CHAPTER 48
PRODUCT C A N N I B A L I S M
Roger A. Kerin, PhD
Harold C. Simmons Distinguished Professor
ofMarketing
Edwin L. Cox School of Business
Southern Methodist University
Dallas. Texas
Dwight R. Riskey, PhD
Vice President, Marketing Research and
New Business
Frjto-Lay, Inc.
Piano, Texas
Slowed economic growth, compressed product life cycles, domestic and foreign competition, and consumer desire for variety have
placed unprecedented pressures on product management. Properly or
improperly, many firms appear to be focusing their efforts on product
opportunities that offer minimal market resistance, leverage existing
technologies, utilize present manufacturing capability and capacity, and
minimize financial investment. These opportunities are often manifested in product line extensions.
It is estimated that 81 percent of new consumer products are
extensions of existing products. No product category seems to be
immune from the surfeit of product offshoots. In 1990 alone, product
line extensions included 64 different sauces, 103 snack chips, 91 cold
remedies, and 69 disposable diapers.1 Line extensions are routinely
achieved in a number of ways.2 Introduction of new sizes (super
economy-size toothpaste), forms (deodorant in liquid, powder, spray,
and stick forms), compositions (regular Head & Shoulders, shampoo
and Head & Shoulders shampoo with conditioner), flavors (Jello
gelatin, which originally began with six flavors, is now available in
more than a dozen), packages (Hi-C fruit flavored drink in glass bottles
or paper or metal containers), and varieties (shampoo for dry, normal,
and oily hair) are just some of the feasible courses of action.
In general, dominant firms in a product category or industry tend
to view product line extension as a relatively low risk means of
maintaining or increasing market share, and as an indispensable tool in
the fierce struggle for retail shelf space. For example, an Association of
National Advertisers study indicated that the majority of line extensions
were considered successful. By comparison, only about one-half of new
products were deemed successful.3 On the other hand, low market share
firms and potential new entrants tend to view the line extension
strategies of dominant firms as a calculated strategy to corner supermarket shelf space, keep out rival brands, and protect their competitive
position.
Even though product-line extension strategies pose minimal risk
of failure for the product being introduced, potential negative effects on
existing products serving existing markets must be considered. These
effects can be called product .cannibalism. While some cannibalism
may be planned or expected, considerable amounts of cannibalism may
be an unexpected consequence of an improperly managed product
development process and line extension program. When AnheuserBusch launched Michelob Light, it fully expected that 20-25 percent
of the new brand would come from its existing base brand because
of its low-calorie appeal among current customers.4 However, despite
PRODUCT CANNIBALISM
MARKETING MANAGER'S HANDBOOK
numerous efforts to differentiate and position Stouffer Food Corporation's Right Course frozen entree, the brand cannibalized the company's
popular Lean Cuisine brand. Six months after the introduction of Right
Course, it was pulled because, according to a company official, "This
brand [Right Course] ended up competing with our Lean Cuisine line."
The purpose of this chapter is to examine several different facets
of product cannibalism. We begin by describing the dynamics underlying product cannibalism. Managerial practices and competitive conditions that foster cannibalism are then enumerated. Because cannibalism
is sometimes justified on the basis of competitive reality, the strategy of
preemptive cannibalism is introduced. The financial consequences of
product cannibalism are then illustrated from the perspective of individual products and product line profitability, and approaches for
identifying cannibalism potential are briefly described. The chapter
concludes with a discussion on the implications of product cannibalism
for product line management.
NATURE OF PRODUCT CANNIBALISM
The theoretical roots of product cannibalism can be traced to the
.theory of cross-elasticity of demand. This theory suggests that the
percentage change in the quantity of product A demanded will be
influenced by the percentage change in the price of product B. The
"demand interrelationship between two products may be described as
complementary or substitutable. For example, if the price of product B
decreases and the quantity demanded of product A increases, the two
products are considered to be complements, providing other things
remain equal. Common examples illustrating complementary products
are razors and razor blades and cameras, and photographic accessories.
