MARKETING

New Criteria for Market Segmentation

Daniel Yankelovich
FROM THE MARCH 1964 ISSUE

The director of marketing in a large company is confronted by some of the most difficult
problems in the history of U.S. industry. To assist him, the information revolution of the past
decade puts at his disposal a vast array of techniques, facts, and figures. But without a way to
master this information, he can easily be overwhelmed by the reports that flow in to him
incessantly from marketing research, economic forecasts, cost analyses, and sales breakdowns.
He must have more than mere access to mountains of data. He must himself bring to bear a
method of analysis that cuts through the detail to focus sharply on new opportunities.

In this article, I shall propose such a method. It is called segmentation analysis. It is based on the
proposition that once you discover the most useful ways of segmenting a market, you have
produced the beginnings of a sound marketing strategy.

Unique Advantages

Segmentation analysis has developed out of several key premises:

 In today’s economy, each brand appears to sell effectively to only certain segments of any
market and not to the whole market.
 Sound marketing objectives depend on knowledge of how segments which produce the most
customers for a company’s brands differ in requirements and susceptibilities from the
segments which produce the largest number of customers for competitive brands.
 Traditional demographic methods of market segmentation do not usually provide this
knowledge. Analyses of market segments by age, sex, geography, and income level are not
likely to provide as much direction for marketing strategy as management requires.

Once the marketing director does discover the most pragmatically useful way of segmenting
his market, it becomes a new standard for almost all his evaluations. He will use it to appraise
competitive strengths and vulnerabilities, to plan his product line, to determine his advertising
and selling strategy, and to set precise marketing objectives against which performance can
later be measured. Specifically, segmentation analysis helps him to—

…direct the appropriate amounts of promotional attention and money to the most potentially
profitable segments of his market;

…design a product line that truly parallels the demands of the market instead of one that bulks
in some areas and ignores or scants other potentially quite profitable segments;

1
…catch the first sign of a major trend in a swiftly changing market and thus give him time to
prepare to take advantage of it;

…determine the appeals that will be most effective in his company’s advertising; and, where
several different appeals are significantly effective, quantify the segments of the market
responsive to each;

…choose advertising media more wisely and determine the proportion of budget that should
be allocated to each medium in the light of anticipated impact;

…correct the timing of advertising and promotional efforts so that they are massed in the
weeks, months, and seasons when selling resistance is least and responsiveness is likely to be at
its maximum;

…understand otherwise seemingly meaningless demographic market information and apply it

in scores of new and effective ways.

These advantages hold in the case of both packaged goods and hard goods, and for
commercial and industrial products as well as consumer products.

Guides to strategy

Segmentation analysis cuts through the data facing a marketing director when he tries to set
targets based on markets as a whole, or when he relies primarily on demographic breakdowns.
It is a systematic approach that permits the marketing planner to pick the strategically most
important segmentations and then to design brands, products, packages, communications, and
marketing strategies around them. It infinitely simplifies the setting of objectives.

In the following sections we shall consider nondemographic ways of segmenting markets.

These ways dramatize the point that finding marketing opportunities by depending solely on
demographic breakdowns is like trying to win a national election by relying only on the
information in a census. A modern census contains useful data, but it identifies neither the
crucial issues of an election, nor those groups whose voting habits are still fluid, nor the needs,
values, and attitudes that influence how those groups will vote. This kind of information,
rather than census-type data, is the kind that wins elections—and markets.

Consider, for example, companies like Procter & Gamble, General Motors, or American
Tobacco, whose multiple brands sell against one another and must, every day, win new
elections in the marketplace:

These companies sell to the whole market, not by offering one brand that appeals to all people,
but by covering the different segments with multiple brands. How can they prevent these
brands from cannibalizing each other? How can they avoid surrendering opportunities to
competitors by failing to provide brands that appeal to all important segments? In neither
automobiles, soaps, nor cigarettes do demographic analyses reveal to the manufacturer what
products to make or what products to sell to what segments of the market. Obviously, some
modes of segmentation other than demographic are needed to explain why brands which differ

2
so little nevertheless find their own niches in the market, each one appealing to a different
segment.

