Google recently announced that it was introducing an advanced attribution solution for marketers: A multi-touch attribution lens that fractionalizes conversion credit across touchpoints as they’re observed to drive consumers through the funnel. In layman’s terms, the media giant is offering to track and credit all of the “clicks” that push consumers to the final purchase of a product, rather than assign full credit to the “click” that immediately precedes the sale, or what is called the “last touch.” Until now, the “last touch” has been Google’s standard operating procedure.
This all sounds good and generous. After all, the service is free, and it seemingly fosters transparency, but the devil is in the detail. Google’s attribution solution won’t import cost data from, and will likely only be able to collect partial (click-only) data from non-Google platforms. In other words, Google will continue to grade its own homework. Not measuring full exposure will be especially bad for most other video and audio platforms, not to mention other “upper funnel” publishers. The data will be incomplete and inaccurate.
Further, we have no way of knowing the assumptions (bias) behind the model’s algorithm. I’m sure that Google won’t compromise huge revenues on an algorithm that potentially decreases the value of search — a long-time benefactor of last-touch measurement at the bottom of the funnel. Sure, YouTube might benefit somewhat, given its more upper-funnel nature, but it also makes it more accountable to business results.
It’s hard not to be cynical: Is Google signaling that the way it has been measuring the billions of dollars through its platforms might have been less than ideal? Is this in principle any different from the Facebook measurement debacle a few months ago? Sure, there is a big difference: For Facebook, the incorrect data was around a transacted media currency. But both examples represent an admission that media value might have been or is being inaccurately represented.
There is a decided benefit of the advanced attribution model vs. the last-touch industry standard way of measuring. Ultimately, it enables marketers to better appreciate the “incrementality" of their advertising and provide more levers for growing their business through understanding channel interplay. I appreciate the move that Google has made in trying to offer a more logical and sophisticated way of measuring ad effectiveness. It signals to the marketing industry that the inertia around employing a simple last-touch perspective on measurement is increasingly insupportable. After all, we — the agencies — are handling clients’ monies for the purpose of investing in growth. Google might even be feeling conspicuous in not sooner having a more democratic offering around advanced measurement.
However, marketers should beware a false sense of security just because this has a Google label. Yes, it’s convenient and, hey, if you are already using Doubleclick ad serving, why not? But one can’t help feeling that Google is creating an illusion of transparency (neutrality) in measurement and another hook into controlling the digital ecosystem. Instead of democratizing more advanced measurement and implementing self-imposed transparency, the reality might be the exact opposite. Moreover, I’d be very surprised if Google pushes this “advanced” attribution model hard and risk hurting its golden goose — search.
As marketers, on either side of the agency-brand fence, we must be careful to not fall victim to the contrivances of Google and Facebook, and their insistence that only they measure themselves, whatever the rationale. As they say: With knowledge comes power.