Five Steps to Superior Consumer Segmentation Research

The word segmentation is one of the most overused words in the marketing world today, and arguably one of the most misunderstood. In its most general sense, clients often think they have done a segmentation when they simply create groups of customers that are defined by some business metric. Typical groupings include ones such as “high value customers”, “millennials” or “lapsed purchasers”. Such groupings do not represent segmentation in the strict sense of the word.

True segmentation research involves the use of statistical tools to create groupings of similar consumers that are highly differentiated, accurate and reproducible. That said – statistical segmentation is only as good as the data available for analysis (the inputs), and only useful if the client understands and takes sustained action on the learning. Below are brief descriptions of five guidelines for superior customer segmentation research.

  1. Do your initial homework

It is critical to fully understand your objectives for the research, the business challenges they face and the context in which the learning will be shared and used. While this sounds self evident, it is often not done or only partially done. Superior segmentation research involves a formal stage of discovery in which feedback is solicited from key organization stakeholders. This feedback can be gathered in one on one interviews and in group workshops. When well done, these efforts ensure that your team is aligned and supportive of the research effort and that the results will be properly socialized in your organization.

  1. Understand what actually motivates consumers – not what they “say” does

In order to connect with the deeper underlying and predictive drivers of consumer behavior in your category it is usually advisable to undertake qualitative research. Employing types of inquiry including psychoanalytics, ethnography and semiotics with small groups of consumers can help uncover the deeper unconscious drivers and motivators of consumer behaviour in your category. Learning from such investigations is critical in facilitating the asking of the “right” questions in the quantitative survey as opposed to having more superficial questions that will not drive a meaningful segmentation.

  1. Ask consumers questions that drive differentiation

Respondents in a survey will always try to provide an answer to questions. Unfortunately, the structure of many types of questions produces results that make segmentation more challenging. One type of question that is routinely overused in segmentation studies is the large rating scale grid where respondents are asked to rate many brands on a large battery of attributes or rate their agreement with a large number of attitudinal statements. In many cases they are unable to give informed scores on all the brands or statements, but they fill in answers anyway. The result of using data from these types questions is a very fuzzy and messy picture of brands and attitudes in the category. The use of choice based non-rating scale questions that do not force your respondents to answer for every brand attribute or attitudinal statement always helps produce more highly differentiated segments at the end of the day.

  1. Use the most powerful segmentation tools possible

The actual output of a segmentation effort is largely dependent on the statistical tools used. A weaker tool will produce a weaker segmentation that often requires a number of adjustments before it appears to provide a useful solution for the client. A powerful segmentation tool will produce highly differentiated segments quickly that are meaningful and reproducible. A Bayesian statistical approach provides a superior solution in all cases. It is based on probabilities, rather than simple clustering derived from a K-Means approach. The Bayesian approach provides much more predictive power as well and forms the basis for powerful simulators and the ability to easily identify segment members in other streams of research.

  1. Fully socialize the results

In order to guarantee superior segmentation research, significant efforts must be made to fully internalize the learning from the research. The presentation from your supplier is not the end of the project. To ensure that segments are fully understood, activities such as workshops and ideation sessions should always be included. Tasks in the sessions can include such things as naming the segments, developing tactics for marketing and communicating to them and creating segment personas. The use of videography to make the segments “come to life” is often a powerful way to summarize the learning in a visual way. All these steps can provide valuable strategic brand advice to your marketing and sales personnel. Finally, superior segmentation should always involve the creation of a powerful simulator and segmentation algorithm so the research “lives on” in your organization.

 

For more information on how you can use Segmentation research to move your business forward, please contact Anne Coulter annecoulter@rsginc.net or Jim Peterson jimpeterson@rsginc.net