In the case of product substitution, or cannibalism, a lowering of the
price on product A will tend to decrease the quantity demanded for
product B, providing other things remain equal. Margarine is sometimes viewed as a substitute for butter.
From a marketing standpoint, however, other things rarely remain
equal. Accordingly, an expanded interpretation of cross-elasticity of
demand is necessary. In addition to price changes, physical and
symbolic attributes of products, alternative means of promoting and
distributing products, and potential end-use interchangeability between
products must be considered. In addition, attention must be given to the
potential for variety-seeking by consumers. That is, for some product
categories (e.g., snack foods and soft drinks) consumers tend to
purchase and consume multiple flavors and sizes.
As shown in Exhibit 1, new entrants and line extensions acquire
their sales revenue from three sources: (1) new consumers who were
not previously buyers of the product type; (2) consumers of competitive
brands within a product category; and (3) consumers of an existing
EXHIBIT 1. Components of New Entrant Sales Revenue
Revenue Source
Revenue Type
Market Expansion
and Market Share
Gains
Incremental
Revenue
New Entrant
Sales Revenue
Product/Brand
Switching within the
Product Line
Redistributed
Revenue
company brand or product who switch to the new entry. The first two
sales revenue sources represent, respectively, incremental revenue for
the product line because of market expansion and the capturing of
competitors' market share or share-building. The remaining sales
revenue source represents redistributed revenue, or cannibalization, in
that existing buyers are substituting one item for another in the
company's product line. Accordingly, product cannibalization can be
succinctly defined as the process by which a firm's new entry gains a
portion of its sales by diverting them from a firm's existing product(s).
This process of sales diversion or redistribution of revenue has a
subtle but managerially important consequence. Assuming that the
change in profits earned by the existing product is reduced or negative
because of substitution, this amount should be added to the incremental
cost curve for the new offering. The implication is clear—sales and
profit gains of a new entry at the expense of an existing product do not
filter down to the bottom line. Rather, the loss of potential profits from
a cannibalized product is real cost that must be absorbed by the new
product. The adage "You can't have your cake and eat it, too" applies
when cannibalism occurs.
883
MARKETING MANAGER'S HANDBOOK
As a generalization, it has been suggested that a product-line
extension must gain two market share points for each share point lost
by cannibalized products to preserve profit margins. One-half of the
gain will come from users of competitors' products and one-half from
new users drawn into the category. Therefore, sales revenue for a
product-line extension is divided equally among incremental sales
volume, cannibalized revenue, and market share gains.6
FOSTERING CANNIBALISM
The erosion of an existing product's share of market resulting
from cannibalism may stem from,management decisions, or it may be a
necessary evil, given competitive conditions. Cannibalism becomes a
severe problem when it provides no incremental competitive or financial benefit to the firm's product line. Several managerial decisions
appear to foster cannibalism of an existing brand's sales volume:
1. management pressure for sales growth from new offerings without
full consideration of financial consequences;
2. preoccupation with developing a full line of products in an attempt
to achieve increases in overall market share in a product class or
category;
3. inadequate positioning of new entries resulting in their seeking the
identity of existing products;
4. unrealistic or excessive market segmentation resulting in "two
segments" with demands for identical product attributes or end-use
needs; and
5. aggressive promotional effort or incentives reflected in sales representatives' overemphasis on new items and neglect of existing
products.
j- Product cannibalism by itself should not be viewed only negatively. Cannibalism sometimes represents an outgrowth of effective and
competitive product management. For example, some firms might
actively pursue line extensions for the purpose of building a base
brand's "brand equity." This rationale has been used by some soft drink
producers to justify cannibalism of regular soft drinks by diet and
caffeine-free varieties. Even though diet and caffeine-free beverages
divert sales from regular drinks, the equity of the base brand name is
believed to be enhanced even though cannibalism among line items is
present. Additionally, a new item with cannibalism potential may be
introduced to eliminate gaps in a product line that might be filled by
competing offerings or to neutralize competitive inroads. In other
words, it may be wiser to have buyers switching among products and
brands within a firm's product line than to have them switching out and
purchasing competitive offerings. Viewed in this manner, preemptive
cannibalism becomes a viable business strategy.
