Interviews, insight & analysis on digital media & marketing

NDA Viewpoints: Marketing attribution is broken. Why does the industry not want to know?

Chris Liversidge, founder and CEO, QueryClick

Indulge me for a moment in some scene-setting. UK companies alone are projected to invest almost £15bn on digital advertising this year. On current trajectories, their digital ad investment will close in on £21bn by 2023.

While we’re digesting those large numbers, bear in mind that this is only two-thirds of overall media spend (rising to a projected almost-three-quarters by 2023). For a sense of scale, this means that total ad spending by UK companies is already not far short of half the UK government’s annual defence budget.

Of course, this is the outlay. For every pound spent on advertising, a return is anticipated — a multiple of the original advertising investment. For simplicity let’s assume the average multiple is just five; this means that there is now over £100bn of revenue at stake, this year alone.

Those big numbers look even bigger now. And this makes the ability to optimise marketing investments based on marketing attribution very important, on two levels.

First, the absolute level of investment is significant; no-one in marketing wants to waste precious budget. Theoretically, if the overall performance of digital marketing investments could be improved by 25%, then companies could spend 20% less and achieve the same commercial results.

This year, that would be a saving of about £3bn — about the same amount that an extra penny on the basic rate of UK income tax raises (and you know how excited we all get about that).

Second: the commercial outcome is important. If investment continued at the same rate but at 25% greater efficacy, our hypothetical five-times multiple becomes 6.25-times, representing an extra £28bn of additional revenues. But this is, of course, nonsensical because the market is more or less a closed system: there’s not suddenly £28bn of extra demand.

What would actually happen is that those companies which optimised their marketing investment would see their revenues and ROI increase, at the expense of those that were failing to optimise and which would see their revenues and ROI fall. (Ironically, their first reaction would likely be to increase their sub-optimised investments, in large part to the benefit of the Google-Facebook duopoly, more on which later.)

Possibly you’re aware of these numbers; if not in their totality, then for your own business. And naturally, with such huge investment and results at stake, marketing attribution has been a ‘thing’, a very important ‘thing’, for many years. I’d bet you’d tell me your marketing team does it. So what’s my point?

Attribution is broken

My point is that current attribution solutions don’t work. I’ll say it again: current attribution solutions don’t work. How do I know? Because we asked, and UK marketers told us.

With a hypothesis that marketing managers over-invest in PPC under pressure from senior executives, sales teams and other stakeholders to provide immediate results, QueryClick commissioned a survey of 200 senior marketers from the UK’s top retail, travel and finance sectors. What we found was that:

  • unreliable or misleading attribution leaves almost 90% of marketers afraid to invest in activities with any kind of long-term payback, because of their inability to prove the value
  • fewer than one-in-seven marketers are able to adjust their marketing investments based on attribution insights to deliver the predicted results
  • in the absence of credible attribution insights, more than two-thirds of respondents (67.5%) report that internal stakeholder pressure restricts their option to invest in marketing activity that has a longer payback period than last-click measures

That current attribution solutions are, apparently, derided to such an extraordinary degree is one thing. What has come as a most unwelcome surprise to me is that nobody in the industry seems to want to know.

Appalled by the slamming UK marketers gave their attribution solutions, and mindful of the enormous levels of wasted investment that must be occurring as a result, we highlighted the key results to every marketing title in the UK.

With the welcome exception of New Digital Age and one other, the response was a deafening silence.

Animal analogies abound: attribution is the elephant in the room, people are making like ostriches and burying their heads in the sand, people have rolled up like hedgehogs until the noisy threat to their well-being — me! — has given up and moved on…

If I were a conspiracy theorist, I might have concluded that publishers were terrified of upsetting the digital advertising duopoly for fear of possible business consequences.

After all, the survey showed that when marketers lose faith in their attribution solutions (such as the market leader Google Analytics 360), they default to immediate payback measures, notably PPC, to the significant benefit of, er, Google. Could that be a reason why no-one wants to talk about this?

But I’m not a conspiracy theorist. The survey results indicate that marketing attribution — hugely important, widely discussed, supposedly mature, vitally important marketing attribution — is actually a shambles and a charade.

So what I actually think is that the inadequacy of marketing attribution is simply too painful, too complex and too far-reaching for the industry to contemplate.

Too painful? Things don’t always turn out well for pointer-outers of elephants in rooms. And too complex?   Possibly. The great promise of digital marketing was that everything would be measurable, and value attributed.

In reality, of course, the interplay between digital channels, and between online and offline, TV, radio, and every other touchpoint, is so complex that it’s beyond the capabilities of mere humans to process or analyse, or even commentate on. It’s just something that we do so we can say we’ve done it, like washing our hands before eating, while not for a moment deceived into thinking we’ve removed all the germs. And too far-reaching?

When billions of pounds have been wasted, inquiries tend to result, with fingers pointed, if not at individuals then at functions and roles, and no-one in marketing is likely to invite that.

So what will happen? Or more, accurately, what is already happening?

Well, attribution is one of those areas of marketing where, to address its complexity and possibilities, machine learning is destined to make a real impact, since current techniques are evidently inadequate. And, while that moment can’t come soon enough, quite probably it will come without fanfare, portrayed by the industry as an obvious evolution of what went before.

Maybe, that way, the elephant in the room won’t ever have to be pointed out, left to just get up and leave of his own accord, never to be spoken of by the industry ever again.   And especially not to anyone from Finance.

Maybe. But these things take time.  The elephant looks likely to sit there for a while yet, while marketers knowingly spend billions of pounds they didn’t need to.