In these days of multi-channel purchasing we know that it’s often more than one contact that drives a purchase. You get an email telling you to watch for your new catalog which gets paged through when it arrives a day or two later, then as you’re seeing what your friends are up to on Facebook, an ad pops up from that same cataloger. A purchase is made and these days it’s most likely done online. So what do we do with all of these “internet buyers”? Test them of course! There are some people out there that are very unlikely to purchase from a catalog and others that want or need to hold a catalog in their hands to shop. Knowing the difference will mean savings on print and postage. How do we test? First, use your matchback results to determine the true Web Only buyers, those people that have purchased but can’t be matched back to a catalog mailing. Then use purchase recency dates to break the universe down into segments - we’ve tested in 12m date buckets, but you also want to be sure you’ve got enough quantity to have a true read on results.
At the very least, keep your 12m Web Only buyers segmented separately because these buyers will likely warrant a different contact strategy than the older Web Only buyers. The next step is to do A/B split holdout testing in at least two mailings so you can watch for consistency in results. Do a random split on each of your Web Only segments, mail one half and retain the other half as the holdout panel. The holdout panel feeds into your next matchback along with the mail file and the results will show how the mailed segments performed versus not mailed. Here is an example: These results were from a first test, and subsequent testing looked very similar for this particular client, so we built our current contact strategy using these results. We continue to mail the 12m Web Only buyers because all segments are performing well over breakeven. We’re mailing the 13-24m Web Only buyers less frequently, and suppressing the 25+ month Web Only buyers in all but one mailing. For this client, the savings are about $32,000 per year based on their cost in the mail.