Many marketers short-change themselves when they send out a direct mail package, run an ad, or do a commercial on TV or radio without testing significant variables. I don’t know why, because those variables can have a significant effect on response.

I am constantly seeing examples where there is a 20%, 30% or even a 100% difference in response because of what might be thought of as insignificant variables.

The “magalog” for Dr. Bruce West’s Health Alert generated up to a 3.38% response, depending on the list. When the tabloid format was tested, response soared to 4.71%. Tabloid format are now “king of the hill” for selling subscriptions to this newsletter.

I have seen reversed type significantly decrease response, a change in premiums dramatically affect returns and paragraphs added in the middle of a letter increase or decrease response.

The omission of testimonials will significantly impact returns. Another good example of “traditional” direct marketing hard at work is the lift note. These notes can increase response by as much as 30%.

Whenever I create a mailing package, advertisement or direct response commercial, I always create two versions and use a major variable. Normally, it is the creative premise of the theme test—like “fear” versus “greed.” The envelope and lead-in of the letter copy, or the commercial theme, are the variables. With other projects, these variables can be pricing, premiums, the offer, a discount or any number of other items that might be important.

Unless you test with “A” and “B” versions, you could be losing a great opportunity to increase response.

Let me give you two examples:

1. Client #1 mailed 60,000 pieces. His Version A package produced .24%, while Version B produced .41%, or about twice the response!

The creative difference: the teaser copy on the carrier envelope and the first two pages of the letter. Everything else was identical.

If only Version A had been mailed, one could have said the offer was barely successful compared to a 0.15% objective. Version B was clearly more powerful, and is currently being rolled out to more than a million names. Is this a fluke? No.

2. Client #2 had a 1.96% response on Version A and a .92% response on Version B. Accept­able response was 1.1%. If only Version B had been mailed, the conclusion would have been that the project could not be successful in the mail. Because of the test, Version A continued to be mailed in large quantities.

To give another example, I recently changed only the testimonials in a TV commercial—nothing else. Response increased so dramatically, that the cost-per-lead dropped by over 30%!

Direct marketing is not just an art and not just a science, but a little of both. To presume one creative approach or offer is better than another can be a costly mistake. It’s best to test and let the marketplace decide.

Pragmatic marketers constantly test new creative approaches against their current “king of the response hill,” looking for even greater response rates and lower costs-per-lead.