Polk Unveils Database, List Modeling Process
TotaList XL is an enhanced version of Polk's large consumer household database, adding automotive ownership, lifestyles, interests and activities data. It features more households, new data, new segmentation methods and a new Internet delivery system.
Polk's new list-modeling process, SuperModels, lets Polk work with a customer's database by integrating proprietary model-development procedures to it and then automating the method by which the procedures are combined to improve response rates and deliver a higher quality of names.
"Automating the system allows a more robust model to be made in a shorter amount of time," said John Kallassy, vice president of Polk's direct marketing product management and analytic services.
Throughout the SuperModels process, Polk combines new tree segmentation techniques, such as CHAID and CART, with predictive modeling methods and sampling techniques. It also integrates and automates Stepwise Bundling, a proprietary technique that allows the sequential modeling of portions of the database to uncover higher-performing names with each succession.
Through the identification of hidden prospect segments, SuperModels can provide a higher response rate and a larger prospecting universe. "We give customers a clear picture of what their customers look like, how they act and why they buy," Kallassy said.