884
PRODUCT CANNIBALISM
Preemptive cannibalism is the conscious practice of stealing sales
from a company's existing products and brands to keep consumers from
switching to competitors' offerings. Two product-line extension strategies are commonly employed in this regard: a flanker brand strategy
and a fighting brand strategy.
Flanker Brand Strategy
Bristol-Meyers' introduction of Datril to compete with McNeil
Laboratories' Tylenol exemplifies a flanker brand strategy. BristolMeyers held a position of strength in the aspirin segment of the
analgesic market with its Bufferin and Excedrin brands. However, this
segment, while large, had plateaued in the mid-1970s. During the same
period, the acetaminophen (noninflammatory compounds) segment of
the analgesic market dominated by McNeil Laboratories had grown
substantially, with a portion of the growth coming from former and
potential aspirin users. Datril's introduction would hopefully attract
aspirin switchers (that is, switching away from company brands) and
tap existing and potential acetaminophen buyers who would most likely
purchase Tylenol. Therefore, even though Datril might cannibalize
Excedrin and Bufferin, aspirin switchers would remain in the BristolMeyers analgesic product line rather than being attracted to Tylenol.
A similar strategy was recently employed by the Miller Brewing
Company with the introduction of Lite Genuine Draft, an extension of
Miller's Genuine Draft beer. The company intended to have Lite
Genuine Draft compete against Coors Lite and Budweiser Light in the
premium segment of the light beer category since Miller Brewing
believed Miller Lite users were switching to Coors and Budweiser
brands. Even though it was likely that Lite Genuine Draft would steal
sales from Miller's Genuine Draft and also Miller Lite, Leonard
Goldman, president of Miller Brewing Company said:
Miller Lite represents a tremendous presence in the fastest growing
segment (light beer), but we are committed to having multiple brands in
each segment to give the consumer the broadest array of choices. This is
a fancy way of saying that if the consumer is going to defect, we want
him to defect to another Miller product. "7
Fighting Brand Strategy
A fighting brand is typically introduced when (1) a firm has a
high relative share in a product category; (2) its dominant brands are
susceptible to having share sliced away by aggressive pricing by
smaller competitors; and (3) it wishes to preserve its profit margins on
existing brands. The introduction of Santitas® brand Tortilla Chips by
Frito-Lay illustrates a fighting brand strategy. As the tortilla chip
category leader, managers for Tostitos® brand Tortilla Chips recognized
the emergence of restaurant-style tortilla chips (RSTC) as a competitive
threat in certain regional markets. Unflavored and typically priced
below national and regional tortilla chip brands, RSTC were also lower
in quality based on consumer taste tests. Nevertheless, lower pricing
coupled with their primary use for dipping with sauces and nacho
preparation resulted in RSTC capturing a sizable portion of tortilla chip
volume. Frito-Lay introduced Santitas® brand Tortilla Chips at a
competitive RSTC price and provided trade promotion to support the
brand. This action, while cannibalizing a portion of Tostitos® brand
Tortilla Chips, resulted in incremental volume growth and category
share-building for Frito-Lay.
ACCOUNTING FOR PRODUCT CANNIBALISM
The previous discussion illustrates the importance of performing
a marginal analysis on the new item in the product line within the
context of present and forecasted market conditions. Incremental
revenue, cost, and investment must be considered.