The point at issue is not that demographic segmentation should be disregarded, but rather that
it should be regarded as only one among many possible ways of analyzing markets. In fact, the
key requirement of segmentation analysis is that the marketing director should never assume in
advance that any one method of segmentation is the best. His first job should be to muster all
probable segmentation and then choose the most meaningful ones to work with. This approach
is analogous to that used in research in the physical sciences, where the hypothesis that best
seems to explain the phenomena under investigation is the one chosen for working purposes.

Ten Markets

In the following discussion we shall take ten markets for consumer and industrial products and
see how they are affected by seven different modes of nondemographic segmentation. The
products and modes are shown schematically in Exhibit I. Of course, these segments are not
the only ones important in business. The seven I have picked are only examples of how
segmentation analysis can enlarge the scope and depth of a marketer’s thinking.

Exhibit I. Example of Segmentation in Different Industries

I. Watches

In this first case we deal with a relatively simple mode of segmentation analysis. The most
productive way of analyzing the market for watches turns out to be segmentation by value. This
approach discloses three distinct segments, each representing a different value attributed to
watches by each of three different groups of consumers:

3
1. People who want to pay the lowest possible price for any watch that works reasonably well. If the watch
fails after six months or a year, they will throw it out and replace it.

2. People who value watches for their long life, good workmanship, good material, and good styling. They are
willing to pay for these product qualities.

3. People who look not only for useful product features but also for meaningful emotional qualities. The most
important consideration in this segment is that the watch should suitably symbolize an
important occasion. Consequently, fine styling, a well-known brand name, the
recommendation of the jeweler, and a gold or diamond case are highly valued.

In 1962, my research shows, the watch market divided quantitatively as follows:

 Approximately 23% of the buyers bought for lowest price (value segment #1).
 Another 46% bought for durability and general product quality (value segment #2).
 And 31% bought watches as symbols of some important occasion (value segment #3).

Defining and quantifying such segments is helpful in marketing planning—especially if a watch

company’s product happens to appeal mostly to one segment or if the line straddles the three
segments, failing to appeal effectively to any. Without such an understanding, the demographic
characteristics of the market are most confusing. It turns out, for example, that the most
expensive watches are being bought by people with both the highest and the lowest incomes.
On the other hand, some upper income consumers are no longer buying costly watches, but
are buying cheap, well-styled watches to throw away when they require servicing. Other upper
income consumers, however, continue to buy fine, expensive watches for suitable occasions.

Timex’s Timely Tactics. The planning implications in value segmentation are very broad for the
industry. For one thing, many of the better watch companies in the years between 1957 and
1962 were inadvertently focusing exclusively on the third segment described—the 31% of the
market that bought a watch only as a gift on important occasions—thus leaving the bulk of the
market open to attack and exploitation.

The U.S. Time Company took advantage of this opening and established a very strong position
among the more than two-thirds of America’s watch buyers in the first two segments. Its new
low-price watch, the Timex, had obvious appeal for the first segment, and it catered to the
second segment as well. At that time, higher-price watches were making the disastrous mistake
in their advertising of equating product quality with water-proof and shock-resistant features.
The Timex also offered these low-cost features, at lower prices, thus striking at a vulnerable
area which the competition itself created. When Timex pressed its attack, it was able within a
few years to claim that “Timex sells more watches than any other watch company in the
world.”

4
Even the timing of Timex’s watch advertising was involved. Much of the third segment was
buying watches only during the Christmas season, and so most of Timex’s competitors
concentrated their advertising in November and December. But since buying by the other two
segments went on all the time, Timex advertised all year-round, getting exclusive attention ten
months of the year.

Thus, nondemographic segmentation in the watch industry has directly affected almost every
phase of marketing, including the composition of the product line. Major watch companies
know that they must plan product line, pricing, advertising, and distribution within the
framework of the three basic value segments of this market.

