The idea that Google’s agenda sometimes diverges from ours as advertisers is not new, or surprising. We see it borne out often in the way the platform evolves.
But recent admissions from a Google executive have shone a light on the extent of that divergence, and on some ugly practices to make sure of whose interests win out.
Those admissions come from Jerry Dischler, vice president for Google’s advertising products, testifying on Monday as part of a federal antitrust trial against Google.
The revelations show – to put it bluntly but accurately – that Google has been in the practice of tweaking their auction algorithms specifically in order to inflate click costs, doing so secretly, and in ways that contradict (at the very least…) the spirit of the law of ad rank.
That news takes some digestion. (At least it does for me… Never a fanboy, but a former Googler whose goodwill towards the company I owe so much is now being tested to the limit.)
According to this writeup from Business Standard
"In one May 2019 email, Dischler and his team discussed how they were “shaking the cushions” to find potential changes to the ad auctions that would ensure Google met revenue targets."
Quizzed on this in court on Monday, Dischler said his goal was “to get creative so we could meet our quota.”
He highlighted one particular change to auction dynamics, called ‘RGSP’. (Reverse Generalized Second-Price Auction? I speculate…)
Under this change, the runner-up is allocated the top spot, while the actual winner takes second place (!)
Major advertisers like Amazon.com win auctions very frequently…
With RGSP, “we flip them,” Dischler said. “Otherwise, Amazon always shows up on top”.
Dischler apparently didn’t know whether that particular practice led advertisers to increasing bids… but it had the desired effect of increasing Google’s revenue.
The ongoing trial isn’t about Google’s treatment of advertisers (though the market dominance under scrutiny of course sets the stage for this kind of exploitation).
But these revelations are about exactly that… and they may well be legally suspect, as well as being clearly unscrupulous.
What does this mean for us?
In the bigger picture – this may be taking us one or two steps closer to a tipping point…
Those of us deep into PPC are acutely aware of how Google has been steering the platform and ecosystem of Google Ads away from advertiser control and towards an automated, aggregated approach that emphases expansion.
Away from the idea that we should ‘only pay for relevant clicks’, to the message ‘pay for as many relevant clicks as possible’. We see it in the tools Google provides (and removes), the insights it offers (and hides) and the ‘best practice’ advice it gives. I first wrote about it in early 2019, and it was far from novel then.
I have long hoped that this direction of travel would start to reverse at some point. But that reversal will only come when Google deems it to be in its own interests.
So either Google will choose to change tack after overstepping the mark and ceasing to appeal sufficiently to advertisers. Or, perhaps, when they overstep the mark legally and are forced into that rethink…
But for the short term, I suggest we take this as a cue to double down on the metrics that matter to us.
We are hearing loud and clear that those diverge from the metrics that matter to Google… and this includes those ‘guiding rating’ metrics like Optimisation score, Ad Strength (and to a much lesser extent, Quality Score). These are not indicators that we are on the right track for our own purposes.
Work backwards from your bottom line. Study revenue (long-term as well as immediate) and cost as deeply as you can… But take every analytical step beyond those metrics with increasing care.
There are plenty of secondary indicators that can be useful – but in all cases (including CTR, CPC, and even Conversion Rate) they are not to be taken as definitive positives without caution.
Focus on what matters to you; not on what Google tells you matters.
And let’s hope they pull themselves away from this Wall-Street driven chicanery fast, because frankly, it’s not a good look.
