Effective Marketing Segmentation is More Than Simplistic Demographics

Dr Brian Monger

If you are basing your marketing on simple demographic titles like “Millenials”  or Generation C  or Baby Boomers.  You should read this Segmentation Philosophy

Segmentation Philosophy

Segmentation, or the idea of targeting the homogeneous components of a heterogeneous market rather than the market as a whole, was initially popularised in 1956 by Wendell Smith’s article in the Journal of Marketing.  Smith defined it as a marketing strategy designed for “a heterogeneous market by emphasising the precision with which a firm’s products can satisfy the requirements of one or more distinguishable market segments.

Wroe Alderson, writing at the same time as Smith, justified segmentation by saying that an advanced market economy harbours a highly diverse demand.’ This heterogeneous demand implies a market composed of different groups or segments, each with unique patterns of income, spending, lifestyle, product preference, and many other characteristics.  These clusters of buyers, forming cohesive groups that differ from one another, invite firms to cater to their needs.  Entrepreneurs exploit the market differences by offering products fitted specifically to the various segments.  In this way, marketers fulfil the social obligations of the marketing concept-serving consumer needs-by matching supply with demand more effectively.

Market segmentation was originally conceived as an alternative to product differentiation.  But good ideas get copied quickly, and a company that finds a profitable market segment soon discovers it has competition.  The firm must then concentrate on keeping the identity of its products unique.  Thus, segmentation and differentiation take place at the same time.

As the segmentation concept grew in popularity, it expanded to include almost any marketing plan that contemplates less than an entire market.  Regional marketing is sometimes categorised under that label.’ Promotions of wine with lunch or dinner can be considered occasion segmentation because a single meal is only one of many situations in which people drink wine.  Product positioning can be called psychological segmentation, for the advertising tries to carve out a “share of mind”. These interpretations may be valid, but a marketer must take care not to define a segment too abstractly to be useful.

While each business incorporates segmentation into its direct marketing program according to its needs, these activities collectively share common criteria distinguishing them from other strategies.  These criteria fall into two categories, those that differentiate segmentation from other marketing efforts and those that examine the process itself and classify the types of segmentation being performed.

Requirements for Effective Segmentation

Any marketing activity must meet certain requirements not mandatory in other marketing practices for it to be considered as using true segmentation.  The most oft-cited requirements are substantiality, differential response, identification, stability, and accessibility.


The matter of size is a relative one.  Does it matter that yours is bigger?

What may be adequate to one firm may be inadequate to another; no two companies possess identical strengths and weaknesses, no two firms have the same resources and managerial philosophies.  But, regardless of its internal capabilities and outlook, the segments a company chooses as its marketing areas must fit its objectives and resources.  The choice of market size alone does not differentiate segmentation from other activities; size is a factor when defining any market.  It is more critical to segmentation, though, owing to its focus on parts or fragments of a market.  If segments are small in relation to a company’s current operation their contribution to overall corporate results may be meagre, but large segments may tax company resources and strain cash flows.  As with any marketing strategy, the optimum ratio between resources and market size must be found.

Differential Response

An absolute requirement for segmentation is that each defined segment must have unique market response characteristics.  The fact that a market segment or a mailing list represents households with selective traits is of no use if the response from it is not substantially different from that of other segments or lists.  Though a selective approach usually means higher communication costs per thousand prospects, a correctly targeted segment will pay off in equally high purchase rates.  Since a uniform response over an entire market renders segmenting that market irrelevant, the choice of specialised media his no functional value other than limiting the scope of operations.


Any segment used in a marketing program must be capable of being identified quantitatively.  If a segment cannot be reduced to statistics there is no way of knowing if the marketing program targeted to it is successful.  Segmentation is a method of resource allocation, and as such must be subject to mathematical analysis; managers, planners, and financial analysts need to know how all the defining characteristics of a segment are distributed among its population in order to judge each one’s effect on the marketing program.


Segments must be relatively stable over time.  Segments that change radically in a short time are precarious investments.  Shortening the payback period reduces risk, but long-term returns are still unreliable.  Unstable segments are often attractive to marketers wishing to cash in quickly on a trend, but they generally hold little attraction for prudent, long-term investors.  Some marketing segments are more volatile than others.

