Stepping In The Same River

Deep

“No man ever steps in the same river twice, for it's not the same river and he's not the same man.”

- Heraclitus, a Greek philosopher (544 B.C., so let's forgive the pronouns)

For roughly my entire career as a marketing and advertising professional, especially one who has been on the inside at places that have held the data, there has been one consistent inquiry.

"How do we get our (insert metric) up?"

Typically this is clicks, but it's also been opens (see email roles). Sometimes it's view based, other times its conversions, your best bet is probably a hybrid measurement that's mid funnel, and there's even been downloads or user time. You name the success metric or KPI, show me some creative, and I can probably tell you a half dozen things that could positively impact success.

In seconds, without research. It's something of a party trick that comes from decades in the space.

But the question also betrays a fundamental misread of the mission.

Short term wins over your control, especially if they are from something as transitory as a design only refresh, gives you a diet of popcorn -- and a very finite amount of popcorn at that. Especially in a typical marketing and advertising mix of multi-channel touch and communication, or with (and here comes the river) a fluctuating supply of impressions.

(You remember the river, right? It's important. Sorry it took me a while to get back to it.)

Especially in broad campaigns and programmatic plays, the quality of traffic can vary wildly, even among people who aren't scouring the Web for low CPMs. Online publishers are under constant pressure to keep the lights on, and that can lead to unfortunate decisions on frequency. There's also the very real spectre of outright fraud, which is slowly getting beaten down due to better tech, but, well, not all at once.

So what's needed is testing. Constant, disciplined, with an emphasis on reporting, preferably with your analysts having a strong dose in significant confidence levels. With a plan that attacks structural differences (i.e., offers) as well as surface changes, and KPIs that don't change with the weather.

Oh, and when you think you've determined, for once and for all, a stronger practice?

Well, that's when you have to run a back test... because the river has changed.

And if you need help navigating those waters, I'm happy to guide your boat.