Testing Still Essential in Marketing

Feb 24, 2009 12:34 AM  By

Testing is a direct marketer’s secret weapon. No other means of advertising can be tested, refined, and tested again with the same mathematical precision.

Once you have rock-solid, up-to-date, detailed customer and prospect data, you can test both the various components of a direct mail campaign—offer, copy, design, and list—but also the various delivery and response methods (direct mail, e-mail, personalized landing pages, phone).

What Should You Test?

Lists: Experts agree that the lists you choose are the single most important component of a direct marketing campaign, comprising about 40% of the impact you’ll have. You can test compiled lists, exchanged lists, or subsets of your house list. Work with a competent list broker who can help you understand more sophisticated list options like merging and purging different lists to create a “new” universe.

Offer: The offer is the second key to success, comprising about 30% of the impact. Test only one component of the offer at a time—the “ask” amount; the price or interest rate; the discount; the premium, etc.

Creative: Together, the copy you write and the design in which you package the offer comprise about 30% of the impact on your results. Creative is made up of a number of sub-components: package design, personalization, type of postage, testimonials, length of letter, use of color, letter signer, adding or subtracting a buck slip or lift note, subject header in an e-mail, etc.

What Quantity Should You Test?
Now that you’ve identified which packages and campaigns to test, you’re next question is likely, “How many pieces should I mail?” When you need a quick number, there is a simple universal rule of thumb called the “Rule of 100.” This means that you should test-mail enough pieces to generate a minimum of 100 responses. To determine the number, you will need to work backwards from the response rate percentage you anticipate (or have last obtained with the control). For example, if you expect a 2.5% response rate, you will need an audience of 4,000 to test (4,000 x .025 = 100 responses).

Compare Apples to Apples
The only way you can tell if one test has beat another is through cost-versus-revenue analysis. For example, let’s assume you are testing 5,000 pieces featuring new creative, against your normal 100,000-piece mailing of the control. In order to accurately evaluate the results of the test vs. the control, you have to calculate the price of mailing the test on the basis of a100,000-piece mailing, even if in the actual test you’re only mailing 5,000 pieces.

  • The cost of putting 100,000 control packages in the mail = 67 cents each
  • The cost of putting 5,000 test packages in the mail = $1.34 apiece
  • The cost of putting 100,000 test packages in the mail = 72 cents each

In short, you have to compare the rate of return and the $.67 vs. $.72 price in order to judge which package really performs best.

The bottom line is testing is an essential activity in direct marketing, and it can be a fun way to connect with new and existing customers.

Crystal Uppercue (cuppercue@euservices.com) is marketing manager at EU Services.