Market segmentation

Market segmentation is the process of dividing a broad market, normally consisting of existing and potential customers, into subsets of consumers (known as segments), that exhibit some type of shared characteristics. Typical shared characteristics include such things as common needs, common interests, similar lifestyles or even similar demographic profiles. The overall aim of segmentation is to identify high yield segments - that is, those segments that are likely to be the most profitable or that have growth potential - so that these can be selected for special attention (i.e. target markets). There are many ways to segment a market. Business-to-business sellers might segment the market into different types of businesses, or countries. While marketers of consumer goods might segment the market into demographic segments, lifestyle segments, behavioral segments or any other meaningful segment.

Market segmentation assumes that different market segments require different marketing programs - that is different offers, prices, promotion, distribution or some combination of marketing variables. Market segmentation is not only designed to identify the most profitable segments, but also to develop profiles of key segments in order to understand their needs and purchase motivations. Insights from segmentation analysis are subsequently used to support marketing strategy development and planning. Many marketers use the S-T-P approach; Segmentation→ Targeting → Positioning to provide the framework for marketing planning objectives. That, is a market is segmented, one or more segments are selected for targeting, and products or services are positioned in a way that resonates with the target markets.

Market segmentation: Historical Overview

The business historian, Richard S. Tedlow, identifies four stages in the evolution of market segmentation:

Figure 1 The Model T Ford Henry Ford famously said “Any customer can have a car painted any color that he wants so long as it is black”

• 'Fragmentation (pre 1880s)': The economy was characterised by small regional suppliers who sold goods on a local or regional basis
• 'Unification or Mass Marketing (1880s-1920s)': As transportation systems improved, the economy became unified. Standardised, branded goods were distributed at a national level. Manufacturers tended to achieve economies of scale by specialising in volume production of a single model or brand e.g. the Model T Ford
• 'Segmentation (1920s-1980s)': As market size increased, manufacturers were able to produce different models pitched at different quality points to meet the needs of various demographic and psychographic market segments. This is the era of market differentiation based on demographic, socio-economic and lifestyle factors.
• 'Hyper-segmentation (1980s+)': a shift towards the definition of ever more narrow market segments. Technological advancements, especially in the area of digital communications, allows marketers to communicate to individual consumers or very small groups.

Contemporary market segmentation was pioneered by brand marketers in the twentieth century in response to two pressing issues. Firstly, demographic and purchasing data was available for groups but rarely for individuals and secondly, advertising and distribution channels were available for groups, but rarely for single consumers and so brand marketers approached the task from a tactical viewpoint. Conventional segmentation was brand-driven process. Until recently, most segmentation approaches have retained this tactical perspective in that they address immediate short-term questions; typically decisions about the current “market served” and are concerned with informing marketing mix decisions.

Criticisms of Market Segmentation

The limitations of conventional segmentation have been well documented in the literature . Perennial criticisms include:

• that it is no better than mass marketing at building brands
• that in competitive markets, segments rarely exhibit major differences in the way they use brands
• that it fails to identify sufficiently narrow clusters
• geographic/demographic segmentation is overly descriptive and lacks sufficient insights into the motivations necessary to drive communications strategy
• difficulties with market dynamics, notably the instability of segments over time and structural change which leads to segment creep and membership migration as individuals move from one segment to another

Market segmentation has many critics. But in spite of its limitations, market segmentation remains one of the enduring concepts in marketing and continues to be widely used in practice. One American study, for example, suggested that almost 60 percent of senior executives had used market segmentation in the past two years.

Market Segmentation Strategy

A key consideration for marketers is whether to segment or not to segment. Depending on company philosophy, resources, product type or market characteristics, a businesses may develop an undifferentiated approach or differentiated approach. In an undifferentiated approach (also known as mass marketing), the marketer ignores segmentation and develops a product that meets the needs of the largest number of buyers. In a differentiated approach the firm targets one or more market segments, and develops separate offers for each segment.

