Boldt: Is that why direct mail tests work?
Cuno: Sure, you don't even have to hit that large of a group to get something predictive. You want to test it a couple of times to make sure you didn't get a fluke, but it's amazing what you can learn by watching people. Direct mail is the perfect place because there's nothing in the mail that tells you it's not the real thing. Offer testing, headlines, etc. It goes way back to John Caples ... who knew that the headline "How to Fix Cars" would do 20 percent better than "How to Repair Cars"? And with predictive behavior, you can then roll out that campaign with confidence.
Boldt: Based on actual results, what is a type of creativity that really works versus one that doesn't?
Cuno: There are flukes, but clear usually beats enigmatic. Usually the latter is a copywriter showing off. But the fact is the market generally is not going to work particularly hard to decipher what you really mean. Whenever you find yourself writing a line and saying, "Wow, that line is good!" then it has to go. The product should shine, not the copy. Falling in love with your own copy is a warning sign. You want to be benefit-oriented, and writers have a hard time with benefits versus features.
I also like to apply the "at a glance" test. They're not going to dig into your ad. They need to find out three things: what's for sale, that it's targeted to them and what is the major benefit to them. They have to get that at a glance; that's what will motivate them to keep reading [and then respond].