II. Automobiles

The nondemographic segmentation of the automobile market is more complex than that of
the watch market. The segments crisscross, forming intricate patterns. Their dynamics must be
seen clearly before automobile sales can be understood.

Segmentation analysis leads to at least three different ways of classifying the automobile market
along nondemographic lines, all of which are important to marketing planning.

Value Segmentation. The first mode of segmentation can be compared to that in the watch
market—a threefold division along lines which represent how different people look at the
meaning of value in an automobile:

1. People who buy cars primarily for economy. Many of these become owners of the Falcon, Ford,
Rambler American, and Chevrolet. They are less loyal to any make than the other segments,
but go where the biggest savings are to be found.

2. People who want to buy the best product they can find for their money. These prospects emphasize
values such as body quality, reliability, durability, economy of operation, and ease of upkeep.
Rambler and Volkswagen have been successful because so many people in this segment were
dissatisfied.

3. People interested in “personal enhancement” (a more accurate description than “prestige”). A handsomely
styled Pontiac or Thunderbird does a great deal for the owner’s ego, even though the car may
not serve as a status symbol. Although the value of an automobile as a status symbol has
declined, the personal satisfaction in owning a fine car has not lessened for this segment of the
market. It is interesting that while both watches and cars have declined in status value, they
have retained self-enhancement value for large portions of the market.

Markets can change so swiftly, and the size of key segments can shift so rapidly, that great
sensitivity is required to catch a trend in time to capitalize on it. In the automobile market, the
biggest change in recent years has been the growth in segment two—the number of people
oriented to strict product value. Only a few years ago, the bulk of the market was made up of
the other segments, but now the product-value segment is probably the largest. Some
automobile companies did not respond to this shift in the size of these market segments in
time to maintain their share of the market.

5
Aesthetic Concepts. A second way of segmenting the automobile market is by differences
in style preferences. For example, most automobile buyers tell you that they like “expensive
looking” cars. To some people, however, “expensive looking” means a great deal of chrome
and ornamentation, while to others it means the very opposite—clean, conservative lines,
lacking much chrome or ornamentation.

Unfortunately, the same words are used by consumers to describe diametrically opposed style
concepts. Data that quantify buyers according to their aesthetic responses—their differing
conceptions of what constitutes a good-looking car—are among the most useful an
automobile company can possess.

The importance of aesthetic segmentation can be pointed up by this example:

When Ford changed from its 1959 styling to its 1960 styling, the change did not seem to be a
radical one from the viewpoint of formal design. But, because it ran contrary to the special
style expectations of a large group of loyal Ford buyers, it constituted a dramatic and
unwelcome change to them. This essential segment was not prepared for the change, and the
results were apparent in sales.

Susceptibility to Change. A third and indispensable method of segmenting the automobile market
cuts across the lines drawn by the other two modes of segmentation analysis. This involves
measuring the relative susceptibility of potential car buyers to changing their choice of make.
Consider the buyers of Chevrolet during any one year from the point of view of a competitor:

 At one extreme are people whose brand loyalty is so solidly entrenched that no competitor can
get home to them. They always buy Chevrolets. They are closed off to change.
 At the other extreme are the open-minded and the unprejudiced buyers. They happened to buy
a Chevrolet because they preferred its styling that year, or because they got a good buy, or
because someone talked up the Fisher body to them. They could just as easily have purchased
another make.
 In the middle of this susceptibility continuum are people who are predisposed to Chevrolet to
a greater or lesser degree. They can be persuaded to buy another make, but the persuasion has
to be strong enough to break through the Chevrolet predisposition.

The implications of this kind of a susceptibility segmentation are far-reaching. Advertising

effectiveness, for example, must be measured against each susceptibility segment, not against
the market as a whole. Competitors’ advertising should appear in media most likely to break
through the Chevrolet predisposition of the middle group. In addition, the wants of those who
are not susceptible must be factored out, or they will muddy the picture. Marketing programs
persuasive enough to influence the uncommitted may make no difference at all to the single
largest group—those who are predisposed to Chevrolet but still open enough to respond to
the right stimulus.