In consumer marketing, it is difficult to find examples of undifferentiated approaches. Even goods such as salt and sugar, which were once treated as commodities, are now highly differentiated. Consumers can purchase a variety of salt products; cooking salt, table salt, sea-salt, rock salt, kosher salt, mineral salt, herbal or vegetable salts, iodized salt, salt substitutes and if that is not enough choice, at the brand level, gourmet cooks are likely to make a major distinction between Maldon salt and other competing brands.The following table outlines the main strategic approaches.

Table 1: Main Strategic Approaches to Segmentation

Number of Segments	Segmentation Strategy	Comments
Zero	Undifferentiated strategy	Mass marketing
One	Focus strategy	Niche marketing
Two or more	Differentiated strategy	Multiple niches
Thousands	Hypersegmentation	One-to-one marketing

Market Segmentation Process

The process of segmenting the market is deceptively simple. Seven basic steps describe the entire process including segmentation, targeting and positioning. Make no mistake, however, the process can be very laborious as it involves poring over loads of data, and requires a great deal of skill in analysis and interpretation. The basic steps are outlined here:

1. Identify market (also known as universe) to be segmented.

2. Identify, select and apply base or bases to be used in the segmentation

3. Develop segment profiles

4. Evaluate each segment's attractiveness

5. Select segment or segments to be targeted

6. Identify optimal positioning

7. Develop and communicate the positioning statement

Bases for Segmenting Consumer Markets

A major step in the segmentation process is the selection of a suitable base. In this step, marketers are looking for a means of achieving internal homogeneity (similarity within the segments), and external heterogeneity (differences between segments). In other words, they are searching for a process that minimises differences between members of a segment and maximises differences between each segment. In addition, the segmentation approach must yield segments that are meaningful for the specific marketing problem or situation. For example, a person's hair colour may be a relevant base for a shampoo manufacturer, but it would not be relevant for a seller of financial services. Selecting the right base requires a good deal of thought and a basic understanding of the market to be segmented.

In reality, marketers can segment the market using any base or variable provided that it is identifiable, measurable, actionable and stable . For example, some fashion houses have segmented the market using women's dress size as a variable. However, the most common bases for segmenting consumer markets include: geographics, demographics, psychographics and behavior. Marketers normally select a single base for the segmentation analysis, although, some bases can be combined into a single segmentation with care. For example, geographics and demographics are often combined, but other bases are rarely combined. Given that psychographics includes demographic variables such as age, gender and income as well as attitudinal and behavioral variables, it makes little logical sense to combine psychographics with demographics or other bases. Any attempt to use combined bases needs careful consideration and a logical foundation.

The following sections provide a detailed description of the most common forms of consumer market segmentation.

Geographic segmentation

Geographic segmentation divides markets according to geographic criteria. In practice, markets can be segmented as broadly as continents and as narrowly as neighborhoods or postal codes. Typical geographic variables include:

• Country e.g. USA, UK, China, Japan, South Korea, Malaysia, Singapore, Australia, New Zealand
• Region e.g. North, North-west, Mid-west, South, Central
• Population density: e.g. central business district (CBD), urban, suburban, rural, regional
• City or town size: e.g. under 1,000; 1,000- 5,000; 5,000 - 10,000... 1,000,000 - 3,000,000 and over 3,000,000
• Climatic zone: e.g. Mediterranean, Temperate, Sub-Tropical, Tropical, Polar,

The geo-cluster approach (also called geo-demographic segmentation) combines demographic data with geographic data to create richer, more detailed profiles. Geo-cluster models break down an area into groups of households based on the assumption that people in any given neighbourhood are likely to exhibit similarities in their demographic composition and other household characteristics.

