Cross-selling without getting crossed up

It’s a fundamental part of retailing. A shopper picks out a pair of pants in a store, and a good salesperson suggests complementary shirts and socks likely to hit home with the customer.

Getting that kind of personalized knowledge into online cross-selling is another matter. But Blair Corp., a multi-channel retailer of apparel and accessories, is using a new cross-sell recommendation engine backed by analytics data and designed to personalize the cross-sell pitch. As a result, revenue from cross-selling on Blair.com has increased by 70%, the retailer says.

Blair has replaced an in-house-developed product recommendation engine that had been capable of processing only a limited amount of clickstream data and, therefore, producing few cross-sell recommendations attractive to customers. Its new cross-sell application, Intelligent Offer from Coremetrics Inc., is designed to continue optimizing cross-sell recommendations based on changing customer shopping behavior and new products.

IRCE 2008 Conference Multi-Media CD-ROM

The Intelligent Offer application produced a complete return on investment within the first month of operation as it helped Blair present online merchandising offers most likely to win over shoppers, Blair says. “With Coremetrics, we have a complete, closed-loop solution that collects and analyzes data, creates relevant recommendations, and reports on performance so we can accurately gauge how cross-sell contributes to revenue,” says Darren Schott, senior director of e-commerce.

The Intelligent Offer application, implemented in less than two weeks, uses algorithms to generate recommendations based on data on online shopping and purchasing information collected on Blair.com. It then loads that data automatically into the IBM WebSphere Commerce platform that powers Blair’s web site, providing more relevant cross-sell offers while saving time for Blair’s merchandisers, Schott says.