News
Yes, There Will Be a Test
Return to News Items February 1, 2006You have a newly designed direct mail package. A new list. A new buy-one-get-one-free offer. But most of all, you have a big budget to roll out your new direct mail campaign.
Before you do, though, you should also have a test. Sending out a completely re-done direct mail campaign could be a costly error. Instead, you should test each new element - one aspect at a time - against your control package, which is the mailer you've been using with a known response rate.
And that's the beauty of direct mail: its individual elements can be tested and evaluated. It's easy to determine, within a relatively short time frame, whether your mail piece will be a winner.
What to Test
The best factors to evaluate are those elements that will have the greatest impact on your response rate. From the creative format (whether it's a letter, brochure or card), to the copy (is it long or short?), to the mailing lists and pricing, all can provide valuable test results.
How to Test
Test only one variable at a time! If you try to test more than one variable at a time, you've tested nothing, because you won't have any way of knowing which variable worked and which didn't. Use key codes, which are numbers or letters to help you identify the mail piece. Make sure your codes tell you which list the responder came from. Or better yet, keep a cheat sheet of your keycodes handy.
If you're testing a list, it's important to select the names randomly. Mail your tests at the same time. Don't make timing another variable. Use sufficient test quantities. A good rule of thumb, in order for your results to be evaluated, you should receive a minimum of 50 responses to the mailing. On a 5,000-piece mailing, this is a one percent response rate, which is considered very respectable.
Finally, even after you've chosen the perfect package, continue testing to improve your results.
Return to News Items