If the marketing director of an automobile company does not break down his potential market
into segments representing key differences in susceptibility, or does not clearly understand the
requirements of each key segment, his company can persevere for years with little or no results
because its promotion programs are inadvertently being aimed at the wrong people.

6
III. Perfume

A segmentation analysis of the perfume market shows that a useful way to analyze it is by the
different purposes women have in mind when they buy perfume.

One segment of the market thinks of a perfume as something to be added to what nature has
supplied. Another segment believes that the purpose of fragrance products is to help a woman
feel cleaner, fresher, and better groomed—to correct or negate what nature has supplied. In
the latter instance, the fragrance product is used to cancel out natural body odors; in the former,
to add a new scent. To illustrate this difference in point of view:

 One woman told an interviewer, “I like a woodsy scent like Fabergé. It seems more intense and
lingers longer, and doesn’t fade away like the sweeter scents.”
 But another woman said, “I literally loathe Fabergé. It makes me think of a streetcar full of
women coming home from work who haven’t bathed.”

These differences in reaction do not indicate objective differences in the scent of Fabergé.
They are subjective differences in women’s attitudes; they grow out of each woman’s purpose
in using a perfume.

Purposive segmentation, as this third mode of analysis might be called, has been of great value
to alert marketers. For instance:

A company making a famous line of fragrance products realized that it was selling almost
exclusively to a single segment, although it had believed it was competing in the whole market.
Management had been misled by its marketing research, which had consistently shown no
differences in the demographic characteristics of women buying the company’s products and
women buying competitors’ products.

In the light of this insight, the company decided to allocate certain lines to the underdeveloped
segments of the market. This required appropriate changes in the scent of the product and in
its package design. A special advertising strategy was also developed, involving a different copy
approach for each product line aimed at each segment.

In addition, it was learned that visualizations of the product in use helped to create viewer
identification in the segment that used perfume for adding to nature’s handiwork, but that
more subtle methods of communication produced better results among the more reserved,
more modest women in the second segment who want the “canceling out” benefits of
perfume. The media susceptibilities of women in the two segments were also found to be
different.

Thus, from a single act of resegmentation, the advertising department extracted data critical to
its copy platform, communication strategy, and media decisions.

7
IV. Bathing Soap

A comparable purposive segmentation was found in the closely related bathing soap field. The
key split was between women whose chief requirement of soap was that it should clean them
adequately and those for whom bathing was a sensuous and enjoyable experience. The
company (a new contender in this highly competitive field) focused its sights on the first
segment, which had been much neglected in recent years. A new soap was shaped, designed,
and packaged to appeal to this segment, a new advertising approach was evolved, and results
were very successful.

V. Hair-Care Market

The Breck-Halo competition in the shampoo market affords an excellent example of another
kind of segmentation. For many years, Breck’s recognition of the market’s individualized
segmentation gave the company a very strong position. Its line of individualized shampoos
included one for dry hair, another for oily hair, and one for normal hair. This line accurately
paralleled the marketing reality that women think of their hair as being dry, oily, or normal, and
they do not believe that any one shampoo (such as an all-purpose Halo) can meet their
individual requirements. Colgate has finally been obliged, in the past several years, to revise its
long-held marketing approach to Halo, and to come out with products for dry hair and for oily
hair, as well as for normal hair.

Other companies in the hair-care industry are beginning to recognize other segmentations in
this field. For example, some women think of their hair as fine, others as coarse. Each newly
discovered key segmentation contains the seeds of a new product, a new marketing approach,
and a new opportunity.

VI. Other packaged goods

Examples of segmentation analysis in other packaged goods can be selected almost at random.
Let us mention a few briefly, to show the breadth of applicability of this method of marketing
analysis:

 In convenience foods, for example, we find that the most pragmatic classification is, once again,
purposive segmentation. Analysis indicates that “convenience” in foods has many different
meanings for women, supporting several different market segments. Women for whom
convenience means “easy to use” are reached by products and appeals different from those
used to reach women for whom convenience means shortcuts to creativity in cooking.
 In the market for cleaning agents, some women clean preventively, while others clean
therapeutically, i.e., only after a mess has been made. The appeals, the product characteristics,
and the marketing approach must take into account these different reasons for buying—
another example of purposive segmentation.
 In still another market, some people use air fresheners to remove disagreeable odors and others
to add an odor. A product like Glade, which is keyed to the second segment, differs from one
like Airwick in product concept, packaging, and type of scent.
 The beer market requires segmentation along at least four different axes—reasons for drinking
beer (purposive); taste preferences (aesthetic); price/quality (value); and consumption level.

