Chris Andrews, head of advertising technology at Wake the Bear, looks at the framing of measurement and effectiveness and how to avoid getting it wrong…

There’s an abundance of irritating cliches that do the rounds in our industry. 

Not just the flavour of the month stuff that gets the eyebrow arched and that must, by law, be followed within ninety business seconds with “maybe it’s the year of the mobile!”

Or the riff on the famous Economist “data is the new oil” headline which was liquid gold for panelists and commentators tripping over themselves proclaiming that it was in fact some other kind of precious commodity/mucky and malevolent ooze, depending on your tastes.

My personal bugbear is the easiest club to pull out of the bag to stimulate a good old-fashioned brand vs performance or “offline” vs. “online” slugfest is John Wanamaker’s famous adage. Sing along at the back, you know the words: “Half the money I spend on advertising is wasted; the trouble is I don't know which half.”

Platforms claim it in their defence to justify shonky last-touch attribution and others claim it in the fight against said platforms to cast aspersions on the validity of attempts to metricise the unmetricisable. Each loudly proclaiming that it couldn’t be their half because look, here’s some evidence that proves me right with no hint of irony!

All of this is to say that it has probably succeeded in framing the recent history of measurement and effectiveness as a dumb (and false) dichotomy between divining value from some numbers showing how many times some lines of code fired in sequence and the vibes theory of media planning. 

Annoyingly, I’m falling into my own trap by using up precious word count on it, because what it masks is a lot of the genuine innovation that has tried to bridge the gap between marketing as art and marketing as science.

Regulation, legal battles and tech infighting

I’ve seen the coming and going of tools, technologies, and standards that were destined to be the silver bullet only for one factor or another to knock them out, like tears in the rain.

Sometimes that was regulation. GDPR took out swathes of matchable indicators and the ePrivacy Directive started trying to deprecate the cookie way before it was cool. 

Fast forward to today and now Google has warmed to fingerprinting (for themselves at least) there’s every chance it’ll just get regulated out of existence by 2026. Or it might be written into UK law and somehow ennobled into the House of Lords in the service of economic growth. Who knows?!

Frequently it’s tech companies fighting other tech companies. But then without iOS14.5 and ATT, would Conversion APIs be where they are now? Would the proliferation of alternative first party IDs exist if the market thought it could keep getting by on cookies, under the false spectre of deprecation or not?

Sometimes the breaker was market demand. One day I’m sure beacons might make a reappearance, exhumed from the same tomb that QR codes once lay.

All of this is to say that maybe some of these tech specific solutions will be around in over five years, and plenty will be distant memories of a bygone time. Relying on them sticking around to give you your long-term measures of success is a fool’s errand. Equally, boldly proclaiming that your proprietary ML algorithm trained on the cycles of the moon on its own will solve measurement, is equally daft.

To give them their appropriately weighted credit, incrementality, MMM or econometrics (lite or full-fat) look to have been among the most durable. They’re obviously flawed in their own ways but generally their proponents gladly own them. Plenty more people put their stock in Binet and Field’s ‘Long and Short of It’ and at the risk of looking like a hypocrite against the rest of what I’ve written, they are obviously objectively right to do so because it allows for nuance.

"The fetid stench of last-touch attribution"

Either way, every time the industry seems to find something that works technologically, it struggles to endure, and a critical mass of the market goes back to safe havens of shit, pointless metrics.

That might be to the fetid stench of last-touch attribution luring people to drown in its swamp, or the misguided and damaging race to reduce CPMs on the buy side to the bottom in the reckless pursuit of more and bigger, as though the market were an infinitely scalable wall that can take an infinite volume of effluent being thrown at it. 

That’s in part because measurement which means anything substantial to advertisers, in a weird way, defies standardisation.

In the outcomes era at least, that should be seen as a good thing. It’s supposed to be hard to measure this stuff (though don’t confuse that as “hard to articulate” because that’s a whole other question). There’s pleasure and intellectual challenge in learning from the old and applying to the new. The challenge of demonstrating value has driven innovation, even if it’s not evenly distributed.

At Wake The Bear there’s an unspoken “don’t trust a silver bullet” rule, instead using our bullshit cards when someone claims to have one. This manifests itself in not over-relying on a single source of truth. 

We know full well that metrics in isolation tell limited stories and that a bespoke, triangulated and relentlessly commercial approach, and one rooted in solid, enduring, principles that give a basis to consistently proving and iterating, is almost always the way that drives sustained success against the metrics that matter to our clients - their business growth.

What does this teach us? Well, at least half of your measurement methods might end up not working. Don’t get in trouble by not knowing which half.