Predictive Marketing Tip #7: 
 Last but not least, the part 
where imagination comes in...

I invested six prior blog posts in tips for valid predictive testing. The process is so darned scientific — which is what accounts for the “valid” part — that some might by now wail, “But advertising is a creative business. Enough with your rules. When do we get to be creative?”

Right now. In Step 5 I urged you not to ask questions, but to create and place customers in real-world situations, where you can watch what they do. Creating situations and passing them off as “real-world” is easier said than done. This is where creativity comes in. I might add that cooking such things up is wicked good fun.

Here are three ingenious examples:

• Marketing guru Paco Underhill learned how changes in store layouts affect sales by moving things around at different times of day, and settling down out of sight with a clipboard to watch and take notes. Caution:  If you decide to try this for yourself and you favor a particular layout, you cannot assume that your observations will be bias-free. Have someone who doesn’t care one way or the other, and who doesn’t know what you prefer, do the hiding and noting for you. (I highly recommend Underhill’s book Why We Buy.)

• Psychologist Richard Wiseman wanted to know what would make people likely to return a lost wallet. Rather than ask people what they would be likely to do, he patiently pretended to drop his wallet — many, many times. People were most likely to return it, cash and all, he found, if there was a picture of a baby inside. (I highly recommend Wiseman’s books Quirkology and :59 Seconds.)

• When a RESPONSE Agency client wanted to know whether rebranding a company they had acquired would increase sales, we created three versions of direct mail for them. The versions were identical except for the brand: one came from the acquired brand, one from the acquiring brand, and one from a third, unknown brand. We divided a valid sample of thousands of customers into three random groups and sent one version to each group. Each version was further subdivided by offering different deals. We learned something of extreme importance to the client: the brand made not one iota of difference in sales — but the offered deal, and only the offered deal, made a huge difference. Upon retesting and attaining the same results, it was clear that neither the acquired nor the acquiring brand had the power our client had fantasized. This does not mean that brands don’t matter. It meant our client’s brand did not have the power they thought it had. And that when our agency continues harping on the importance of the offer, it is for good reason. (I highly recommend my book Prove It Before You Promote It.)

It takes ingenuity to set up a valid test. Consider this your license to have fun at work.

—Steve Cuno