Groups that adopt fads are prime examples, as are segments that follow the vagaries of fashion and style.  Any analysis should determine a segment’s stability and whether it is worth the risk and expected returns to market a given product in that segment.


Accessibility dictates that direct marketers must be capable of reaching their chosen segments efficiently through one or more media.  Each firm determines its own level of efficiency by correlating its resources and objectives with certain factors common to all marketing programs.

First, the efficiency of message transmission is positively correlated with market size.  For example, the cost per thousand names for a mass-produced list is normally lower than for a selective one.  Smaller, more specialised lists can be justified only if their greater response offsets their additional expense.

Second, Promotional (eg advertising) unit production costs tend to decline as volume expands.  Longer production runs bring economies of scale.  Larger order size, especially of identical units, reduces transportation and handling costs.  These quantity-cost relationships hold for all media though their effects are greatest with printed material.

A third issue is communication wastage.  Assuming a company initially targets segments with the highest spending potential, any further expansion must lower efficiency.  The general rule is to proceed until the marginal net return equals the marginal outlay.  As this process continues it deviates more and more from segmentation’s selective product-market philosophy.

Fourth, a firm’s market size criterion or other marketing goals may conflict with segmentation’s specialised approach.  For example, holders of American Express Prestige cards form a distribution that skews towards the upper end of the socioeconomic scale.  From this standpoint, American Express serves a selective market.  But the company is only practicing segmentation to a limited extent when it advertises on network television.  With the exception of certain events, the medium is a generalised one; it reaches all people.  The fact that the higher-income, better-educated portion of the audience is more apt to respond is beside the point.  If segmentation is a marketing strategy then the focus falls on what the company does; what buyers do are consequences of marketing action.  When commercials go out as generalised messages, addressed “To Whom It May Concern, segmentation is compromised.

Compromises are always present in varying degrees, reflecting the ever-present possibilities of trade-offs.  Marketers must select criteria for use that both define a market segment and target those elements in it that advertising is to be directed toward, for only then will the returning information tell them whether their chosen compromise is the best strategy for any particular market mix.

TABLE-1     Most Often Used Criteria for Segmentation Marketing


Benefits sought



States of Being


Opinions, interests

Perceptions, images



Consumption patterns

Product usage

Recency of purchase


Frequency of purchase

Purchase volume/order size



Criteria for Segmentation

The criteria used to segment markets are of three types – behaviouristic, states of being, and psychological.

Behavioural criteria divide a market on the basis of purchase behaviour; what was bought, in what quantities, where, and at what times.

States of being are attributes of the population, including sex, age, education, occupation, income, and geographical distribution,

Psychological criteria involve individual states of mind and their collective effect on the market.  Because they are subjective, states of mind are the most difficult criteria to quantify.  There is no agreement among market researchers as to whether the measurements are in effect ordinal or interval scales.’ Ordinal scales rank items into graded groups and thus cannot yield all measures of central tendency, such as mean, median, and average.  Interval scales, in contrast, are continuums beginning at arbitrary zero points, and therefore do yield central tendency measurements.  In contrast, behaviour and states of being can often be measured on a ratio scale, the only one that permits comparisons of absolute magnitudes.

These classifications can be broken down into innumerable subcategories, limited only by the imagination of the analyst.

 The criteria in Table 1 can be used independently or in combination.  These single or multiple variables are then associated with predetermined general response rates to yield market segments with unique combinations of buying traits.

The predetermined response rates used in this analysis fall into two categories: response to the product itself and response to previous promotions of the product.  The former quantifies a product’s long-term sales pattern; the latter involves consumer reactions to price deals, coupons, and other promotional appeals.”

Response can be measured over a single defined purchasing period or series of such periods.” The former requires cross-sectional data, the latter longitudinal measurements, obtaining repeat observations from the same consuming units.

The selected segmentation criteria provide a conceptual framework for a market partitioning strategy.  It then remains to gather data from the segment according to the established criteria.  This is the operational phase of the segmentation strategy.