Geographic segmentation may be considered the first step in international marketing, where marketers must decide whether to adapt their existing products and marketing programs for the unique needs of distinct geographic markets. Tourism Marketing Boards often segment international visitors based on their country of origin. By way of example, Tourism Australia undertakes marketing in 16 core geographic markets; of which China, UK, US, NZ and Japan have been identified as priority segments because they have the greatest potential for growth and are extremely profitable segments with higher than average expenditure per visit. Tourism Australia carries out extensive research on each of these segments and develops rich profiles of high priority segments to better understand their needs and how they make travel decisions. Insights from this analysis are used in travel product development, allocation of promotional budgets, advertising strategy and in broader urban planning decisions. For example, in light of the numbers of Japanese visitors, the city of Melbourne has erected Japanese signage in tourist precincts.

A number of proprietary geo-demographic packages are available for commercial use. Examples include Experian's Mosaic (geodemography) Segmentation (active in North America, South America, UK, Europe, South Africa and parts of Asia-Pacific ) or Helix Personas (Australia, New Zealand and Indonesia). It should be noted that all these commercial packages combine geographics with behavioural, demographic and attitudinal data and yield a very large number of segments. For instance, the NZ Helix Personas segments New Zealand's relatively small population into 51 discrete personality profiles across seven geographic communities ( too numerous to itemise on this page). These commercial databases typically allow prospective clients access to limited scale demonstrations, and readers interested in learning more about how geo-demographic segmentation can benefit marketers and planners, are advised to experiment with the online demonstration software via these companies' websites.

Geographic segmentation is widely used in direct marketing campaigns to identify areas which are potential candidates for personal selling, letter-box distribution or direct mail. Geo-cluster segmentation is widely used by Governments and public sector departments such as urban planning, health authorities, police, criminal justice departments, telecommunications and water boards.

Demographic segmentation

Segmentation according to demography is based on consumer- demographic variables such as age, income, family size, socio-economic status, etc. Demographic segmentation assumes that consumers with similar demographic profiles will exhibit similar purchasing patterns, motivations, interests and lifestyles and that these characteristics will translate into similar product/brand preferences. In practice, demographic segmentation can potentially employ any variable that is used by the nation's census collectors. Typical demographic variables and their descriptors are as follows:

• Age: e.g. Under 5, 5–8 years, 9–12 years, 13–17 years, 18-24, 25-29, 30-39, 40-49, 50-59, 60+
• Gender: Male, Female
• Occupation: Professional, self-employed, semi-professional, clerical/ admin, sales, trades, mining, primary producer, student, home duties, unemployed, retired
• Social class (or socio-economic status):
• Marital Status: Single, married, divorced, widowed
• Family Life-stage: Young single; Young married with no children; Young family with children under 5 years; Older married with children; Older married with no children living at home, Older living alone
• Family size/ number of dependants: 0, 1-2, 3-4, 5+
• Income: Under $10,000; 10,000- 20,000; 20,001- 30,000; 30,001-40,000, 40,001- 50,000 etc
• Educational attainment: Primary school; Some secondary, Completed secondary, Some university, Degree; Post graduate or higher degree
• Home ownership: Renting, Own home with mortgage, Home owned outright
• Ethnicity: Asian, African, Aboriginal, Polynesian, Melanesian, Latin-American, African-American, American Indian etc
• Religion: Catholic, Protestant, Muslim, Jewish, Buddhist, Hindu, Other

Figure 2 Scouts develop products for people of all ages

The Scouting movement offers an excellent example of demographic segmentation in practice. Scouts develops different products based on relevant age groups. In Australia, the segments are Joeys for boys and girls aged 6–7 years; Cubs for children ages 8– 10 years; Scouts for those aged 11–14 years; Venturers ages 15–17 years and Rovers aged 18–25 years. The Scouting movement provides members of each cohort with different uniforms and develops different activity programs for each segment. Scouts even cater to the needs of the over 25s offering them roles as scout leaders or volunteers. Scouts' segmentation is an example of a simple demographic segmentation analysis which utilises just a single variable, namely age.

