 11.5%
of the variance (Adjusted R-square = .1147). FOCUS and POS*FOC are
significant, but are inversely related to CONTROL. POSITION is positively
related to CONTROL although nonsignificant (p < .54). The model is surprising
in as much as the negative relationship between FOCUS and CONTROL and the
low significance of POSITION as a predictor of CONTROL is unexpected. Both
phenomena will be discussed in detail in the next section.
119
Table 5.17
Original Framework: Regression Model Part I
SYSLIN Procedure: Two-Stage Least Squares E stim ation
Dependent v a r ia b le : CONTROL
A n a ly s is o f V arian ce
Source
Model
E rro r
C T o ta l
DF
2
71
73
Sum o f
Squares
1.87182
33.89171
35.76353
Root M SE 0.69090
Dep Mean -0.00131
C.V. -52594.69848
Mean
Square
0.93591
0.47735
R-Square
Adj R-SQ
F Value
1.961
0.0523
0.0256
Prob>F
0.1483
Param eter E stim a te s
Param eter Standard T fo r HO:
V a ria b le DF E stim ate E rro r Parameter=0 Prob > ]T j
INTERCEP 1 0.015509 0.080837 0.192 0.8484
POSITION 1 0.069883 0.123194 0.567 0.5723
FOCUS 1 -0.273037 0.148859 -1.83 4 0.0708
120
Table 5.18
Original Framework: Regression Model Part II
SYSLIN Procedure Two-Stage Least Squares E stim atio n
Dependent v a r ia b le : INTENSE
Analysis o f Variance
Source
Model
E rro r
C T o ta l
DF
8
65
73
Sum o f
Squares
30.16363
12.15758
42.32120
Mean
Square
3.77045
0.18704
F Value
20.159
Prob>F
0.0001
Root M S E
Dep Mean
c.v.
0.43248
-0.05800
-745.63515
R-Square
Adj R-SQ
0.7127
0.6774
NOTE: The model is n o t f u l l ra n k. Least Squares s o lu tio n s fo r th e param eters a re
not un iq u e . C e rta in s t a t is t ic s w i ll be m is le a d in g . A re p o rte d degree o f
freedom o f 0 o r B means th e e s tim a te is b iase d. The fo llo w in g param eters
have been s e t to z e ro . These v a ria b le s are a lin e a r c o m b in a tio n o f o th e r
v a ria b le s as shown.
CONTROL = -0.0345 * INTERCEP -0.13 97 * POSITION +0.0312 * FOCUS
+0.2013 * FIN 1 +0.2309 * INVEST +0.0271 * UNCERTTY
+0.0730 * C R IND +0.1815 * SUPPORT +0.1373 * INTERACT
Param eter E stim ate s
Param eter Standard T fo r HO:
V a ri a b le DF E stim ate E rro r Parameter=0 Prob > IT ]
INTERCEP B -0.093517 0.055905 -1.673 0.0992
POSITION B -0.504010 0.088776 -5 .6 7 7 0.0001
FOCUS B -0.178440 0.111055 -1.60 7 0.1130
CONTROL 0 0 .
FINANCE B 0.131939 0.072957 1.808 0.0752
INVEST B -0.016738 0.065291 -0.25 6 0.7985
UNCERTTY B 0.131715 0.060158 2.189 0.0322
CONTRACT B -0.124184 0.056257 -2 .2 0 7 0.0308
SUPPORT B 0.031537 0.059471 0.530 0.5977
INTERACT B 0.354451 0.157963 2.244 0.0283
121
Figure 5.1: Expanded Framework of Distribution Intensity
H 4: (+) H 5: (+)
H3: (-) H 2: (-)
H 6: (+)
H 8: (-)
Barriers to
Intermediary
Participation
H7c: (-)
H7b:/(+) H7d\(+)
H7a: (+)
Interaction
(POS*FOC)
Need for
Control
Financial
Attractiveness
Investment
Level
Management
Support
Contractual
Restrictiveness
Target
Focus
Product
Positioning
Relationship
Uncertainty
Distribution
Intensity
Manufacturer
Aspect
Intermediary
Aspect
Results for the overall model, presented in Table 5.20, show that the
overall model is significant (F = 14.272, p c.0 0 0 1 ) and explains more than 62% of
the variance (Adjusted R-square = .6207). All variables related to the
manufacturer aspect of the framework (POSITION, FOCUS, CONTROL) are
inversely related to INTENSE. POSITION and CONTROL are significant at the p
< .0001 and p < .0660 levels respectively, while FOCUS is nonsignificant.
Three variables related to the intermediary aspect of the framework, FINANCE,
UNCERTTY and INTERACT), are significant at p < .05 and show positive
relationships with INTERACT. While FINANCE and INTERACT are expected to
display such positive relationship, for UNCERTTY it is unexpected. The other
three variables of INVEST, CONTRACT and SUPPORT are all nonsignificant and
only CONTRACT shows an expected inverse relationship to INTENSE.
5.2.3 Hypotheses
5.2.3.1 Manufacturer Aspect of Framework
The three direct influences on distribution intensity are assumed to be
product positioning, target focus and need for channel control. In addition to the
direct influence, product positioning and target focus were also modeled as
indirectly influencing intensity via their effect of need for channel control.
123
Hypothesis 1 deals with the impact of product positioning on distribution
intensity and expects high end products to be distributed less intensively than low
end products. Hypothesis 1 is completely confirmed by the model. The parameter
estimate for the model shows a inverse relationship between POSITION and
INTENSE and significance at the p < .0001 level.
Hypothesis 2 expected firms with a very narrow target focus to distribute
their products through fewer retail outlets than firms which target several consumer
segments. While the model estimation confirms the inverse relationship between
FOCUS and INTENSE, the relationship appears to be nonsignificant. A possible
explanation could be the fact that FOCUS is highly significant in explaining the
CONTROL variable and CONTROL is significant in explaining INTENSE,
