Bring 'Em Back
You want to use this information to develop a profile of people who stayed longer versus those who only stayed a short period of time, according to Briley. "In general, the people who are more likely to come back had a rich history with the company, as opposed to the casual surfer," he says.
Adds Rubin: "The key is not just looking at one variable; you're looking at the interaction from all those data themes and types of data."
Sometimes, however, your data may not be rich enough to support your scoring and marketing efforts.
"In many cases, a client data sample will be too thin to provide statistical validity because either they haven't been tracking properly or they just don't have enough inactives that have converted to actives over a period of time," shares Ruf.
In this event, he indicates, you can score your inactive files by either analyzing current responders to a mailed file or by analyzing your best customers with RFM as the dependent variable in a custom model or household cluster study.
"Ultimately that isn't going to be as good as using a custom model of new actives from the inactive file, but it does give us a targeted, aimed shot versus a shotgun blast at reactivation opportunities," says Ruf.
How Deep to Dig
"When you're done with your model, you're usually able to focus on the top 10, 20 or 30 percent of your population with a high degree of confidence that you're going to be able to find good targets for reactivation," indicates Rubin.
How deep you go into your pool of inactives, however, will be dictated by your model results, budget and profitability. You need to understand the costs associated with reactivating each segment of your inactive file. For example, it may make better sense to mail your best offer to the top 10 percent of the file and reduce the cost of the mail piece for those lower down on the list, offers Briley. "At the end of the day, you want to know what is the expected incremental revenue relative to the incremental cost of [creating] the campaign."