An analysis of a hypothetical multiproduct firm serves to demonstrate the financial consequences of product cannibalism. This firm has
an existing product that was expected to capture 5 percent of a market
forecasted at 15 million units, or 750,000 units. At a $2.00 per unit
price and a $1.00 per unit gross margin, forecasted sales are $1.5
million with a $750,000 gross margin. Budgeted marketing expenditures plus allocated overhead total $300,000, which will provide a
$450,000 profit before taxes and a 10 percent return on investment. An
abbreviated pro forma income statement showing these figures is shown
in part A of Exhibit 2.
A new offering is introduced that satisfies several, but not all,
buyer requirements met by the existing product in addition to providing
several other benefits. The new item is value-priced at $1.50 per unit
with a $0.75 per unit gross margin. The lower price and modified
product benefits are expected to expand the market for this product type
by 20 percent to 18 million units. Both products combined are expected
to capture 10 percent of the expanded market, or 1.8 million units,
which represents a 240 percent increase over forecasted volume for the
single existing product! Marketing expenditures plus allocated overhead
for the new product are budgeted at $450,000. Incremental investment
for the new offering is $1 million. Most of the volume captured by the
new product comes from market expansion because of the lower price
and differentiated product benefit structure. However, slightly more
than 25 percent cannibalism rate occurs from the existing product.
Part B in Exhibit 2 shows the effects of the activities and events
described for the existing product and the new item, individually and
combined. Also shown is an incremental analysis comparing the
existing product alone versus the existing and the new product combined. Given the conditions of the example cited, the apparent new item
886
S „
1
II
O.
88
K
CO
CN
rx O
co 10
-« K
£*
5?
— "-o "l<
v» K v> Is. <
8
88
O» •— CM
CO
8
CO I
O. •
£888
8
8 8:
f O Q
*O O
O O
IO O
s s »»„-
8
fi»
O
*O
s
888:
~8'
"1141«
iisilL.lii
2O
1/1
_S "" ~Z £ F
T!
§
MARKETING MANAGER'S HANDBOOK
profit is much less when cannibalized volume is subtracted from the
existing product's contribution. The apparent return on investment for
the product line with the new offering is inflated; it is actually less than
the return on investment for the existing product when cannibalized
volume is taken into account. Finally, the incremental analysis reveals
that the incremental profit from the new entry is only 30 percent of
what it appears to be without consideration of cannibalism's affects.
This example highlights several possible ramifications of product
cannibalism:
1. Without accounting for product cannibalism, sales volume and
profits for a new item may be more illusionary than real.
2. New-entries examined in an isolated fashion—without also considering cannibalized volume—provides a distorted view of productline profits and return on investment.
3. Market share growth for a product line resulting from a new entry
may represent Pyrrhic victories in terms of product-line profitability
and individual item volume.
4. Both the amount and source of potential new item volume must be
considered in product-line planning to calculate the impact of
1
cannibalism on product-line profitability.
Effects of Preemptive Cannibalism
This analysis can also be used to illustrate the potential effects of
preemptive cannibalism. Suppose in our hypothetical situation dial the
new item described was used as a retaliatory device (fighting brand) to
meet a competitor whose lower-priced product was capturing a portion
of the existing product's market share. If one considers the new item's
cannibalized volume as potentially lost to the competitor, then these
buyer> are being kept by the firm, albeit at a lower rate of return. If the
new item were not introduced, 200,000 units would be lost, resulting in
a five percent return on the existing product's investment. Even with
cannibalism considered, the firm's new product will virtually preserve
the return on investment percentage, thus showing the benefit of
preemptive cannibalism.
Incremental Analysis
It is also possible to calculate the incremental unit volume
necessary to overcome the effects of cannibalism. This measure can be
used as a benchmark for evaluating market capacity and the quality of
introductory marketing programs early in the business analysis stage.
The expression is as follows:
Incremental volume
Ratio of the
to offset
: Cannibalized x old and new
unit volume
cannibalism effect
product margins
888
Using figures from the hypothetical example, the incremental
volume necessary to overcome the effects of cannibalism is approximately 267,000 units: (200,000 units) x ($1.00 margin/$0.75 margin).