therefore taking away direct explanation power of FOCUS. Using multiple
regression without the indirect effects via CONTROL shows significance for
FOCUS on INTENSE at the p <0.11 level, thereby lending support for this
explanation.
The third hypothesis dealt with the inverse relationship between a
manufacturer’s need for channel control and the intensity of distribution found for
its products. The model confirms both the inverse relationship between
CONTROL and INTENSE by estimating a negative parameter as well as its
significance (p<.066).
124
Table 5.19
Expanded Framework: Regression Model Part I
SYSLIN Procedure Two-Stage Least Squares E s tim a tio n
Dependent v a r ia b le : CONTROL
A n a ly s is o f V arian ce
Source
Model
E rro r
C T o ta l
DF
3
70
73
Sum o f
Squares
5.40169
30.36183
35.76353
Root M SE 0.65859
Dep Mean -0.00131
C.V. -50134.81124
Mean
Square
1.80056
0.43374
R-Square
Adj R-SQ
F Value
4.151
0.1510
0.1147
Prob>F
0.0091
Param eter E stim a te s
Param eter Standard T fo r HO:
V a ria b le DF E stim ate E rro r Parameter=0 Prob > |T|
INTERCEP 1 0.180658 0.096380 1.874 0.0650
POSITION 1 0.071399 0.117433 0.608 0.5452
FOCUS 1 -0.386184 0.147336 -2.621 0.0107
POS*FOC 1 -0.436112 0.152874 -2.85 3 0.0057
125
Table 5.20
Expanded Framework: Regression Model Part II
SYSLIN Procedure Two-Stage Least Squares E s tim a tio n
Dependent v a r ia b le : INTENSE
A n a ly s i s o f V arian ce
Sum o f Mean
Source DF Squares Square F Value Prob>F
Model 9 31.00806 3.44534 14.272 0.0001
E rro r 64 15.45045 0.24141
C T o ta l 73 42.32120
Root M SE 0.49134 R-Square 0.6674
Dep Mean -0.05800 Adj R-SQ 0.6207
C.V. -847.11040
Param eter Est im ates
Param eter S tandard T fo r HO:
V a ri a b le DF E stim ate E rro r Parameter=0 Prob > }T j
INTERCEP 1 -0.114642 0.064510 -1.77 7 0.0803
POSITION 1 -0.589462 0.110724 -5.324 0.0001
FOCUS 1 -0.159352 0.126581 -1.259 0.2126
CONTROL 1 -0.611619 0.327023 -1.870 0.0660
FINANCE 1 0.255063 0.105849 2.410 0.0189
INVEST 1 0.124495 0.105852 1.176 0.2439
UNCERTTY 1 0.148315 0.068919 2.152 0.0352
CONTRACT 1 -0.079517 0.068229 -1.165 0.2482
SUPPORT 1 0.142527 0.089926 1.585 0.1179
INTERACT 1 0.438419 0.184992 2.370 0.0208
126
In regard to the relationship between POSITION, FOCUS and CONTROL,
the estimation results are mixed. While the relationship between FOCUS and
CONTROL is highly significant (p< .0107) it shows an unexpected inverse
relationship to CONTROL. In other words, firms with a narrow target market
experience less control over their channel of distribution than firms which go for a
large and diverse customer base. An explanation for the surprising relationship
could be the fact that the majority of small, specialized niche players do not
possess the capability of exerting control in the channel. By choosing their
retailers carefully, they eliminate the need for control. On the other hand,
manufacturers with broad brand appeal will have a divergent dealer base and more
resources. Having more resources to administer control, they may be more open
in their acceptance of diverse intermediaries. In their case the need for
coordination and control is higher simply due to the diversity of intermediaries.
The regression of CONTROL on POSITION shows the hypothesized
positive relationship, but it is nonsignificant (see Hypothesis 5). A reason for this
lack of significance could be the introduction of the interaction term, which shows
a high level of significance in explaining CONTROL (p< .0057). The interaction
of high product positioning and narrow target focus seems to pick up a lot of the
explanatory power of POSITION.
127
5.2.3.2 Intermediary Aspect of Framework
For the intermediary aspect of the framework, three hypotheses were
developed. They deal with the financial attractiveness of a brand, barriers to entry
a dealers faces and the interaction of those two.
Hypothesis 6, which stated that financially attractive brands will have more
dealers willing to distribute them and therefore higher distribution intensity than
brands with lower financial incentives for dealers was fully supported by the
model. The model estimation yields a positive parameter for the FINANCE
variable and shows statistical significance (p <.0189).
The barriers to entry were conceptualized as four individual constructs
(CONTRACT, INVEST, UNCERTTY, SUPPORT). The only statistically
significant relationship between these barriers to entry variables and intensity is
shown for UNCERTTY (p < .0352). Surprisingly, the model estimates a positive
parameter while hypothesis 7b assumes an inverse relationship. While this result is
unexpected, it can be explained by looking at the background of this study. Only
manufacturers were interviewed, even for the intermediary aspect. Under the
assumption that higher relationship uncertainty is a sign of non-committal
manufacturer dealer relationships, it is plausible that brands have to distributed
through more outlets in each trade area. A manufacturer cannot be sure or is not
128
willing to keep dealers on a longterm basis, which explains the need to have more
