Direct-Mail Testing Strategy

By Shari Altman

What is the biggest direct response marketing medium in terms of total spending? Between catalogs, envelopes and assorted other formats, more money is spent on direct mail than any other approach. So if you are successful with DRTV or DR radio, direct mail may well be worth considering to grow your customer file.

Acquiring customers means being able to reach prospects. In direct mail, this means renting lists. So your first decision point is to determine if there are enough targeted lists of prospects for your brand. The key here is the word targeted. Selecting names based on things like age or income is useless if the prospects haven't also shown by some behavior that they are likely to be interested in what you have to offer. Remember actual buying behavior trumps attitudes or demographics every time in terms of results.

LOOKING AT THE TEST
If there are enough targeted prospect lists available, it's time to decide how many to test and which ones. Organize your prospective lists into categories (e.g., magazine subscribers, catalog buyers, other product buyers, etc.). Ensure that you test at least two or three lists from each category you want to test. If you only test one list in a category, you will not be able to conclude anything about the category, only about the single list. Your goal in any direct-mail test plan should be to move your program forward over time, which means learning about which categories of lists perform best. Each list test should be a minimum of 5,000 names for statistically valid results.

Once you've determined which lists to test, consider other tests such as offers. Offer tests in direct mail can be executed with more statistically reliable results than in DRTV or DR radio, because you can truly select random segments of lists to assign to each offer test panel. All the mail goes out at the same time, so there is no concern for environmental factors impacting one segment of the mail differently than the other. Because of this, offer tests make great sense in direct mail.

Distribution of offer tests among your lists can be done in a variety of ways. The two most popular are to rent larger numbers of names for some lists, and divide these all in two or three for each of the offers. An alternative method is to select a random group of 10,000 names or some culled from all your other lists, and to re-assign those to the test group. Which method you use depends on how many names you plan to mail for your test overall. Your list broker and data prep house can help you determine which approach will be best.

When you are all done, look back and be sure you aren't testing too many things. You need to ensure that for each test you have sufficient results to make valid decisions. Usually this means 50 to 100 orders minimum for each test panel.

Direct mail can be effective for growing your customer file and gaining knowledge via testing. Planning and managing your overall testing plan and strategy are key to getting useful statistically valid and reliable results.

Shari Altman is president of Altman Dedicated Direct, a DR marketing consultancy based in Rural Hall, N.C. She can be reached at (336) 969-9538, or via e-mail at saltman@altmandedicateddirect.com.