Software Product Segments Transaction Text
Normal segmentation packages can read only ZIP code, merchant code, purchase price and other numerics. They can tell that a transaction occurred at a jeweler, for example, but can't determine whether it took place at a discount chain like Zales or a high-end retailer like Tiffany.
Select Profile from HNC, San Diego, can read, understand and make decisions on unstructured text data through its intelligent Content Mining feature. As HNC marketing director Patsy Campbell explained, it can figure out that a number of disparate merchant abbreviations represent the same company.
Industry consultant Katharin Dyer said credit card issuers can currently read text transactions at the merchant level to pinpoint where consumers shop and issue them coupons for those merchants. This text, however, doesn't provide enough information to be used for segmentation or one-to-one marketing purposes.
"Select Profile lets you see in a more insightful way what realms people are interested in, that's very powerful," she said.
SelectProfile extracts data from an issuers transactional database to construct a profile of each customer and segment them into categories. Segments change as the customer profile changes which helps card issuers adapt who they target for particular offers. Campbell claims the software will be able to see life changes. The purchase of a significant amount of baby clothes, for example, signals a trend and segments the consumer as a parent.
"Credit card issuers have a lot of segmentation programs, but they deal in pretty broad segments," she said. "SelectProfile can tell the difference between frequent budget traveler and the first-class budget traveler. It has a much finer ability to segment to the right thing the person wants."