Case Study: Recommend Segments of Respondents who have Similar Consumption Patterns

StatGenius has recently launched an intuitive feature called, “Recommend a Segment” that allows a market researcher to find segments in their SurveyMonkey or Qualtrics-collected survey data. Statisticians call this “cluster analysis” because it is used to find groupings of data points (aka ‘segments’) in a survey file. Regardless, with StatGenius, you only need to know where to click the link for, “Recommend a Segment” and the software does the rest for you.

Jill A. is a research manager, working for a boutique advertising agency on behalf of a  burger-themed fast-food restaurant chain. This restaurant wanted to drum up support for a new line of frozen yogurt drinks – and wanted to best understand “which purchasers would purchase which shakes”. In executing this project, Jill ran a questionnaire in SurveyMonkey, and then used StatGenius to analyze the segments that resulted.  

In the past, Jill had to outsource this project to a senior statistician, however she decided to use StatGenius to get her the exact same results – but without the wait time. Using StatGenius, Jill simply selects the variables that she wants to be considered for the segment. This was easy, because she knew her client had spoken repeatedly about:

  • Amount of Sugar Preferred
  • Preferred Flavor
  • Calories
  • Income
  • Education

That was it! StatGenius does the rest of the work. Although Jill’s seasoned statistician typically needed to enter numerous parameters into SPSS (taking hours to complete) – StatGenius is smart enough to analyze the clusters without human intervention. The results came in seconds, and Jill was stunned. She found three distinct groupings of consumers that had very specific preferences, as described below:

  • “Mango Sweet Tooth”: these high sugar, high calories, Mango flavor lovers make  28% of all yogurt consumers.
  • “Fresh and Light lovers”: a segment that makes 18% of all yogurt consumers, who prefer a lower amount of calories and sugar, as well as Citrus flavor
  • “Vanilla High Market”: High Income, High Education (Master’s or Doctorate), Vanilla lovers, that make 10% of the consumers, and are an attractive segment due to their particular Purchasing Power.

Jill’s client was thrilled at the results, and used two of them to develop specific flavors and promotions across their restaurant chain. And Jill was happy that she was able to complete segment analysis without the use of her outsourced statistician.

What is Cluster Analysis?