By Barbara Silverman
I find myself a bit entertained by this new and exciting practice of A/B testing. I see it in context for evaluating the effectiveness of email blasts and web pages, mobile apps and who knows what else!
Now, I’m not taking issue with the idea of testing. It affords terrific opportunities to create “intelligent” marketing campaigns. Otherwise, we are left to instinct and guesswork that can be effective but more often than not, leaves something to be desired.
What I find amusing is that I learned about A/B testing back in the 70’s and I know it began way before then.
I was waiting to be interviewed for a new job – a production artist at a direct-response ad agency. As I gazed around the lobby I focused of a few framed newspaper ads. They appeared more like articles than ads. Bold, screaming headlines predicting financial doom, packed to the brim with text. Not a breathing space to be had. At the bottom, the foundation of the ad, was a coupon. This part was the salvation. “Buy this newsletter and you will be saved from financial run.” Check boxes, fill-ins and dashed lines surrounded it, clarifying the directions to clip it out and send it in with your check.
Subtlety tucked away in a corner of the coupon lay a piece of nonsense. Yes. A code. When this coupon would get processed and the advertiser would know which ad this coupon came from. The compiled returns would quantify the success of each derivation of each versioned ad. Testing could include publications, headlines, offers, length of copy (back in the day, a 16 page direct mail piece often out-performed the shorter read) – whatever was tested. Winners would become controls for additionally tested criteria. Then tested again, and again. Obviously, I was viewing a big winner.
Of course, I didn’t understand any of this as I reviewed this wall-of-pride. My naïveté designer’s eye just saw ugly. Ugly, ugly, ugly. I considered fleeing the scene of this horrendous crime when I was called in for my interview.
And so began my education: Pretty design does not (necessarily) a good advertisement make. There is an old adage in direct marketing: the three elements of a direct mail test, in order of importance are
Does that hold true today? We might ask our friends at Apple. Anyway, I digress. Except to the point that testing can prove or disprove any of this. And testing is a good thing. And testing has been around for decades. And good marketers test. Testing has been happening for a long time and is still a hot topic today. So test.