“Marketing to unknowns is continuing to bedevil the industry,” says Lou Mastria, vice president of public affairs for Westminster, CO-based customer acquisition and retention firm NextAction. “With the induction of multichannel marketing, you’re continuing to add people to your house files who you’re not exactly sure where they came from.”
Indeed, multichannel marketing is helping catalogers and direct marketers augment their house files. But quantity is not better than quality, direct marketers have learned, when mailing to their in-house lists. And the same goes for e-mail lists, especially when consumers can sign up to be on your lists just because they are interested in an offer that doesn’t necessarily target them.
A multichannel merchant can use modeling to build a targeted e-mail list, however. Mastria outlines a recent predictive personalization project NextAction handled to test demonstrated interest in an e-mail for one of its clients, a cataloger of scientific gadgets.
The goal was to segment the client’s house file based on the type of product each consumer was most likely to be interested in purchasing. The client determined the product categories that it want to promote via e-mail. NextAction built custom product-level models designed to predict the propensity of the consumer to purchase from specific product categories, analyzing the buying patterns of each record from both the marketer and across the NextAction co-op databases.
By analyzing each consumer’s scores on the product-level models and the relationship of the scores to each other, a highly relevant product mix was selected for each consumer using either customized landing pages or dynamically generated e-mail messages.
The results were a 46% increase in response, a 61% increase in conversion, a 146% increase in average order size, and an increase in revenue per e-mail of 270%.
“What that means is the rule of thumb of personalization and segmentation that has been around since the 1800s continues to hold true today,” Mastria says.