alternative outlets available.
Contractual restrictiveness (CONTRACT) and investment level (INVEST)
were also hypothesized to be negatively related to INTENSE. While the model
shows an inverse relationship of CONTRACT to INTENSE, it also shows non
significance. The explanation for this phenomenon is the heavy bias towards the
low end of the variable. Most of the firms in the survey do not have any contracts
regulating the relationship with their dealers. As long as dealer credit remains in
good standing manufacturers will supply them. No other restriction in addition to
the credit standing determines the relationship between the two parties. Therefore,
the nonsignificant result in regard to hypothesis 7d is no surprise and could be
remedied by a larger sample with a larger number of manufacturers who have
contracts with their intermediaries.
INVEST, which in hypothesis 7a was construed to be inversely related to
INTENSE, shows a positive relationship in the model, although nonsignificant.
Technology has not changed much over the last years in regard to speaker design.
Therefore, training salespeople on brand specific knowledge is not a major factor
for dealers in their decision to take on new speaker brands or not. Investment into
brand specific setups within the dealerships are limited due to the nature of the
product. Sound-rooms used to show the actual speaker sound are usually already
129
in place and shared for several speaker brands. As a result, the limited explanation
power of INVEST, seems to be inherent of the business analyzed and could
radically change for other products.
The last of the four hypotheses states that brands with higher managerial
support for their dealers can be expected to have more outlets per trade area. It is
based on the logic that dealers prefer help from manufacturers and are more
willing to carry a product which has manufacturer support than products without
any such supplier help. The model supports the positive relationship, although the
variable less significant than what was expected (p < .1179).
The last hypothesis deals with the interaction effect between FINANCE, the
barriers of entry and INTENSE. Hypothesis 8 states that barriers of entry are
mediated by financial attractiveness. Even with high barriers of entry, financial
attractiveness may cause more dealers to carry a brand than if a brand has only
low entry barriers but no financial attractiveness. This hypothesis was modeled as
an interaction term (INTERACT) and the model estimation fully supports the
hypothesis. The results show a positive and highly significant relationship (p <
.0208).
130
5.3. Summary
Table 5.21 summarizes the main findings of the research. Overall, the
manufacturer aspect of the framework and its hypotheses appears reasonably well
supported, while the intermediary aspect hypotheses show no or mixed support.
The main reason for the lack of support in the intermediary aspect could be the
operationalization of the constructs. Manufacturers had to evaluate the items on
investment level, relationship uncertainty, contractual restrictiveness and
managerial support on an aggregated basis. Of course, these firm dealer
relationship characteristics may differ highly across dealers. Therefore,
manufacturer responses may have been biased by perceptions based on major retail
accounts instead of reflecting a true average. As a result, the information gathered
for the four constructs may not truly reflect the situation and have less variance
across brands than in reality. The framework aspect concerned with the influence
of product positioning, target focus and their interaction on need for channel
control does not explain much. Even though target focus and the interaction effect
are significant, the explained variance remains low (R squared = 11%). Again, the
aggregate character of the inquiry may have caused this result. The differences
between local one store dealers and national retail chains and their relationships
with the manufacturer are too diverse to be captured by aggregate measures.
131
Table 5.21
Summary of Empirical Findings
H y po th esis Su p p o r t e d
1. Products positioned at the high end of the price quality spectrum
will exhibit low distribution intensity while products positioned at
the low end will exhibit high distribution intensity.
Yes
2. With increasing focus in a brand’s targeting strategy, the intensity
of distribution for the brand will decrease. No
3. Manufacturers’ need for channel control is inversely related to the
level o f distribution intensity. Yes
4. Product positioning in regard to the price/quality spectrum is
positively related to manufacturer need for control in the channel. No
5. With increasing focus in a brand’s targeting strategy, the
manufacturer need for channel control rises. No
6. Financial attractiveness of a manufacturer’s brand is positively
related to distribution intensity. Yes
7a. The required level o f investment on the part of the retailer will be
inversely related to distribution intensity. No
