List Testing - Parameters & Strategies
No one will dispute the fact that testing lists is important. The list has the greatest impact on the success or failure of the mailing. "Whereas creative tests may cause 1.5-percent to 3-percent variations in response, the right or wrong mailing list can sway results by a thousand percent!" suggests David O. Schwartz, president of 21st Century Marketing, a list broker/manager in Farmingdale, NY. But how much should you test and where should you look for list test ideas?
The amount of list testing a mailer should do is a direct function of the maturity of its direct mail program, Schwartz explains. "If it's a mature program, I recommend 25 percent to 30 percent of budget for testing. In a very mature program, maybe 20 percent is enough," he says.
Realize, of course, that proven lists—those you've mailed successfully before—as well as your housefile, are going to perform better on the whole than any new outside lists you test. But, as Schwartz warns, you must constantly test lists to "bring new blood" into the customer file or your business will die off due to attrition. "Most mailers should be testing more. The problem is it's simple human nature to stick with what looks good at present," he explains.
Leland Kroll, president of Kroll Direct Marketing in Plainsboro, NJ, agrees that due to budget issues, many mailers don't test enough. "A handful of names a year is not sufficient," he notes. "It's always important to keep testing in order to keep the prospecting pool expanding." At least 15 percent to 20 percent of a campaign budget should be dedicated to testing lists, Kroll says.
Setting Up a Test Grid
A common misunderstanding or misconception among mailers is that you have to mail 5,000 or 10,000 names in a test because that's the rental minimum. Not so, says Schwartz. "The law of statistics doesn't care how many you mail—only that you get enough of a response for the data to be statistically projectable."