The hypotheses are to represent proposed relationships between customer characteristics and the goodness of the customer, as measured by the quality score. For the generation of an initial list of such segmentation hypotheses, it is necessary to analyze (i) the structure of the market, (ii) market information residing within the organization, (iii) data available through market experts and their publications, (iv) competitive information collected by reviewing competitor websites for their marketing communications, promotions, sales content, and product features, (v) structurally similar industries by reviewing industries with similar organizational characteristics, and (vi) standard, ‘a priori’ segmentation schemes. Once the hypotheses have been identified, they need to be tested for their validity with the data available from internal and external sources.
Developing variables and hypotheses is important for a variety of reasons, but its primary purpose is to provide a framework for the customer segmentation research process. Once you have established a clear hypothesis and the variables that you need to test, you can begin executing the intricate process that will help you identify your best current customer segments.
While most companies possess enough market knowledge to predict or anticipate which customer segments are their most profitable, the leaders of those businesses also know that scaling a business is not best left to guesswork or instinct. That’s why, in a customer segmentation process like the one described in this guide, it is critical to develop customer segment hypotheses and variables, and validate them with a well-developed, scientific research process.