7b. Uncertainty about the future prospects of a business relationship
will be inversely related to distribution intensity. No
7c. The level of manufacturer sponsored management support will be
positively related to distribution intensity. Yes
7d. Contractual restrictiveness will be inversely related to distribution
intensity. No
8. High financial attractiveness of the brand will moderate the
influence of entry barriers on distribution intensity. Yes
132
Chapter 6
Summary and Conclusions
The purpose of this chapter is to summarize the present dissertation and offer some
suggestions for future research. The research objectives are reexamined. A brief
overview of the methodology is provided and the empirical findings are
summarized. The chapter concludes with a discussion of the limitations inherent to
this study and points out some directions for future research.
6.1 Research Objectives
The objective of this study was to build an integrated conceptual framework
of distribution intensity. Distribution Intensity is the number of retail outlets per
trade area, which carry a product or brand and is an important factor of how easily
available the product is for consumers. Two problems with the existing research on
distribution intensity were identified. Both are addressed by this study.
The first problem deals with lack of direct research. Contrary to the
importance of the construct, a comprehensive research stream for the topic is
missing. Although many researchers have touched upon the issue of number of
133
intermediaries in a distribution network, there is no integrated and comprehensive
approach in existence. Copeland (1923) establishes a framework of categories of
goods and only a side aspect is devoted to distribution intensity. Aspinwall (1958)
and Miracle (1965) expand the framework but distribution intensity is again only a
side aspect. The little research that deals with distribution intensity directly, does
not focus on the factors influencing the intensity level, but instead looks at how
distribution intensity effects firm performance (cf. Hartung and Fisher 1965).
In addition, existing research only looks at intensity levels and their
relationship to one market participant at the time. Either consumer shopping
patterns, manufacturer strategy or the distributor aspects and their relationship to
intensity are conceptualized (cf. Bucklin 1965, Hartung and Fisher 1965, Webster
1976). This research project integrates two of the three dimensions and establishes
a framework which looks at manufacturer and intermediary aspects of distribution
intensity.
The second problem is the lack of empirical research. No empirical
research, which looks at the factors influencing distribution intensity, exists. A
possible reason for the lack of empirical research could be that the connection
between product categories of specialty, shopping, and convenience goods and their
inherent level of distribution intensity appears so common sense that empirical tests
seem uninteresting.
134
This point of view misses the fact that most of the competition takes place
within product classes and therefore variation in distribution intensity may be a
major determinant of success. This is supported by existing research linking
convenience goods, their distribution intensity and firm performance (Farris, Olver
and De Kluyver 1989). The present study aims to expand our understanding of
distribution intensity by looking at the construct within one product category and
testing the framework empirically.
6.2 Methodology
The framework developed was tested in the Hi-Fi speaker industry by
gathering data from manufacturers regarding their marketing and especially channel
strategy. A questionnaire was used to collect data on product positioning, target
focus for consumer segments, need for channel control, financial attractiveness of
the brand, dealer investment levels for carrying the brand, managerial support for
intermediaries, contractual restrictiveness, and relationship uncertainty for the
dealer. Given the nature of data required, close cooperation and high levels of trust
in the confidential treatment of provided information was necessary. The industry
is relatively small in regard to the number of manufacturers in the marketplace.
Both aspects combined made it very difficult to convince firms to participate in our
study. Initially, there seemed to be an unsurmountable lack of trust and interest in
partipation. Only personal communication in which any manufacturer question
could be answered and any concern addressed appeared to be having an effect on
135
firm willingness to participate. Consequently, the researcher tried to establish
manufacturer trust and commitment to participation with face-to-face meetings or
telephone conversations. The Consumer Electronics Show proved to be an almost
ideal place to establish the communication, because it had most of the
manufacturers in one place for several days.
As a result of the extreme efforts, 85 usable responses were generated as a