8
VII. Retail soft goods

Although soft-goods manufacturers and retailers are aware that their customers are value
conscious, not all of them realize that their markets break down into at least four different
segments corresponding to four different conceptions of value held by women.

For some women value means a willingness to pay a little more for quality. For others, value
means merchandise on sale. Still other women look for value in terms of the lowest possible
price, while others buy seconds or discounted merchandise as representing the best value.

Retailing operations like Sears, Roebuck are highly successful because they project all these
value concepts, and do so in proportions which closely parallel their distribution in the total
population.

VIII. Adding machines

In marketing planning for a major adding machine manufacturer, analysis showed that his
product line had little relationship to the segmented needs of the market. Like most
manufacturers of this kind of product, he had designed his line by adding features to one or
several stripped-down basic models—each addition raising the model price. The lowest priced
model could only add; it could not subtract, multiply, divide, or print, and it was operated by
hand.

Since there are a great many features in adding machines, the manufacturer had an extremely
long product line. When the needs of the market were analyzed, however, it became clear that,
despite its length, the line barely met the needs of two out of the three major segments of the
market. It had been conceived and planned from a logical point of view rather than from a
market-need point of view.

The adding machine market is segmented along lines reflecting sharp differences in value and
purpose:

 One buyer group values accuracy, reliability, and long life above all else. It tends to buy
medium-price, full-keyboard, electric machines. There are many banks and other institutions in
this group where full-keyboard operations are believed to ensure accuracy.
 Manufacturing establishments, on the other hand, prefer the ten-key machine. Value, to these
people, means the maximum number of labor-saving and timesaving features. They are willing
to pay the highest prices for such models.
 Both these segments contrast sharply with the third group, the small retailer whose major
purpose is to find a model at a low purchase price. The small retailer does not think in terms of
amortizing his investment over a period of years, and neither labor-saving features nor full-
keyboard reliability count for as much as an immediate savings in dollars.

Despite the many models in the company’s line, it lacked those demanded by both the
manufacturer and small retailer segments of the market. But, because it had always been most
sensitive to the needs of financial institutions, it had developed more models for this segment
than happened to be needed. Product, sales, and distribution changes were required to enable
the company to compete in the whole market.

9
IX. Computers

One pragmatic way of segmenting the computer market is to divide potential customers
between those who believe they know how to evaluate a computer and those who believe they
do not. A few years ago only about 20% of the market was really open to IBM’s competitors—
the 20% who believed it knew how to evaluate a computer. By default, this left 80% of the
market a virtual captive of IBM—the majority who did not have confidence in its own ability
to evaluate computers and who leaned on IBM’s reputation as a substitute for personal
appraisal.

Another segmentation in this market involves differences in prospects’ attitudes toward the
inevitability of progress. Although this factor has been widely ignored, it is a significant
method for qualifying prospects. People who believe that progress is inevitable (i.e., that
change is good and that new business methods are constantly evolving) make far better
prospects for computers than those who have a less optimistic attitude toward progress in the
world of business.

X. Light trucks

The market for light trucks affords us another example of segmentation in products bought by
industry. As in the computer example, there are both buyers who lack confidence in their
ability to choose among competing makes and purchasers who feel they are sophisticated
about trucks and can choose knowledgeably. This mode of segmentation unexpectedly turns
out to be a key to explaining some important dynamics of the light truck market.