In other words, at the estimated cannibalism rate, the new item must
generate an incremental volume from new and competitors' customers
of 267,000 units to offset the loss of margin dollars from the existing
product. In effect, for this illustration, a 21 percent increase in
incremental volume over forecasted levels would be required. Issues
surrounding market capacity and the quality of the market entry
program assume a different light in this context.
In some instances, the dollar gross margin of the new item might
exceed the dollar gross margin of the existing product that it cannibalizes. In this instance, for each unit of the existing product that is
cannibalized, the firm records a higher dollar gross margin. Therefore,
the firm would benefit from product cannibalism. It should be emphasized, however, that the incremental costs associated with advertising
and promotion or any additions in manufacturing capacity costs must
be considered to determine the net effect of cannibalism. If these
incremental costs exceed the incremental benefits of higher dollar gross
margin, then product-line profitability will suffer.
IDENTIFYING CANNIBALISM POTENTIAL
The importance of identifying cannibalism potential cannot be
overemphasized. Cannibalism effects should be considered in light of
product-market structure and throughout the product development
process, beginning with concept testing and continuing through commercialization.
Analysis of Product-Market Structure
\ . Conventional approaches for defining product-market structure
focuses on product segmentation. Exhibit 3 illustrates this approach for
chip and vegetable dips sold through supermarkets. As shown, dips
divide into two distinct categories. Prepared dips are ready-made and
divided into two categories: (1) refrigerated dips such as Marie's and
(2) shelf-stable dips such as Frito-Lay's Dips® which are sold in metal
cans and require no refrigeration. Dip bases, which require at-home
preparation, also divide into two categories: (1) dry soup mixes and (2)
dry dip mixes. The implication of this structure is that prepared dips
(refrigerated and shelf-stable) compete with each other, but not with dip
bases. However, this conclusion might be incorrect when usage situations are considered.
Recent research suggests that attention to usage context offers a
broadened perspective on how consumers view. products and their
substitutability, hence cannibalism potential. The substitution-in-use
approach (SIU) simultaneously examines products and their usages to
889
MARKETING MANAGER'S HANDBOOK
EXHIBIT 3. Product Segmentation of Vegetable and Chip Dips Sold Through Supermarkets
reveal product-market structure. A key assumption underlying this
approach is that products represent "benefit bundles" and when
grouped on the basis of appropriateness for similar usages are presumed
to deliver similar benefit combinations. The SIU approach implies that
products perceived to be similar by consumers are considered substitutable as a means for the same ends or usages. Therefore, prepared
refrigerated dips could be substitutes for dry dip mixes prepared at
home for specific usage situations, e.g., parties. Accordingly, a dry dip
mix line extension by a refrigerated dip producer could cannibalize the
firm's existing product(s), at least for specific usage contexts.
Applications of the SIU approach result in maps, dendrograms,
and other visual depictions that attempt to portray product-market
structure in a manner corresponding to consumers' mental representations of a product category (e.g., vegetable and chip dips). The
mechanics of this approach are beyond the scope of this discussion.8
Nevertheless, by directing managerial attention toward consumer perceptions of products as a means to achieving an end inherent in a usage
context, the SIU approach provides useful insights into product-market
structure, product similarities, and product substitution potential.
