Getting the Measure of Attribution

Getting the Measure of Attribution

Why the obsession with online metrics and the quest for true media attribution are leaving marketers jaded and confused.

As the digital media marketplace matures there is an ever-increasing array of means by which we all seek to quantify and understand what we are buying and exactly how it is working. In fact, after seven years working within the online planning and buying department at Space and Time Media, I’ve realised that there are more metrics for measurement now than have ever existed at any point since the advent of online advertising as a mass medium.

CPA, CPC, CPM, CTR, CPA, CPE, CPV, impressions, views, reach, frequency…and that’s just considering metrics from buying the media. Then also consider post-click actions, post-view actions, engagement, dwell time, bounce rates, exit pages, cart abandonment rates, sign ups, brochure downloads, visits to website by device, downloads, appointments, phone calls, retweets, mentions, shares… aaaand breathe!

In this age of omnichannel marketing there persists a major problem for all marketing professionals, and that’s proving which, of all their marketing efforts, actually delivers results, and more importantly understanding how best to interpret and report those results.

Within my role at space and time, I am presented on an almost daily basis with the latest measurement tool, attribution model, or analytics tech that is going to BLOW MY MIND. Whether using Google Analytics or call tracking analytics such as Infinity, or adserving tech such as doubleclick or mediaplex, or omnichannel data hubs such as Datorama or Adalyser. There are so many tools out there that we must be able to finitely understand everything that has happened, to whom and when.

Finally, our craft is a grown-up science and not a dark art. That over-used quotation of John Wannamaker’s is finally redundant. We know exactly which half of your budget is being wasted.

But in many regards the influx of data can also raise more issues than it solves. If you add together all the sales that each channel has generated, you seem to have sold everything three and a half times. Because, of course, one touchpoint does not equal one purchase. And not all touchpoints were created equal.

The fragmentation of the media marketplace is comensurate with a massive increase in the volume of advertising each one of us sees as we go about our day. Your media agency is using your budget to stack media precisely because that’s how your customer consumes it. So even a simple decision to purchase an FMCG product might now be influenced by half a dozen different messages over the course of a few days- some of which are visible to the agency and some of which very much aren’t.

So, is a Facebook view worth more than a network click? Or if I sold to this person a month ago from an organic link and then they came back again yesterday from a paid search link, which channel delivered the value of the secondary sale? I wish people would stop emailing each other links to our website because I can’t track those. And for heaven’s sake, don’t write down the press phone number and then use it after you’ve visited us online…

For all that we would like media and attribution to be an exact science, and despite the indisputable value and insight that some of these tools can deliver, a huge amount of what we all do still boils down to old-fashioned human insight, experience and instinct.

Something that looks to all the world like a cold, hard number is often actually some sort of a crypto-feeling: somebody in your own department decided that phone calls longer than 30” are the ones that are worth counting, so that’s what gets measured; a team at the IAB decided that a “view” is when 50% of the ad is on screen for longer than one second. Between us, we decided that it’s the video interactions of longer than ten seconds that count, or that the first touch should be given 20% of the value of the sale.

In addition to the perils of overstating the accuracy of the figures we’re using, there’s a real risk here of us all being so busy measuring the height of the trees that we forget there’s even a forest there. The presentation of so many measurement points has diverted attention away from what really matters, or should I say, from what always mattered, and that is a simple understanding that marketing and advertising are necessary practices when trying to sell a product or a service.

Going back in time to the pre-digital era, simply knowing that your media will cost X and your creative will cost Y, so, therefore, we need to sell Z amount of product to make it pay, was pretty much the only way anyone measured advertising success.

So what is my point I hear you cry? Should we all just give up and carry on throwing our money blindly at the latest media opportunities without giving a care as to what works and what doesn’t?

Well.. no. Of course not. I’m not pro-cave dwelling, or about to suggest we all start smashing looms. True attribution might be impossible to achieve, but it’s still an ideal that is worth pursuing. For instance, one thing analytics tools can teach us is what isn’t working. And by not working I mean we can quickly gauge which digital or even offline channels are not driving traffic, and furthermore, traffic that sticks, and traffic that converts into leads or sales.

But what we have to accept is that for the four or five channels that are working, there is probably no outright winner, and no perfect combination; that each channel will influence the performance of the other in ways that are too complex and numerous for anybody to aggregate, average and analyse. And we have to appreciate that while many of the available tools and the metrics they provide are invaluable, they are not the goal itself; that in the absence of true science, sometimes anecdote, heuristic impulse and experience do still have immense commercial value.

 

Nick Beckingham
Head of online advertising

 

Social Share