A casino needed their customer segmentation to be scored across their entire membership database. This came with a catch: the segmentation itself was created using survey information only (attitudes, demographics and behaviours), but we could only use the casino database behavioural and demographic information to score customers.
Firstly, we found the ‘killer questions’ which were most crucial in defining the segments and conducted an additional short survey among a number of casino database members. These survey results were directly combined with database information at an individual level. We then utilised a combination of both survey and database information to update the segments, ensuring the new segments were as similar as possible to the original ones, but also including database information. A subsequent analysis identified the most distinguishing database variables to use in reconstructing the new segments. A final algorithm was produced, which the casino used to determine the customer segment of every member on their entire member database.