Customer Retention: Get the Best Response From Your Customer File

What’s the single biggest mistake catalogers make when attempting a strategy to improve customer retention? “We most often see over-investment in the least productive customers and under-investment in the most productive [customers],” says Jim Wheaton, co-founder and principal of Wheaton Group, a Chicago-based data mining practice.

Wheaton offers the following advice to catalogers looking to make the most of their customer files.

1. Focus on identifying customers who’ve ceased to respond because they’re no longer receiving your promotions. The first step is getting rid of the dead weight, says Wheaton. In addition to NCOALink, “nixie” and deceased processing, statistics-based predictive modeling can determine which customers are unlikely to be receiving your promotions. This information enables you to focus on customers who actually can respond to what you send them, he notes.

2. Work towards an optimal contact strategy. “Everything from modeling to test strategies to simulations you use to better your contact strategy will improve customer retention,” says Wheaton.

3. Treat lapsed customers differently. If a customer hasn’t bought from you in a while, adapt your contact strategy. “You can offer different merchandise or merchandise bundles; change the look of the promotion itself; or even send a smaller book with the hope that promotional costs will decline less than corresponding response rates,” says Wheaton.

Special offers are certainly another way to get lapsed customers to respond, but Wheaton cautions that customers who are reactivated in this manner tend to look for similar deals in the future and may again fall inactive once they’re treated like regular customers.

4. Measure incremental rather than total response to your promotions. Wheaton and his colleagues have run tests that show, for certain customer subsets of a catalog housefile, 50 to 75 percent of purchase volume would have occurred whether those customers were promoted to or not. These customers tend to have found the catalog merchandise via the Web, and never give source codes, even after they’ve begun to receive catalog mailings, says Wheaton.

“The lesson here is that testing is going to be the key to determining what kind of customers you have and how to best mail to those customers,” he says. “You ought to logically be mailing to the point where incrementally you’re generating just enough return to pay for the mailing. But if you’ve established that, for some customers 50 percent of their revenue would’ve happened anyway, you’re overmailing.”

5. Matchback processing isn’t the whole answer. The arbitrary nature of matchbacks often doesn’t provide an accurate picture of promotional response, says Wheaton. Measurement of response over time is key to understanding incremental response, he notes further. Within a six-month period of time, a cataloger knows what a customer ordered, when the customer ordered and how much the customer spent.

Since the cataloger also knows when and how often that customer was promoted, as well as the cost of that promotion, it can definitively determine how much profit and revenue came from a certain amount of promotional cost, says Wheaton.

Wheaton can be reached at (919) 969-8859 or via his Web site, www.wheatongroup.com.