result of the study. The response rate approached 40% which is acceptable for
research of this kind.
6.3. Empirical Findings
The empirical findings of this study were generally supportive of the
conceptual framework developed. The hypotheses concerning the manufacturer
aspect of the framework were more strongly supported, while the hypotheses
concerning the intermediary aspect received only mixed support. The framework
part with hypotheses concerning product positioning and target focus influence on
need for control is not significant. The results are shown in Figure 6.1.
136
Figure 6.1: Empirical Results
H3 : (-) p < .066
H 1: (~)
p < .0001
Need fo r
Control
Product
Positioning
Distribution
Intensity
H7c: (-)
p <.118
(-)
Barriers to
Intermediary
Participation
H 8: (-)
p < .020
Management / \
Support
/ \
/ \
/
Investment
Level
/
Relationship Contractual
Uncertainty Restrictiveness
-J
H7b: (-) H7d: (+)
p < .035 p < .248
unexpected sign
Focus
H 2: (-)
p < .21
H 6: (+)
p<.019
Financial
Attractiveness
H7a: (-)
p < .244
unexpected sign
Overall Model:
F Value: 14.272
Prob > F: .0001
Adj. R square: .6207
Basically, high end product positioning and a high need for channel control
are significant factors influencing distribution intensity. The analysis shows an
inverse relationship with higher positioning or need for control associated with
lower distribution intensity. Target focus seems to be an insignificant factor, which
may be caused by industry specific factors for this study. Regarding the
intermediary aspect of the framework, it was shown that financial attractiveness is
a strong determinant for intensity. Barriers to entry on the other hand are generally
not supported as influential factors for distribution intensity. The sole exception is
managerial support provided by the manufacturer. Of course, this factor is more of
an enabler than a barrier to dealer participation.
There are two main results associated with this study. First, we have shown
that distribution intensity is a construct in itself and not just synonymous or only
driven by product positioning. Secondly we developed and showed the influence of
several strategic dimensions on distribution intensity.
6.4 Limitations
At least two important limitations of this research need to be addressed.
First, the issue of generalizability of results has to be discussed. It includes the
question about boundary conditions for this study. The second limitation is the
issue of construct validity.
138
This study was conducted at the manufacturer level in the Hi-Fi speaker
industry, which leads to two questions in regard to generalizability. First, are the
results generalizable with the industry and second, can the results be generalized
across product lines? The level of cooperation needed and the general knowledge
requirements on the side of the respondents meant that both aspects of the
framework were covered by manufacturer responses. While sampling
manufacturers for the manufacturer aspect of the framework is ideal, the
intermediary aspect should ideally have been covered by an intermediary sample.
The dealer sample was sacrificed in favor of manufacturer responses, based on
manufacturer willingness to participate and the level of knowledge in regard to
marketing and especially channel issues in their industry. This may seriously
impair the generalizability of the results. However, given the fact that many
dealers were not willing or able to provide accurate data on their part of the
framework, a focus on manufacturers as informers on the intermediary aspect of
the framework is preferable to leaving out the second part of the model. On the
other hand, pre-test interviews were employed to verify the compatibility and
congruence of manufacturer provided information with dealer perception.
The product line of Hi-Fi speakers used for this study is generally
considered a shopping good. Different product classes may show different results
as some of the factors in the present framework become more or less important.
On the other hand, there is reason to assume that the framework would generally
139
hold for an analysis within the product classes of convenience or specialty goods.
The framework tries to explain variance within a product class with brands ranging
from the lower end of the spectrum to the higher end. In as much as other product
classes have a similar positioning distribution of brands the framework should
hold.
The Hi-Fi speaker industry has some unique conditions which might make it
necessary to reexamine the framework under different circumstances. The number
of major competitors is very small with a large number of small manufacturers.
While the bigger firms have marketing strategies, some of the smaller suppliers act
on an opportunistic basis. Of course, the responses of these opportunistic
manufacturers must be considered carefully as they have the potential to introduce
bias to the sample. These small manufacturers may also be the reason for a major
portion of non-response. Another issue to be considered is the difficulty to obtain
