Direct Mail Testing Today is So Much "One and Done"

As a 20-year veteran of direct mail I have seen many approaches to testing.

Over the last five years, there has been an enormous shift in overall attitude—at least as evidenced by the precipitous decline in test to continuation ratios. Now, I doubt many mailers—or brokers, for that matter—will agree with this assessment. The former creates the measurement equations and the latter co-signs them.

No one expects a test to make money. But it is a waste of money to test a list once and drop it if it does not index well. Mailers should be more judicious in their test picks. After all, mailing a list at least twice will give a much better idea of future results

Why? One reason is the inherent "unfairness" of most merges. If a mailer is doing a priority merge, and they are bringing in a large number of lists/names, multi-buyers
are going to be assigned more preferably to the higher ranked lists. Even with a random allocation merge, it is still left to chance which multis get reallocated to which source file. If a list only has 25-50% single buyers/unique names, the quality of the particular reallocated multis can vary widely.

Mixing test files in with proven continuation files can also skew test results unfavorably. While it can be argued that all tests in that merge face the same obstacles, the argument does not save the process from the scrutiny I am giving it. Each of those tests is facing an uphill battle.

Another challenge is the ever-increasing number of sources from which names are derived. Are mailers setting themselves up for failure by dumping all lists into a merge rather than segmenting them by source? Is it fair that all tests get the control offer or something very similar? Do Internet names fail because they are sourced from the Internet. Or, do they fail because mailers have failed to create new offers that speak to an internet mindset? So many new names are coming to market from the Web. It would be an utter waste if mailers could not find a way to mail to these names successfully. It might be argued it is too expensive to design a creative in this manner. However, with the dearth of new postal files on the market, it may be too expensive not to take this action.

When I started in this industry many more mailers tested files at least twice. Some larger mailers I work with today still do. Now sadly, the vast majority are repeating the "One and Done" mantra. Brokers may want to be more discriminating in their recommendations for tests, so their clients can commit to at least two test mailings of a list. Mailers need to start seeing test to continuation ratios go up to one in four or even one in three again. It is not unusual to see these ratios at 20% or below, even on traditional direct mail sold files. Then and only then we might see the demise of a mindset I do not believe serves the direct mailers very well.

Tom Mack Jr. is vice president of list management at Infocore Inc., Carlsbad, CA.

Like to submit an article to List Talk? Contact Jim Emerson at jim.emerson@penton.com