A large telco client requested a segmentation study of its entire fixed service residential consumer base. It had already developed segments that were meaningful to the business, but there were two problems:
- The segments developed relied on information held for only one third of its residential customers.
- The company was unsure whether the segmentation it had performed was statistically valid. In particular, were the segments different from each other, and were the customers within each segment similar?
Datamine was able to solve both these problems. The first was solved using a technique called “reverse engineered clustering”, a technique that begins with clusters (segments), and then identifies the structure and reliability of the clusters using the data.
Solving the second problem required some segments to be amalgamated, some to be split, and some customers to be moved between segments. The result was a statistically valid segmentation that made ‘sense’.
As a result of this analysis, the telco was able to gain a better understanding of their entire customer base, not just the third of it that had previously been segmented. With confidence that the new segmentations accurately represented their customers, the organisation was able to execute better-targeted marketing campaigns.