data. In other environments where data is more easily obtained, the framework
should be tested from both sides, manufacturers as well as dealers. An analysis
based on both sides’ input will certainly clarify some of the mixed results. In either
case, the generalizability of the study needs to be examined further.
The second issue deals with construct validity, based on the operational
measures for the intermediary aspect of the framework. The present study had to
overcome the handicap of lack of intermediary participation. Therefore, some of
140
the measures for the intermediary aspect were summarized over the population of
dealers available to a manufacturer instead of focusing on individual dealers. Given
the character of the measures, their validity in correctly reflecting the situation can
be questioned and is reflected in the results of our empirical findings.
6.5 Future Research
For a complete framework of distribution intensity, the consumer aspect
cannot be left out. While this study represents an important first step towards the
goal of construct integration, the consumer aspect remains empirically untested.
The question lingers on, whether the conceptually intuitive categories of goods
framework and its implications for distribution intensity can be empirically
supported. In the same vein, an analysis of the framework across product classes
will be helpful on the road to establishing a comprehensive framework for the
construct.
Finally, the element of channel performance needs to be examined in the
context distribution intensity. Why do firms establish certain levels of distribution
intensity and what are the net costs and benefits of their respective strategies. The
study of this normative aspect of the framework would be of high value to both,
the academic field of channels research as well as the marketing practioners.
Although there is some conceptual work in this area, empirical investigations of
141
performance within product classes but between product lines and also across
product classes need to be conducted.
142
References
Aaker, David A. and J. Gary Shansby (1982), "Positioning Your Product,"
Business Horizons. 25 (May-June), 56-62.
Anderson, Erin and Barton Weitz (1986), Determinants of Continuity in
Conventional Industrial Channel Dvads. Working Paper 86-024, University of
Pennsylvania.
Anderson, James C. and David W. Gerbing (1988), "Structural Equation Modeling
in Practice: A Review and Recommended Two-Step Approach," Psychological
Bulletin. Vol. 103 (3), 411-23.
Aspinwall, Leo (1958), "The Characteristics of Goods and Parallel Systems," in
Managerial Marketing. Harold Kelly and William Lazer, eds., Homewood, IL:
Richard D. Irwin, 434-450.
Assael, Henry (1985), "Marketing Management: Strategy and Action. Boston,
MA: Kent Publishing Company.
Bagozzi, Richard P. (1986), Principles of Marketing Management. Science
Research Associates, Inc.
Balderston, Frederick E. (1958), "Communication Networks in Intermediate
Markets," Management Science. 4 (January) 159-63.
Baligh, Helmy H. and Leon E. Richartz (1964), "An Analysis of Vertical Market
Structures," Management Science. 10 (July), 667-689.
Bobrow, E.E. (1976), Marketing Through Manufacturer’s Agents. Sales Builders:
New York.
Bucklin, Louis (1963), "Retail Strategy and the Classification of Consumer
Goods," Journal of Marketing. 27 (January), 50-55.
_______ (1966), A Theory of Distribution Channel Structure. Berkeley: Institute of
Business and Economic Research.
_______ (1973), "A Theory of Channel Control," Journal of Marketing. 37
(January), 39-47.
143
Cespedes, Frank V. (1988), "Control versus Resources in Channel Design:
Distribution Differences in One Industry," Industrial Marketing Management. 17,
215-227.
Copeland, Melvin (1923), "Relation of Consumers’ Buying Habits to Marketing
Methods," Harvard Business Review. 1 (March-April).
Corey, E. Raymond, Frank V. Cespedes and V. Kasturi Rangan (1989), Going to
Market: Distribution Systems for Industrial Products. Boston, MA: Harvard
Business School Press.
Corstjens, Marcel and Peter Doyle (1979), "Channel Optimization in Complex
Marketing Systems," Management Science. 25 (October), 1014-1025.
Dickson, Peter and James Ginter (1987), "Market Segmentation, Product
Differentiation, and Marketing Strategy," Journal of Marketing. 51 (April), 1-10.
Doyle, Peter and John Saunders (1985), "Market Segmentation and Positioning in
Specialized Industrial Markets," Journal of Marketing. 49 (Spring), 24-32.
Dommermuth, William P. (1965), "The Shopping Matrix and Marketing Strategy,"
Journal of Marketing. 2 (May), 128-132.
______ and Edward W. Cundiff (1967), "Shopping Goods, Shopping Centers, and
Selling Strategies," Journal of Marketing. 31 (October), 32-36.
Farris, Paul, James Olver and Cornelius De Kluyver (1989), "The Relationship
between Distribution and Market Share," Marketing Science. 8 (Spring), 107-128.
Groeneveld, Leonard (1964), "A New Theory of Consumer Buying Intent,"
Journal of Marketing. 28 (July), 23-28.
Hardy Kenneth G. and Allan J. Magrath (1988), Marketing Channel Management.