In practice, most demographic segmentation utilises a combination of demographic variables. For instance, a segmentation approach developed for New Zealand by Nielsen Research combines multiple demographic variables including age, life-stage and socio-economic status. The proprietary segmentation product, known as geoTribes, segments the NZ market into 15 tribes, namely: Rockafellas- Affluent mature families; Achievers-Ambitious younger and middle aged families; Fortunats- Financially secure retirees and pre-retirees; Crusaders-Career-oriented singles and couples; Preppies- Mature children of affluent parents; Indpendents- Young singles and couples; Suburban Splendour- Middle class mature families; Twixters- Mature children living at home; Debstars-Financially extended younger families; Boomers -White collar post family pre-retirees; True Blues -Blue collar mature families and pre-retiree singles or couples; Struggleville -Struggling young and middle aged families; Grey Power-Better off retirees; Survivors-Retirees living on minimal incomes and Slender Meanz-People living in underprivileged circumstances.

Figure 3 Visualisation of two approaches to demographic segmentation using one and two variables

The use of multiple segmentation variables normally requires analysis of databases using sophisticated statistical techniques such as cluster analysis or principal components analysis. It should be noted that these types of analysis require very large sample sizes. However, data-collection is expensive for individual firms. For this reason, many companies purchase data from commercial market research firms, many of whom develop proprietary software to interrogate the data. Proprietary packages, such as that mentioned in the preceding geoTribes by Nielsen example, offer clients access to an extensive database along with a program that allows users to interrogate the data via a 'user-friendly' interface. In other words, users do not need a detailed understanding of the 'back-end' statistical procedures used to analyse the data and derive the market segments. However, users still need to be skilled in the interpretation of findings for use in marketing decision-making.

Psychographic segmentation

Psychographic segmentation, which is sometimes called lifestyle, is measured by studying the activities, interests, and opinions (AIOs) of customers. It considers how people spend their leisure, and which external influences they are most responsive to and influenced by. Psychographics is a very widely used basis for segmentation, because it enables marketers to identify tightly defined market segments and better understand the motivations for selecting particular products.

One of the most well-known psychographic segmentation analyses is the so-called Values And Lifestyles Segments (VALS), a proprietary psychometric method that measures attitudes, behaviours and demographics that align with brand preferences and new product adoption. The approach was originally developed by SBI International in the 1970s, and the typologies or segments have undergone several revisions over the years. As it currently stands, the VALs segments that describe the adult American population are: Achievers (26%), Strivers (24%), Experiencers (21%), Innovators, Thinkers, Believers, Makers and Survivors. The VALs segments are country specific and the developer offers VALs segmentation typologies for use in China, Japan, the Dominican Republic, Nigeria, UK and Venezuela.

Outside the USA, other countries have developed their own brand of proprietary psychographics segmentation. In Australia and New Zealand, Roy Morgan Research has developed the Values Segments which describes ten segments based on mindset, demographics and behaviours. The Values Segments are: Visible Achievers (20%); Traditional Family Life (19%); Socially Aware (14%); Look-At-Me (11%); Conventional Family Life (10%); Something Better (9%); Young Optimists (7%); Fairer Deal (5%); Real Conservatives (3%) and Basic Needs (2%) Market research company, Nielsen offers PALS (Personal Aspirational Lifestyle Segments), a values-based segmentation that ranks future lifestyle priorities, such as the importance of career, family, wanting to have fun, or the desire for a balanced lifestyle. PALS divides the Australian market into six groups, namely Success Driven (25%); Balance Seekers (23%); Health Conscious (22%), Harmony Seekers (11%), Individualists (11%) and Fun Seekers (8%).

While these proprietary psychographic segmentation analyses are well-known, the majority of studies based on psychographics are custom designed. That is the segments are developed for individual products at a specific time. One common thread among psychographic segmentation studies is that they use quirky names to describe the segments.