Behaviouristic Criteria

 Consumption and usage patterns

The most popular method of segmentation in direct marketing is to divide up customers according to consumption and product usage patterns since purchase and usage rates are usually correlated.  This criterion is especially attractive when promotions are directed to names kept in an internal database of previous and current customers.

This approach rests on the premise that for many products a relatively small percentage of buyers are responsible for the lion’s share of the volume purchased.

Goods and services

Any marketing effort should be directed to the main source of business, the heavy buyer, for sales and profits are obviously better where the propensity to consume is highest.

The heavy user strategy has been in use since long before the segmentation concept was developed.

Decisions based on buying patterns can be uni-dimensional, two, or three factored, using recency, frequency, and amount bought singly or in combination.  Groupings are relatively few in number and large in size, which raises questions about the extent of homogeneity within a segment.” The less the similarity in buying behaviour, the less the value of the segment.  In practice, the various segments based on the tripartite criteria are used primarily to determine whether solicitations should be made or withheld, and not to match offers to segments as envisioned by the segmentation philosophy.

When recency, frequency, volume weights are used, they should be related to a particular product mix.  Different products have different replacement rates, and hence varying buying frequencies.  Items at different price levels would also have diverse elasticities, and account for different proportions of a total purchase.  Many marketers assign arbitrary percentage values to the three factors, with recency weighted .50, frequency of purchase .35, and amount of purchase .15.” The large weighting for recency is based on the rationale of attrition among buyers, which may easily come to 50 percent a year.  When the multiple purchase pattern, or buying frequency, is added to the recency factor, the pull of a mailing is said to increase substantially.  Regardless of the logic, these weightings have no theoretical basis.  Arbitrarily derived ratios are too general to apply to specific products.

Segments based on patterns of consumption lend themselves best to being marketed using customer lists from multi product marketers.  But firms must also often go beyond their own databases to mass-produced lists and possibly to general media.  Here purchase data are not recorded for each name.  Heavy users must be identified by indirect methods, relating consumption and usage patterns to other variables, the most common of which denote states of being.


Geographic variation is of two basic types: unevenness in population distribution and locational differences in product tastes and preferences.  The first source of variation involves the placing of physical inventories and is of minor concern to this text.   The second factor is inherent in market demand, and heavily influences segmentation practices.

Taste and preference differences can occur at various geographic levels.  For example, real estate developers of recreational communities are confronted with regional differences.

Airline travel similarly has strong regional biases, and tours promoted by direct marketing must consider these geographic variations.

To arrive at viable segments, direct marketers must link geography with response rates and to other state of being measurements.  Otherwise, their promotions to broad, geographical areas are regional marketing rather than segmentation.

Demographic and socioeconomic variables

The Bureau of Statistics systematically updates its socioeconomic data through a program of continuing surveys.  Estimates for large geographic regions or major metropolitan areas contain relatively small sampling errors, but such sample bases are wholly inadequate for estimating current population characteristics for small units, such as zip code areas; these must await the  Census to be updated.

It is possible in certain instances to make local estimates from sources other than those of the Census.  Real estate values can be updated from local assessment rolls.   These methods, however, are tedious and time-consuming.  Compilers of mass lists and directories update small area counts but not changes in distribution of demographic and socioeconomic data.

Despite their shortcomings, socioeconomic characteristics are invaluable for pinpointing opportunities among population segments.

Multiple states of being 

Using segmentation properly requires establishing a relationship between purchases or usage and demographic variables.  The discussion so far has implied a simple relationship between only two factors.  The simultaneous use of more than one state of being may yield more discriminating consumer targets.

Another combination of state of being categories could improve the buying rate but this can only be determined by repeating the analysis until the most profitable combination is found.

Another approach to prediction is multiple regression analysis.  This uses more than one independent variable and, assuming a linear relationship

Psychological Segmentation

Consumption patterns and demographics are useful ways of conceptualising market segments.  But age, income, and occupation alone do not account for consumer behaviour.  Insofar as they describe markets they indicate prospective buyers, but they cannot tell what causes a person to respond to a particular offer.  Such causal factors are to be found in people’s minds, not in their economic or social status.

