I recently listened to the CD of a talk given by Joseph Sugarman, founder of JS&A and widely recognized as one of the greatest mail order marketers of all time.

JS&A is the company that sells those Blue Blocker sunglasses you see advertised on TV and in magazines.

The glasses really work, by the way. Put a pair on, and everything blue is converted to a shade of gray.

Anyway, during the talk, Mr. Sugarman mentioned that he often tested three or four different ads for the same product ? and in some cases as many as 10 different versions!

?Typically, nine of the ads would fail but one would work spectacularly well,? said Mr. Sugarman, ?The profits from that one ad would more than cover the losses from the other nine.?

Do you do that ? create multiple ads and then test them to see which works best?

Or do you — like most businesses — create and test just one ad ? or one postcard ? or one e-mail ? or one sales letter or direct mail package?

If so, you are significantly reducing your odds of getting a winning promotion?.

The reason is that not all promotions work. In fact, most don?t.

Say one out of four promotions is a winner. And that?s being optimistic.

Jerry Huntsinger, a well-known copywriter in fundraising, once told me ?9 out of 10 of the things I do don?t work.?

And in his speech, Joseph Sugarman reported similar results ? sometimes having to write and test 10 ads to get one winner.

But let?s stick with the ?one winner out of every four tests? figure for now.

Based on those odds, if you run just one ad, or mail just one version of a sales letter, your chances of hitting a winner are only one out of four ? and the odds are 75% that your marketing effort will bomb.

What commonly happens is that a business decides to ?try? direct mail ? sends out a poorly written, amateurish letter or postcard ? and when they get no response, they proclaim that ?direct mail doesn?t work.?

Sure it doesn?t ? tell that to Nightingale-Conant ? or Boardroom ? or Publishers Clearinghouse ? or Day Timer.

On the other hand, if you create and test four different ads or letter versions, the odds are in your favor that at least one will work and be profitable for you.

My rule of thumb for improving direct marketing results is: Look at what the big players ? the successful direct marketers ? are doing. And do what they do.

And the one thing every successful direct marketer has in common is ? they test. A lot.

What do they test?

Headlines ? outer envelopes ? direct mail formats ? copy approaches ? sales appeals ? mailing lists ? prices ? offers ? guarantees ? terms ? anything with the potential to generate a big lift in response rates. Or even a small one, for that matter.

Does all this testing make sense?

On one such test, a marketer increased response to an e-mail marketing message by 50% ? just by changing the subject line.

In another test, a software company increased orders from a direct mail package tenfold ? simply by varying the wording of the offer.

And a computer school doubled the response rate to its newspaper advertising when they added the offer of a free career booklet.

Does all this testing make sense?

You bet it does!

Imagine ? just by changing a few words on a piece of paper or a computer screen, you can double your sales ? revenues ? and profits.

If there?s another area of business that gives you that kind of leverage, I?d like to hear about it.

One other point?.

In direct marketing, no one can predict with any degree of certainty which ad or mailing is going to work.

You only learn what works by testing and keeping track of the results.

You may have your subjective opinions about what you like and don?t like in advertising ? we all do ?.

But in direct marketing, you simply can?t argue with results.