Behavioral segmentation

Behavioral segmentation divides consumers into groups according to their observed behaviors. Many marketers believe that behavioral variables are superior to demographics and geographics for building market segments. Typical behavioral variables and their descriptors include:

• Purchase/Usage Occasion: e.g. regular occasion, special occasion, festive occasion, gift-giving
• Benefit-Sought: e.g. economy, quality, service level, convenience, access
• User Status: e.g. First-time user, Regular user, Non-user
• Usage Rate/ Purchase Frequency: e.g. Light user, heavy user, moderate user
• Loyalty Status: e.g. Loyal, switcher, non-loyal, lapsed
• Buyer Readiness: e.g. Unaware, aware, intention to buy
• Attitude to Product or Service: e.g. Enthusiast, Indifferent, Hostile
• Adopter Status: e.g. Early adopter, late adopter, laggard

Note that these descriptors are merely commonly used examples. Marketers must ensure that they customize the variable and descriptors for both local conditions and for specific applications. For example, in the health industry, planners often segment broad markets according to 'health consciousness' and identify low, moderate and highly health conscious segments. This is an example of behavioral segmentation, using attitude to product or service as a key variable.

Purchase/ Usage Occasion

Purchase or usage occasion segmentation focuses on analyzing occasions when consumers might purchase or consume a product. This approach customer-level and occasion-level segmentation models and provides an understanding of the individual customers’ needs, behavior and value under different occasions of usage and time. Unlike traditional segmentation models, this approach assigns more than one segment to each unique customer, depending on the current circumstances they are under.

For example, Cadbury has segmented the market into five segments based on usage and behavior:

Figure 4 Cadbury segment consumers of their chocolate range according to usage occasion

• Immediate Eat (34%): Driven by the need to snack, indulge or an energy boost. Products that meet these needs include brands such as Kit-Kat, Mars Bars
• Home Stock (25%): Driven by the need to have something in the pantry to share with family in front of TV or for home snacking. Products that meet these needs are blocks or multi-packs
• Kids (17%): Driven by need for after school snacks, parties, treats. Products that meet these needs include Smarties and Milky Bars.
• Gift-giving (15%): Products purchased as gifts, needs include a token of appreciation, a romantic gesture or a special occasion. Products that meet these needs include boxed chocolates such as Cabury's Roses or Quality Street
• Seasonal (3.4%): Driven by need to give a present or create a festive atmosphere. The products are mainly purchased on feast days such as Christmas, Easter, Advent. Products that meet these needs include Easter eggs and Christmas tree decorations.

Benefit-Sought

Benefit sought (sometimes called needs-based segmentation) divides markets into distinct needs or benefits sought by the consumer.

Other Types of Consumer Segmentation

Generational segments

A generation is defined as "a cohort of people born within a similar span of time (15 years at the upper end) who share a comparable age and life stage and who were shaped by a particular span of time (events, trends and developments)." Generational segments refer to the process of dividing and analysing a population into cohorts based on their birth date. Generational segmentation assumes that people's values and attitudes are shaped by the key events that occurred during their lives and that these attitudes translate into product and brand preferences.

Demographers, studying the end of the population explosion, disagree about precise dates for each generation. Dating is normally by population peaks, which can occur at different times in each country. For example, in Australia the post-war population boom peaked in 1960 while it occurred somewhat later in the USA and Europe, peaking in 1964. Accordingly, Australian Boomers are normally defined as those born between 1945-1960; while American and European Boomers are normally defined as those born between 1945-64. Thus, the generational segments and their dates discussed here must be taken as approximations only.

The primary generational segments identified by marketers are:

• Ancients (born prior to 1945)
• Baby Boomers (born about 1945- 1965)
• Generation X (born about 1966- 1976)
• Generation Y also known as Millenials ( born about 1977- 1994)
• Generation Z also known as Centennials (born 1995- 2015)

Cultural segmentation

Cultural segmentation is used to classify markets according to cultural origin. Culture is a major dimension of consumer behavior and can be used to enhance customer insight and as a component of predictive models. Cultural segmentation enables appropriate communications to be crafted to particular cultural communities. Cultural segmentation can be applied to existing customer data to measure market penetration in key cultural segments by product, brand, channel as well as traditional measures of recency, frequency and monetary value. These benchmarks form an important evidence-base to guide strategic direction and tactical campaign activity, allowing engagement trends to be monitored over time.