Glenview, IL: Scott, Foresman and Company.
Hartung, Philip H. and James L. Fisher (1965), "Brand Switching and
Mathematical Programming in Market Expansion," Management Science. 11
(August), B231-B243.
Hauser, John (1986), "Theory and Application of Defensive Strategy," in The
Economics of Strategic Planning. L.G. Thomas, ed. Lexington, MA: Lexington
Books, 113-39.
144
Hlavacek, James D. and Tommy J. McCuistion (1983), "Industrial Distributors -
When, Who and How?" Harvard Business Review. 61 (March-April), 96-101.
Holbrook, Morris and John Howard (1977), "Frequently Purchased Nondurable
Goods and Services," in Selected Aspects of Consumer Behavior. Robert Ferber,
ed., Washington, DC: National Science Foundation, 189-222.
Holton, Richard H. (1958), "The Distinction between Convenience Goods,
Shopping Goods, and Specialty Goods," Journal of Marketing. 24 (July), 53-56.
Jeuland, Abel P. and Steven Shugan (1983), "Managing Channel Profits,"
Marketing Science. 2 (Summer), 239-272.
Kaish, Stanley (1967), "Cognitive Dissonance and the Classification of Consumer
Goods," Journal of Marketing. 31 (October) 28-31.
Kleimenhagen, Arno K. (1966), "Shopping, Specialty, or Convenience Goods?"
Journal of Retailing. 42 (Winter), 32-39.
Kotler, Philip (1989), Marketing Management. 7th Edition, Englewood Cliffs, NJ:
Prentice Hall.
______ and Gary Armstrong (1991), Principles of Marketing. 5th Edition,
Englewood Cliffs, NJ: Prentice Hall.
Lilien, Gary and Ambar G. Rao (1976), "A Model for Allocating Retail Outlet
Building Resources across Market Areas," Operations Research. 24 (January-
February), 1-14.
Little, Robert W. (1970), "The Marketing Channel: Who Should Lead This Extra-
Corporate Organization," Journal of Marketing. 34 (January), p.32.
Luck, David J. (1959), "On the Nature of Specialty Goods," Journal of Marketing.
25 (July), 61-66.
Lusch, Robert F. (1979), "Erase Distribution Channel from Your Vocabulary and
Add Marketing Channels," Marketing News. (July 27), p. 12.
Marcus, Burton, et al. (1975), Modern Marketing. New York, NY: Random
House.
Mason, Joseph B. and Morris Mayer (1972), "Empirical Observations of
Consumer Behavior," Journal of Retailing. 48 (Fall), 17-31.
145
Mayer, Morris L., Joseph B. Mason, and Morris Gee (1971), "A
Reconceptualization of Store Classification as Related to Retail Strategy
Formulation," Journal of Retailing. 47 (Fall), 27-36.
McCarthy, E. Jerome and William D. Perreault (1984), Basic Marketing.
Homewood, IL: Richard D. Irwin, Inc.
McGuire, T. W. and R. Staelin (1983), "An Industry Equilibrium Analysis of
Downstream Vertical Integration," Marketing Science. 2, 2, 161-191.
Miracle, Gordon E. (1965), "Product Characteristics and Marketing Strategy,"
Journal of Marketing. 29 (January), 18-24.
Murphy, Patrick and Ben Enis (1987), "Classifying Products Strategically,"
Journal of Marketing. 50 (July), 24-42.
Naert, Philippe A. and Alain V. Bultez (1975), "A Model of Distribution Network
Aggregate Performance," Management Science. 21 (June), 1102-1112
Park, C. Whan, Bernard J. Jaworski and Deborah J. Mclnnis (1986), "Strategic
Brand Concept-Image Management," Journal of Marketing. 50 (October), 135-145.
Pegram, Roger (1965), Selecting and Evaluating Distributors. New York, NY:
National Conference Board.
Porter, Michael (1980), Competitive Strategy. New York, NY: The Free Press.
________(1985), Competitive Advantage. New York, NY: The Free Press.
Rangan, V. Kasturi (1987), "The Channel Design Decision: A Model and an
Application," Marketing Science. 6 (Spring), 156-174.
______ , Andris A. Zoltners and Robert J. Becker (1986), "The Channel
Intermediary Selection Decision. A Model and an Application," Management
Science. 32 (September), 1114-1122.
Rosenbloom, Bert (1978), "Motivating Independent Distribution Channel
Members," Industrial Marketing Management. 7, 275-281.
________(1991), Marketing Channels. 4rd Edition, Chicago: The Dryden Press.
146
Rueckert, Robert W ., Orville C. Walker and Kenneth J. Roering (1985), "The
Organization of Marketing Activities: A Contingency Theory of Structure and
Performance," Journal of Marketing. 49 (Winter), 13-25.
Shipley, David D. (1984), "Selection and Motivation of Distribution
Intermediaries," Industrial Marketing Management. 13, 249-256.
Sibley, S. D., and Teas K. R. (1979), "The Manufacturer’s Agent in Industrial
Distribution," Industrial Marketing Management. 8, 286-292.
Skinner, Steven J. and Joseph P. Guiltinan (1985), "Perceptions of Channel
Control," Journal of Retailing. 61 (Winter), 65-88.
Stern, Louis and Adel El-Ansary (1988), Marketing Channels. 3rd Edition,
Englewood Cliffs, NJ: Prentice Hall.
Webster, Frederick E. (1976), "The Role of the Industrial Distributor in Marketing
Strategy," Journal of Marketing. 40 (July), 10-16.
_______ (1975), "Perceptions of the Industrial Distributor", Industrial Marketing
Management. 4, 257-264.
Zusman, Pinhas and Michael Etgar (1981), "The Marketing Channel as an
Equilibrium Set of Contracts," Management Science. 2 (Summer), 284-302.
147

