Charlie Nelson
director foreseechange
August 2001
Consumer diversity is increasing rapidly and firms have long sought to differentiate their products relative to competitors. This is where market segmentation comes in. Long gone are the homogeneous markets that Henry Ford conquered with his mass production of one model of car (mass customisation is the new objective). While there has been a strong move towards one-to-one marketing in recent years, there are few examples of successful implementation, particularly in consumer markets. Market segmentation provides a proven way of disaggregating markets in a way that can improve profitability without the investment in systems and sales resources needed for one-to-one marketing.
This document provides an overview of market segmentation
and links to more detailed information sources.
Market segmentation is the first of three important steps in developing marketing strategy. Segmentation groups customers with similar needs and responses; targeting determines which segments to serve; positioning is about how the product (or product portfolio) should compete with others in the market.
The objectives of market segmentation are to more accurately meet the needs of selected customers in a more profitable way.
Precisely how this can be achieved will vary by company
capability. For example, a single
product company may be able to boost sales and cut advertising costs if they can
target consumers with a high likelihood of product purchase.
On the other hand, a company with several brands in a category will
benefit by positioning each brand within the portfolio against a distinct set of
consumer needs – ideally each brand should be sufficiently distinct so that
there is little cannibalisation.
There is a large array of possible segmentation bases.
Some of these are briefly described below.
Consumers can be grouped on the basis of characteristics
such as age or household composition. This
is easy to do and it is easy to reach such segments with media.
But age and other demographics are only loosely related to behaviour.
Similarly, characteristics such as income, occupation and education can be used to derive segments that are easy to reach. Such segments are indicators (although not perfect) of behaviour such as lifestyle, price sensitivity, and brand preference.
Potential to use the firm’s product is a behaviourally
based segmentation basis. Potential
could be determined by administering questions about disposition to use (such as
awareness, used in the past, would consider using) in a survey and respondents
grouped accordingly. The problem is
then how to reach the most attractive segments. This is done either by using a large-scale single source
survey (such as ACNielsen Panorama) that asks consumers about product
disposition and media usage or by relating product disposition to demographics.
Both approaches are usually imperfect as behaviour is rarely strongly
correlated with demographics or media usage.
Personality, attitudes, opinions, and life styles are often
used a segmentation bases. These
characteristics have some relationship to behaviour and provide insight into how
to communicate with chosen segments. Reaching
the chosen segments is then the issue, as discussed under product usage, above.
Generation, or cohort, refers to people born in the same
period of time. For example, the
Baby Boomer generation can be defined as those people born between 1946 and
1955. Such cohorts share much in
common. Not only are they of a
similar age, but they experienced similar economic, cultural, and political
influences in formative years. Thus
generation is probably a better segmentation basis than age and just as easy to
reach.
Some people are price sensitive, others seek quality or
service. Some people are brand
loyal, while others frequently switch brands.
It is possible to group consumers on the basis of these factors.
Note that price/quality sensitivity can vary by category.
Some people are very concerned about the quality of the food they eat but
will buy cheap laundry detergent. Others
will feed themselves any rubbish but are fastidious about cleanliness. This is a very powerful basis for segmentation but it is not
easy to buy media on this basis. These
segments can be reached by the message (self-selection) but this is not
necessarily cost effective.
There are two reasons why people who live in the same area
may share similar characteristics. First,
some areas have more expensive properties than others and so people with similar
socioeconomic characteristics may cluster together.
Second, they have similar transport and shopping options.
It is easy to reach particular areas by using local newspapers, cinema,
outdoor, and selective direct mail but mass media is less effective.
There are several commercial geodemographic segmentation schemes available, that combine demographics and geography as a segmentation basis. This approach aims to identify groups of small geographic areas that have similar demographic profiles. These tend to suffer from the fallacy of averages. Some areas may be genuinely relatively homogenous but many are not and this can be very misleading.
More on geodemographic segmentation (pdf format, 1,351k).
The segmentation basis used depends on the decision to be made. For pricing decisions, for example, the segmentation basis should be price sensitivity and deal proneness. For advertising decisions, the bases could include benefits sought; media use; or psychographics (or some hybrid of these).
Clearly, one segmentation basis will not be ideal for all marketing decisions. Nor will one segmentation basis be ideal for all industries – food, detergents, clothing, and motor vehicles all satisfy different needs and have different levels of purchase involvement.
Nonetheless, many companies do use “generic”
segmentation schemes. They need to
satisfy themselves that in doing so:
The market segments identified should satisfy three
criteria.
This means that potential customers within a segment should
have similar responses to the marketing mix variable of interest but a different
response to members of other segments.
This is the degree to which the segmentation makes every
potential customer a unique target. That
is, the segmentation should identify a small set of groupings of substantial
size.
This is the degree to which marketers can reach segments separately using observable characteristics of the segments.
There is a large array of analytical techniques applicable
to market segmentation. The most
frequently used are briefly described below.
Cluster analysis is a set of techniques for discovering structure, or groups of individuals, within a set of data comprising measures on each individual. The measures could be, for example, an attitudinal battery. There is no dependent variable – all variables are treated equally.
This technique aims to decompose preference into component
parts, such as brand, quality, and price. This
technique views products as bundles of attributes and uses an experimental
design to vary attribute levels to create product descriptions.
Survey respondents then rank the products and the analysis works out how
much each attribute contributes to preference.
It is a good technique for benefit segmentation.
This was called Automatic Interaction Detection for a long
time and now also goes under various names used by software vendors, including
Regression Tree, Answer Tree, Classification Tree and CART.
It is a technique frequently employed in Data Mining and it is a useful
exploratory analysis technique prior to regression analysis.
It can quickly analyse a large set of candidate explanatory variables to
determine the most influential variables on a dependent variable.
The basic idea is to hierarchically segment the population
on the database based on a dependent (categorical) variable such as bought/did
not buy a product. The explanatory
variables are categorical too, such as:
This technique is used to quantify the relationship between
segment membership (eg bought, did not buy) and explanatory variables such as
income and attitudes. It is often
used after CHAID identifies candidate explanatory variables, to formally
quantify and test the significance of relationships.
“Marketing Engineering” by Gary L. Lilien and Arvind
Rangaswamy, Addison-Wesley, is an excellent book on the application of marketing
models to marketing strategy. It
includes Excel-based software to demonstrate the application of these
techniques.
A good text on the technical details of the whole range of
techniques for segmentation and more is “Multivariate Data Analysis” by
Joseph F. Hair et al, Macmillan.
For details of algorithms, see “Finding Groups in Data,
An Introduction to Cluster Analysis” by Leonard Kaufman and Peter J. Rousseeuw,
Wiley.
www.futuretoolkit.com/segfor.htm
www.futuretoolkit.com/conjoint.htm
Other Articles
Using Futures Research and Market Segmentation to Identify Opportunities (pdf, 212k)
The Diversity of Australia's Consumers
Geodemographic Segmentation (pdf, 1,351k)