890
PRODUCT CANNIBALISM
Analysis of Product Features, Preferences, and Perceptions
Substitution-in-use analysis offers a useful, although coarse-grain
assessment of product comparability in usage contexts. This approach
has a "black-box" character in that it does not measure directly the
product features sought in a specific usage context nor consumer
perceptions of the benefits afforded by individual products. Conjoint
analysis is typically employed to study the linkage of product features
to preferences and the linkage of features to consumer perceptions. This
analysis technique is of special value in concept testing where its
application is most prevalent.9
Conjoint analysis, which actually represents a family of multiattribute choice models, concerns itself with measuring the joint effect of
two or more independent variables (e.g., product features) on the
ordering of a dependent variable (e.g., preference). Conjoint analysis
makes it possible to take a consumer's overall evaluation of a set of
concepts and decompose those evaluations into separate scores—utility
functions—for the various attributes or features of concepts that led to
the consumer evaluating them the way he/she did. Consumer choices
based on their individual utility functions can result in estimated share
of choices and brand-switching matrices showing the (company or
competitive) brands from which any of a number of new concepts (i.e.,
line extensions) can draw customers. For example, Sunbeam Appliance
Corporation routinely uses conjoint analysis to assess new model
concepts for its small appliance line both in terms of draw from
competitive brands
(e.g., Sears, Cuisinart) as well as cannibalization of
its own models.10
Conjoint analysis also can be used as a tool for product positioning within a product line. For example, applications are possible that
can provide insight into the value of a brand name (when brand names
1
are treated as independent variables) as well as the appropriateness of
various brands for providing certain types of features and benefits. The
latter application treats brands as dependent variables and combinations
of product features are sorted according to their similarity to or
appropriateness for various brands.
An extended description of the statistical properties and application of conjoint analysis is beyond the scope of this discussion.11
Nevertheless, it is worth noting that as a concept-testing practice,
conjoint analysis does show promise in its ability to predict actual
consumer choice and produce valid market share estimates for new
concepts.12 Since actual product sales and market share are often
confounded by marketing mix variables such as advertising and
distribution and competitive behavior, conjoint analysis as a predictive
technique for assessing sources of volume for line extensions—hence,
product cannibalism—is not a substitute for pretest market modeling
and actual field market tests.
PRODUCT CANNIBALISM
Pretest Market Modeling and Field Test Marketing
Cannibalism potential can be further identified and possibly
refined beyond the concept-testing stage. As product ideas evolve from
concepts to prototypes, segmentation and positioning strategies become
clearer, introductory marketing programs can be drafted, and spending
levels can be finalized. At this time, cannibalism potential can be
assessed within the context of pretest market models. Pretest market
models combine executive judgment on marketing strategy, historical
data, laboratory data using product preference, and trial and repeat
purchase models of buyer behavior to make sales and market share
projections for new concepts.13 Most notably, pretest market models
typically incorporate cannibalization estimates that are often based on
data gathered from simulated market tests (or laboratory test markets)
and preference modeling (e.g., conjoint analysis).
Despite the benefits of pretest market models, they are not
without limitations. Their application is limited to studying new brands
seeking to enter well-defined product categories. They are not particularly suitable for products with long purchase cycles or products whose
benefits are not immediately realized. Consequently, primary use of
these models exists in evaluating consumer package goods and food
products.. Furthermore, these models are limited to the analysis of
individual products; they do not provide information on product line
dynamics per se.
Estimating a new offering's sales volume potential and its
tendency to cannibalize existing products can be finally assessed by
field test markets. Test markets allow product sales and product item
interactions to be observed in an actual—although not always
representative—competitive setting. Even in a test market setting,
however, assessing cannibalism is not a trivial exercise. Although it is
possible to perform product-by-product sales trend analysis to infer
levels of cannibalism, trend analysis only uncovers cannibalism effects
when they are large. Even then, alternative explanations for an existing
product's sales decline can be invoked.
The advent of controlled (scanner) test markets for consumer
package goods sold using household panels offers considerable promise
in assessing the sources and magnitude of product cannibalism. For
example, algorithms for studying household panel data have been
developed for analyzing a household's history of purchases before and
after the introduction of a new item. These algorithms have proven
useful in studying sources of volume for existing products and line
extensions. Analysis of household panel data reveals that cannibalism
rates ranging from 60 to 70 percent are common for minor line
extensions to 10 to 20 percent or less for new offerings with significant,
previously unaddressed consumer benefits. A new flavor of potato chip,
for instance, typically generates 50 percent incremental volume for its
892
base brand. An altogether new and unique snack chip brand may
produce incremental sales volume in the range of 75 to 90 percent.