Those who do not trust their own judgment in trucks tend to rely very heavily on both the
dealer’s and the manufacturer’s reputation. Once they find a make that gives them reliability
and trouble-free operation, they cease to shop other makes and are no longer susceptible to
competitive promotion. Nor are they as price-sensitive as the buyer who thinks he is
sophisticated about trucks. This buyer tends to look for the best price, to shop extensively, and
to be susceptible to the right kind of competitive appeals, because he puts performance before
reputation.

These ways of looking at the truck market have far-reaching implications for pricing policy, for
product features, and for dealers’ sales efforts.

Conclusion

To sum up the implications of the preceding analysis, let me stress three points:

1. We should discard the old, unquestioned assumption that demography is always the best way of looking at
markets. The demographic premise implies that differences in reasons for buying, in brand
choice influences, in frequency of use, or in susceptibility will be reflected in differences in age,
sex, income, and geographical location. But this is usually not true. Markets should be
scrutinized for important differences in buyer attitudes, motivations, values, usage patterns,
aesthetic preferences, or degree of susceptibility. These may have no demographic correlatives.
Above all, we must never assume in advance that we know the best way of looking at a market.

10
This is the cardinal rule of segmentation analysis. All ways of segmenting markets must be
considered, and then we must select out of the various methods available the ones that have the
most important implications for action. This process of choosing the strategically most useful
mode of segmentation is the essence of the marketing approach espoused in this article.

In considering cases like those described, we must understand that we are not dealing with
different types of people, but with differences in peoples’ values. A woman who buys a
refrigerator because it is the cheapest available may want to buy the most expensive towels. A
man who pays extra for his beer may own a cheap watch. A Ford-owning Kellogg’s Corn
Flakes-eater may be closed off to Chevrolet but susceptible to Post Toasties; he is the same
man, but he has had different experiences and holds different values toward each product he
purchases. By segmenting markets on the basis of the values, purposes, needs, and attitudes
relevant to the product being studied, as in Exhibit I, we avoid misleading information derived
from attempts to divide people into types.

2. The strategic-choice concept of segmentation broadens the scope of marketing planning to include the
positioning of new products as well as of established products. It also has implications for brand
planning, not just for individual products but for the composition of a line of competing
brands where any meaningful segment in the market can possibly support a brand. One
explanation of the successful competing brand strategy of companies like Procter & Gamble is
that they are based on sensitivity to the many different modes of market segmentation. The
brands offered by P & G often appear very similar to the outsider, but small, marginal
differences between them appeal to different market segments. It is this rather than intramural
competition that supports P & G successes.

3. Marketing must develop its own interpretive theory, and not borrow a ready-made one from the social
sciences. Marketing research, as an applied science, is tempted to borrow its theoretical
structures from the disciplines from which it derives. The social sciences offer an abundance
of such structures, but they are not applicable to marketing in their pure academic form. While
the temptation to apply them in that form is great, it should be resisted. From sociology, for
example, marketing has frequently borrowed the concept of status. This is a far-reaching
concept, but it is not necessarily the most important one in a marketing problem, nor even one
of the important ones. Again, early psychoanalytic theory has contributed an understanding of
the sexual factor. While this can sometimes be helpful in an analysis of buying behavior in a
given situation, some motivation researchers have become oversensitive to the role of sex and,
as a result, have made many mistakes. Much the same might be said of the concept of social
character, that is, seeing the world as being “inner-directed,” “other-directed,” “tradition-
directed,” “autonomous,” and so forth.

One of the values of segmentation analysis is that, while it has drawn on the insights of social
scientists, it has developed an interpretive theory within marketing. It has been homegrown in
business. This may explain its ability to impose patterns of meaning on the immense diversity
of the market, and to provide the modern marketing director with a systematic method for
evolving true marketing objectives.

A version of this article appeared in the March 1964 issue of Harvard Business Review.

11
Daniel Yankelovich is chairman of Viewpoint Learning, a firm that promotes problem
solving through dialogue, and DYG, a market research firm that tracks social trends. He
is based in San Diego.

https://hbr.org/1964/03/new-criteria-for-market-segmentation&cm_sp=Article-_-Links-_-
Top%20of%20Page%20Recirculation

12