Sometimes “one size” of user experience does not fit all. Propensity modeling allows you to find dimensions of user groups, allowing you to predict how different groups will respond to what you are offering: Enjoy your product, visit your website, request your service, etc.
LOGISTIC REGRESSION is a machine learning method that is relatively simple to run and highly effective. Using it I predicted purchase probabilities for a new VMWare product line by different market segments (e.g., Government, Small Business, Enterprise, etc.).
This simple graphic shows how it could work with human users. Some user groups have a low propensity to appreciate your design, others have a greater propensity, and so on. Use this to figure out whom to please and how to please them!