IMPLICATIONS FOR PRODUCT-LINE MANAGEMENT
The popularity of product-line extensions as a strategy of choice
by many firms raises the spectre of product cannibalism. The previous
discussion suggests four implications of product cannibalism for
product-line management.
First, with companies coming under greater pressure to add new
items to their lines, the need to systematically assess the incremental
sales and profit contribution of product additions is becoming ever
more critical. Specifically, it is important to quantify incremental
product-line (versus single-product) profits before deciding whether to
add a new item to an existing product line. Such incremental analysis
requires an understanding of product cannibalism.
Second, preemptive cannibalism is frequently necessary to confront competitor product development efforts. Nevertheless, productline filling and stretching efforts to meet competitive inroads need to be
assessed by comparing cannibalized sales volume with projected lost
volume. If cannibalized volume is substantially larger than projected
lost volume, preemptive product cannibalism may not be justified.
Third, understanding demand interdependencies is a necessary
first step in assessing cannibalism potential. Traditional notions of crosselasticity of demand based solely on price are inadequate. Numerous
other factors and conditions result in product substitutions. In this regard,
effective product-line management must consider a new offering's
"substitution-in-use" possibilities and hence its cannibalism potential.
Finally, attention to the sources and magnitude of cannibalism
produced by new offerings should be studied throughout the product
1
development process from concept through commercialization. To
ignore product cannibalism is to invite unwise product proliferation and
its frequent derivative—product cannibalism.
NOTES
1. "New-Product Troubles Have Firms Cutting Back," Wall Street Journal
(January 13, 1992), p. Bl; "Multiple Varieties of Established Brands Muddle
Consumers, Make Retailers, Mad," Wall Street Journal (January 24 1992) pn
Bl, B5.
2. Roger A. Kerin, Vijay Mahajan, and P. Rajan Varadarajan, Contemporary
Perspectives on Strategic Market Planning (Boston: Allyn & Bacon Inc
1990), pp. 242-243.
3. Prescription for New Product Success (New York: Association of National
Advertisers, 1984).
4. "Anheuser-Busch, Inc., Has Another Entry in 'Light Beer' Field," Wall
Street Journal (February 19, 1978), p. 4.
USER'S HANDBOOK
5. Judann Dagnoli, "How Stouffer's Right Course Veered Off Course,"
Advertising Age (May 6, 1991), p. 34.
6. Douglas J. Dalrymple and Leonard J. Parsons, Marketing ManagementStrategy and Cases, 4th ed. (New York: John Wiley & Sons, Inc., 1986), pp.
381-82.
7. "Miller Brewing to Test-Market New Light Beer," Wall Street Journal
(April 20, 1988), p. 24.
8. For a recent description of the substitution-in-use approach, see S. Ratneshwar and Allan D. Shocker, "Substitution in Use and the Role of Usage Context
in Product Category Structure," Journal of Marketing Research, (August 1991),
pp. 281-295.
9. Dick Wittink and Philippe Cattin, "Commercial Use of Conjoint Analysis:
An Update," Journal of Marketing (July 1989), pp. 91-96.
10. Albert L. Page and Harold F. Rosenbaum, "Redesigning Product Lines
with Conjoint Analysis: How Sunbeam Does It," Journal of Product Innovation
Management, 1987, pp. 120-137.
11. For further reading on conjoint analysis and possible applications, see Paul
E. Green, Donald S. Tull, and Gerald Albaum, Research for Marketing
Decisions, 5th ed. (Englewood Cliffs, NJ: Prentice-Hall, 1988).
12. Paul E. Green and V. Srinivasan, "Conjoint Analysis in Marketing
Research: New Developments and Directions," Journal of Marketing (October
1990), pp. 3-19.
13. For an extended discussion on pretest market models, see Allan D. Shocker
and William G. Hall, "Pretest Market Models: A Critical Evaluation," Journal
of Product Innovation Management, 1986, pp. 86-107.