Testing Versus Winning

Something Like This

Here's a dirty little secret about marketing and advertising.. a significant percentage of the people who do this kind of work pretty much *hate* testing.

There are reasons for this, of course.

Testing takes time, and discipline. It can be screwed up in any number of ways, many of them just plain maddening, and if you don't catch the mistake, you can do true damage with a false reading. It requires you to be willing to "waste" a significant percentage of your inventory on creative that no one is rooting for. Worst of all, it can take your brand-new work, the stuff that you are exceptionally proud of, and fast-track it to the dumpster, because data just won't be stopped, really. And if you want to be truly doctrinaire about it, once you start testing, you never really *stop*, because it acts as your de facto insurance policy, to ensure that your control is still optimal.

I've had any number of clients refuse to run a test, just because they were so in love with the new art, and/or that dissatisfied with the control. In each and every case, I've tried to push back for all of the direct marketing virtues. In most of these cases, the client stayed with their gut and ran without a test, and (here's where the direct marketing purist in me feels ill) it sometimes really worked out for them.

Note the pronoun there: them, not me.

This is also where a couple of cross purposes come into play. Part of being a marketer is being a scientist, and that science doesn't really have an end goal. The journey is the thing. Creative can always be optimized more, there's always some new clue or option not tried from the data, and the world will give you clues, if you're open to hear them.

The executive can look at this and wonder when the law of diminishing returns kicks in, or question the talent involved from creative professionals who would subject themselves to the long work of incremental steps to optimal. It all seems like something that you wouldn't get from top tier agency work, or a process that would lend itself to automation... but that's never been the way it's worked out for me, or how it seems to operate in the real world.

Now that we've gone through all of the reasons why people don't do it, the reasons why it's still the best way to work: it ensures that you never damage your client. It ensures job security, because you've either got a lift, or you've got learnings that will later result in a lift. (Or marketability for a future client.) It creates either single variable steps that take you were you want to go, or if those aren't driving enough of a data difference to pass statistical significance, bigger swings. And if you're fortunate enough to either work in a position where you can see a lot of tests go through the pipe, or in a cross-medium or category house, one where you can bring in learnings from another field, you can seem a lot smarter than you actually am.

I've been fortunate enough to work in this kind of business for decades, and have never felt "burned out"... because we've tested, and learned, and used the results from that testing to fuel the next chapter in the story. Plus, with technological changes, the ability to beat a control has never been "easier", or more important.

So if you're one of those marketing and advertising pros that considers test to be just another four letter word that's not worth the trouble, or beneath your talents...

Well, actually, stay just the way you are.

Because you might be smarter than me, or more talented...

But you won't be more effective.

And I might need to beat you one day.

* * * * *

Feel free to comment, as well as like or share this column, connect with me on LinkedIn, or email me at davidlmountain at gmail dot com, or hit the RFP boxes at top right. RFPs are always free, and we hope to hear from you soon.