When approaching market segmentation, the overall goal is to create distinct groups of customers in order to create more effective marketing strategies and tactics for each group. Organizations cannot afford to be all things to all customers; differences in the customer base must be taken into account when developing marketing plans. The underlying concept is straightforward: organizations don’t typically want to treat all their customers the same way. Different groups of customers are presumed to have different needs and contribute differently to an organization’s bottom line. Consequently, marketers need to be able to develop clear and insightful marketing strategies and messages for its diverse customer base. These strategies and tactics work together to achieve the organization’s business goals concerning revenue and profitability. The goal of segmentation is to identify the needs and customer characteristics which contribute most clearly to the development of successful marketing strategies.
Benefits-based and a priori segmentation
Traditional segmentation strategies have had two distinct focuses: benefits-based strategies and a priori strategies. Benefits segmentation theory creates customer segments based on what specific product benefits appeal to certain groups. For example, consumers who are concerned more with a snack product’s nutritional value are presumed to be different from other consumers focused on the product’s appeal to the children in the home. A priori segmentation theory focuses on brand usage patterns, demographics or media usage behaviors which are known beforehand (hence, a priori). In our snack example, a marketer may segment the market into families with and without young children and develop marketing messages accordingly.
Advantages and disadvantages of traditional techniques
Each of these traditional techniques has its benefits and drawbacks. Marketers definitely need to know which product features and benefits are of most interest to their customers and use benefits-based segmentation to address this issue. However, if these behaviors create groups which are not unique in terms of brand usage or other quantifiable behaviors, the marketing message’s effectiveness will be diluted and unclear. Consider the dilemma created with a great-tasting snack that is also healthy. Will the nutritionally-oriented consumer believe the marketing message if it focuses on taste alone? Will the consumer who has been greeted with skepticism at home about good-tasting and healthy snacks believe that the product will still get eaten if the message focuses only on how healthy it is?
A priori segmentation has its benefits and drawbacks. It is relatively easy to divide customers into groups based on what you know beforehand. However, demographics and brand usage patterns of behavior are not completely reliable as predictors of which benefits are of the greatest appeal to customers. In our example, some consumers without children may prefer the benefit of great taste more than the benefit of nutritional value just as other consumers with children may select a product associated with higher nutritional value and choose other strategies on how best to get the kids to eat it.
What to do? Predictive segmentation!
Clearly, if benefits-based segments and a priori segments don’t provide complete insight into brand usage or other profitable behaviors, then marketing strategies, messages and even brand identity are harder to clarify. A strong segmentation solution provides differentiation based not only on customer needs but also on insights into customer responses to the brand. Positive responses to the brand may involve growth in sales, market share, perception of the brand or other dependent variables important to the business.
Predictive segmentation takes both sets of challenges in hand and effectively parses the market into easily-identifiable and discrete market segments that differ based on both needs and behavior.
How does it work?
Customer attitudes, perceptions and behaviors about a brand need to be measured and analyzed together. By combining information about which product benefits are most important (benefit utilities), along with information about brand usage information, the analysis reveals the underlying relationships between them. Benefit utilities and brand behaviors are summarized based on their inter-correlations to get at a description of the overall “personality” of the brand. These data are then aggregated into two statistically-driven measures that are specific to the brands and utilities being studied. This enables the graphic representation of both the brand behaviors and benefits utilities on two axes in a perceptual map.
Perceptual maps are interpreted as follows:
- The farther apart any two brands are on the map, the more they differ in terms of which attributes describe them.
- The farther a brand is from the center of the map, the more the attributes differ from each other in terms of which attributes describe that brand.
- Roughly speaking, the closer an attribute is to a specific brand, the more frequently that attribute is said to describe the brand.
- The farther an attribute is from the center of the map, the more the brands differ in terms of the degree to which they are described by that attribute.
An example is shown in Figure 1.
Next, the relationship of each individual customer in the study’s data set to this structure of benefit utilities and behaviors is computed. In other words, customers themselves are placed within the context of this brand usage/brand utility structure to identify which customers are most likely to be associated with the brand’s usage and utilities.
Last, clusters of brands, brand benefits and people are created, which are interpretatively cohesive and distinct from other groups. This type of clustering takes into account behaviors, attitudes and customer characteristics that create better predictions of real marketplace behavior than other segmentation methods. In our example, we could describe consumers who are most likely to purchase Treat C as “health-conscious spenders” while those who purchase Treat A as “economical flavor fans.” In some studies, predictive segmentation can also yield an interactive tool which will help to tag additional or future customers into the identified segments simply by asking them a few key questions. This type of tool can be very useful for customer relationship management practices.
Lastly, not only can predictive segmentation be implemented in a new study, the technique can also be used with existing data, provided the data have benefit or attribute importance measures. Furthermore, predictive segmentation can also be used to predict other key variables in addition to brand preference, such as certain demographics or media behavior.
In practice, predictive segmentation provides a quantum-leap improvement over utilities-based or a priori segmentation practices. The benefit-seeking patterns of the predictive segments are typically well-differentiated. If they are not, the benefit utilities are usually weakly correlated with the brand choice behavior being studied. Most importantly, brand usage differentiation will be significantly higher than with conventional segmentation.
Figure 2 shows full results from a study where both techniques were implemented for comparison purposes. For every one of the 16 brands being studied, predictive segmentation provided greater differentiation around brand usage behavior.
For example, using conventional segmentation, the market share for Brand M shows little variation, ranging from 16 to 23. Variations in market share this small can easily be explained by sampling error. However, when using predictive segmentation, market share varied widely, from 9 for Segment 4 to 37 for Segment 1. Differences in market share this dramatic are much more likely to be statistically significant, showing real measurable and actionable differences in the segments created.
Statistical procedures used in conventional benefit utilities or a priori segmentation studies usually lead to mathematically elegant solutions which nonetheless do not adequately explain the relationships between “real-world” variables like brand usage, demographics and media preference. In contrast, predictive segmentation takes the analysis out of the statistical ivory tower and anchors solutions in the real marketplace. It imposes a structure which assuredly conforms better to the real world: what a person actually buys, who the customer is (in regard to demographics and attitudes) and how the customer can be reached most effectively. In today’s challenging economy, research solutions need to address the real-world challenges in reaching and marketing to customer. Predictive segmentation is a solid technique to address these challenges.
Customer Lifecycle is a global research-based consultancy committed to helping our clients avoid costly mistakes by focusing on thorough front-end planning, appropriate support for research execution and in-depth deployment consulting and implementation at the back end. Outcomes are rigorous and balanced customer-focused performance metrics, improved financial results and a superior total customer experience. Its mission is to provide companies with insight into their industries and staff by deploying sophisticated analyses to answer tough business questions and intelligence that clients can act on with confidence, thereby offering an edge in understanding customer choice, engagement, loyalty and advocacy.
Each stage in the customer lifecycle – acquisition, service, growth, retention – has its own unique challenges and solutions to address specific business issues. Customer Lifecycle helps both B2B and B2C-focused organizations plan and conduct research to accurately identify and measure customer requirements for satisfaction, loyalty and retention at every stage of the relationship and to deploy and integrate customer requirements for performance into the processes and internal performance metrics of the organization.
The author welcomes any questions you may have about this thought piece.
Karin A. Ferenz | Principal