Cultural segmentation can also be mapped according to state, region, suburb and neighborhood. This provides a geographical market view of population proportions and may be of benefit in selecting appropriately located premises, determining territory boundaries and local marketing activities.

Census data is a valuable source of cultural data but cannot meaningfully be applied to individuals. Name analysis (onomastics) is the most reliable and efficient means of describing the cultural origin of individuals. The accuracy of using name analysis as a surrogate for cultural background in Australia is 80-85%, after allowing for female name changes due to marriage, social or political reasons or colonial influence. The extent of name data coverage means a user will code a minimum of 99 percent of individuals with their most likely ancestral origin.

Selecting Target Markets

Another major decision in developing the segmentation strategy is the selection of market segments that will become the focus of special attention (known as target markets). The marketer faces a number of decisions:

• What criteria should be used to evaluate markets?
• How many markets to enter (one, two or more)?
• Which market segments are the most valuable?

When a marketer enters more than one market, they are often labelled the primary market, secondary market and so on. There are no formulas for evaluating the attractiveness of market segments and a good deal of judgement must be exercised. Nevertheless, a number of criteria can be used to evaluate market segments for attractiveness.

Market Size:

• How large is the market?
• Is the market segment substantial enough to be profitable?
• Market size can be measured in number of customers, but superior measures are likely to include sales value or volume

Market Growth:

• Is the market growing or contracting?
• What are the indications that growth will be sustained in the long term?

Competitive Environment:

• To what extent are competitors targeting this market?
• Can we carve out a viable position to differentiate from any competitors?

Reach:

• Is this market accessible with our current distribution infrastructure?
• What are the costs/ logistics of reaching members of this market?

Compatibility:

• Is this market aligned with our company's operating philosophy?
• Do we have the resources necessary to enter this market?

Market segmentation and the Marketing Program

When the segments have been determined and separate offers developed for each of the core segments, the marketer's next task is to design a marketing program (also known as the marketing mix) that will resonate with the target market or markets. Developing the marketing program requires a deep knowledge of key market segment's purchasing habits, their preferred retail outlet, their media habits and their price sensitivity. The marketing program for each brand or product should develop a marketing program, based on the market understanding revealed in the market profile.

Table 1 provides a brief summary of the marketing mix used by Black and Decker, manufacturer of three brands of power tools and household appliances. Each brand is pitched at different quality points and targets different segments of the market. This table is designed to illustrate how the marketing program must be fine-tuned for each market segment.

Table 2: Black & Decker Market Segmentation

Segment	Product/ Brand	Product Strategy	Price Strategy	Promotion Strategy	Place Strategy
Home-owners/ D-I-Y	Black & Decker	Quality Adequate for occasional use	Lower priced	TV advertising during holidays	Mass distribution (lower-tier stores)
Weekend Warriors	Firestorm	Quality Adequate for regular use	Higher priced	Ads in D-I-Y magazines, shows	Selective distribution (top tier hardware stores)
Professional users	De Walt	Quality adequate for daily use	Highest price	Personal selling (sales reps call on job sites)	Selective distribution (top tier hardware stores)

Bases for Segmenting Business Markets

Businesses may be segmented according to industry, business size, turnover, number of employees or any other relevant variables.

Firmographics

Firmographics (also known as feature based segmentation) is the business community's answer to demographic segmentation. It is commonly used in business-to-business markets (it’s estimated that 81% of B2B marketers use this technique). Under this approach the target market is segmented based on features such as company size (either in terms of revenue or number of employees), industry sector or location (country and/or region).

Multi-variable account segmentation

In Sales Territory Management, using more than one criterion to characterize the organization’s accounts, such as segmenting sales accounts by government, business, customer, etc. and account size or duration, in effort to increase time efficiency and sales volume.

Using segmentation in customer retention

The basic approach to retention-based segmentation is that a company tags each of its active customers with four values:

Is this customer at high risk of canceling the company's service?

One of the most common indicators of high-risk customers is a drop off in usage of the company's service. For example, in the credit card industry this could be signaled through a customer's decline in spending on his or her card.

Is this customer at high risk of switching to a competitor to purchase product?

Many times customers move purchase preferences to a competitor brand. This may happen for many reasons those of which can be more difficult to measure. It is many times beneficial for the former company to gain meaningful insights, through data analysis, as to why this change of preference has occurred. Such insights can lead to effective strategies for winning back the customer or on how not to lose the target customer in the first place.

Is this customer worth retaining?

This determination boils down to whether the post-retention profit generated from the customer is predicted to be greater than the cost incurred to retain the customer, and includes evaluation of customer lifecycles.

What retention tactics should be used to retain this customer?

This analysis of customer lifecycles is usually included in the growth plan of a business to determine which tactics to implement to retain or let go of customers. Tactics commonly used range from providing special customer discounts to sending customers communications that reinforce the value proposition of the given service.

Algorithms and approaches

The choice of an appropriate statistical method for the segmentation, depends on a number of factors including, the broad approach (a-priori or post-hoc), the availability of data, time constraints, the marketer's skill level and resources.

A-priori segmentation

According to the Market Research Association (MRA),a priori research occurs when "a theoretical framework is developed before the research is conducted". In other words, the marketer has an idea about whether to segment the market geographically, demographically, psychographically or behaviorally before undertaking any research. For example, a marketer might want to learn more about the motivations and demographics of light and moderate users in an effort to understand what tactics could be used to increase usage rates. In this case, the target variable is known - the marketer has already segmented using a behavioral variable - user status. The next step would be to collect and analyse attitudinal data for light and moderate users. Typical analysis includes simple cross-tabulations, frequency distributions and occasionally logistic regression or CHAID analysis.

The main disadvantage of a-priori segmentation is that it does not explore other opportunities to identify market segments that could be more meaningful.

Post-hoc segmentation

In contrast, post-hoc segmentation makes no assumptions about the optimal theoretical framework. Instead, the analyst's role is to determine the segments that are the most meaningful for a given marketing problem or situation. In this approach, the empirical data drives the segmentation selection. Analysts typically employ some type of clustering analysis or structural equation modeling to identify segments within the data. Figure 5 shows how segments might be formed using clustering, however note that this diagram only uses two variables, while in practice clustering employs a large number of variables. Post-hoc segmentation relies on access to rich data sets, usually with a very large number of cases.

Statistical Techniques Used in Segmentation

Marketers often engage commercial research firms or consultancies to carry out segmentation analysis, especially if they lack the statistical skills to undertake the analysis. Some segmentation, especially post-hoc analysis, relies on sophisticated statistical analysis.

Figure 5 Visualisation of market segments formed using clustering methods.

Common statistical techniques for segmentation analysis include:

• Clustering algorithms such as K-means or other Cluster analysis
• Conjoint Analysis
• Logistic Regression (also known as Logit Regression)
• Chi-square Automatic Interaction Detector CHAID; a type of decision-tree
• Structural Equation Modeling (SEM)
• Multidimensional scaling and Canonical Analysis
• Statistical mixture models such as Latent Class Analysis
• Ensemble approaches such as Random Forests
• Other algorithms such as Neural Networks

Data Sources Used for Segmentation

Marketers use a variety of data sources for segmentation studies and market profiling. Typical sources of information include:

Internal databases

• Customer transaction records e.g. sale value per transaction, purchase frequency
• Patron membership records e.g. active members, lapsed members, length of membership
• Customer relationship management (CRM) databases
• In-house surveys

External sources

• Commercial Surveys or tracking studies (available from major research companies such as Nielsen and Roy Morgan)
• Government agencies or Professional/ Industry associations
• Census data
• Observed purchase behavior - collected from online agencies such as Google
• Data-mining techniques

See also

• Demographic profile
• Mass marketing
• Mosaic (geodemography)
• Niche market
• Precision marketing
• Psychographic
• Target market
• Industrial market segmentation
• Sagacity Segmentation