Attempts to map buyers’ states of mind have applied numerous theories of consumer behaviour and used a variety of measurements.  These make up a vast literature and are beyond the scope of this book.  Among direct marketers using segmentation two psychological categories have gained prominence: the benefits sought by consumers and their lifestyle psychographics.

 Benefits sought (by consumer)

Benefits segmentation assumes that marketing should address itself to consumer needs.  Its underlying rationale is that consumers are not interested in products per se, but in the benefits ensuing from their consumption.  Accordingly, the variations in benefits sought create natural market segments.”

The search for product benefits focuses not on a product’s physical attributes, but on the consequences of its use.  Thus an automobile is not a four-wheeled vehicle but personal transportation.  Consumers do not buy toothpaste, but cavity-free teeth.  However, not all consequences are necessarily utilitarian, or obvious.  An automobile can be a status symbol or an object of vicarious pleasure.  A toothpaste can be a love amulet.  A perfume offers not only a distinctive scent, but an erotic fantasy.  A wine may evoke not only sensations of the palate, but images of romantic escapades and sidewalk cafes.

The search for benefits must therefore take into account the obvious as well as the imperceptible, the utilitarian as well as the imaginary.  The investigation also concerns itself with discrepancies between preferences and products; it seeks to discover gaps between what consumers want and what they are getting.  On the whole, larger gaps lower the probability of a purchase, but since not all discrepancies are of equal import it becomes necessary to determine which are decisive in consumer choices.

Lifestyle psychographics 

Lifestyle segmentation is more complicated and less straightforward than product benefit segmentation.  This is because lifestyles can be defined from various perspectives.  Sometimes they are seen as coinciding with social class, and are ultimately delineated by the conventional mixture of education, occupation, and financial status.” On other occasions they are viewed as combinations of different social and economic variables, and hence classifications differ little from those using only demographic and socioeconomic criteria.” The activities and states of mind making up a lifestyle are studied using an analytical tool called a psychographic profile.”

The usual lifestyle psychographic profile in marketing is based on consumer activities, interests, and attitudes.

It distinguishes between activities, which denote behaviour, and psychographics, which connote a mapping of mental states.  One reason for this dichotomy is direct marketing’s widespread use of lists representing activities, such as attendance at conventions or trade shows, membership in professional associations, subscriptions to magazines, or use of credit cards.  It is therefore useful to distinguish between the activity-oriented and psychologically-oriented aspects of a psychographic profile.

Interests and attitudes are inferred from activity-oriented data.  Thus, a subscriber to a gardening publication is presumed to be interested in shrubs and flowers.  Homes that entertain heavily are regarded as good prospects for wine, liquor, and snack sales.  The psychological segments derive their classification by direct measurement of the motives, interests, and attitudes that supposedly trigger behaviour.  Inference nevertheless plays a part, for the measurements are of indirectly observable variables.

On the whole, psychological criteria have been the least used in direct marketing for purposes of market segmentation, for they are more limited than either consumption or states of being criteria.

One major limitation is that these studies have defined segments in psychological terms that have no counterpart in population statistics.  A liquor study, for example, characterised groups drinking for “mood modification” and “social lubrication.” Another defined its segments as innovative, opinion leaders, socially active, and community-minded.  A Psychographic study with a sample of 4,000 adult males divided the population into eight groups: quiet family men, traditionalists, discontented, ethical highbrow, pleasure oriented, achievers, he-men, and sophisticated.”

Another shortcoming is that psychographic profiles have no theoretical underpinnings.” In the absence of a well-defined theory there is little understanding of causality, how variables relate to each other.  This lack of defined relationships leads to what Richard Bagozzi called the two most common faults of contemporary marketing research, “blind empiricism” and “sterile tautologies.

Did you find this article useful?  Please let us know

These are outline notes from a MAANZ course.  If you are interested in obtaining the full set of notes (and a PowerPoint presentation) please contact us – info@marketing.org.au

Also check out other articles on http://smartamarketing2.wordpress.com

MAANZ International website http://www.marketing.org.au

Smartamarketing Slideshare (http://www.slideshare.net/bmonger)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s