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Cross-sectional tests of tax and regulatory costs of a change in depreciation methods in the railroad industry

PDF

A contingent claims analysis of the equity interest in property-casualty insurance companies

PDF

A quantitative model for the analysis of warranty policies

PDF

Communications medium and product class: Their effect on negotiation outcome

PDF

A biogeographic theory of industrial market structure and competitive dynamics

PDF

An examination of the effect of environmental cues on audit judgments

PDF

An investigation of information requirements determination and analogical problem solving

PDF

A study on the content and determinants of close exchange relationships in channels of distribution

PDF

Career advancement in managerial hierarchies in the United States firms: A multi-theoretic model and empirical tests of the determinants of managerial promotion

PDF

American motion picture distribution with an analysis of the independent distribution of "successful" American feature films: 1970-1983.

PDF

An empirical study of executive stock options and company performance

PDF

An analysis of the Nigerian economy within the framework of the wedge model

PDF

A theory of deterministic consumer choice behavior: Applying generalizability theory of measurement to consumer panel data

PDF

Assessing effectiveness in four corporate universities

PDF

Corporate Political Action Committee contribution tactics: Prime defense contractors, federal election cycle 1985-1986

PDF

Auditors' risk attitudes: A hierarchical levels study within various decision contexts

PDF

An investigation of the impact of congruent managerial values on appraisal behavior

PDF

A highly efficient and general method for the chemical synthesis of beta,gamma-fluoro- and difluoromethylene analogs of purine and pyrimidine 5'-ribo- and deoxyribonucleotides

PDF

A laboratory study of the effects of accounting valuation methods on manager performance under different inflation environments

PDF

A study of the English Canadian feature film industry 1977-1981

Asset Metadata

Creator Lassar, Walfried M (author)

Core Title A conceptual framework of distribution intensity

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Business Administration

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

Tag business administration, general,OAI-PMH Harvest

Language English

Advisor [illegible] (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-761615

Unique identifier UC11348186

Identifier DP22691.pdf (filename),usctheses-c17-761615 (legacy record id)

Legacy Identifier DP22691.pdf

Dmrecord 761615

Document Type Dissertation

Rights Lassar, Walfried M.

Type texts

Source 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 au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA