Google Ads announced in February that they would be sunsetting the Average Position metric, replacing it with Impression Share metrics for “Top” and “Absolute Top”. Ready or not, the time has come to say goodbye and we have everything you need to know about how to navigate the change.
For many people in the PPC world, this is earth shattering. Average Position has been our friendly helper for so long, letting us know where our ads were appearing in a way that we could point to on the rectangle of our screens. We could connect this performance metric to our experience as users of Google search. It’s been consistent and tangible in a world that’s constantly changing.
When I was first learning PPC, Average Position was a warm blanket of security that I was doing the right thing. As long as the average position was low (i.e. high on the page) for key terms, I felt that I at least wasn’t making any huge errors. Google Ads had such an overwhelming thicket of metrics and data that Average Position’s folksy straightforwardness gave me a common-sense gut check that everything was OK.
The problem, as I later learned, was that this metric could lie to us.
Average Position is a Liar
Ok, ok, Average Position doesn’t so much lie, as it misleads. The key distinction is between precision vs accuracy. Average Position is precise in that it gives us a decimal place – we get reporting down to 10 notches between each ad position. But too many people take this decimal place to mean that the ads are showing up within a narrow band.
For example: an Average Position of 2.1 may sound like the ad showed mostly in position #2 and a few times in position #3. Something like this:
However, the distribution of impressions probably looks much closer to this:
Average Position is deceptively simple and can lure PPC managers into making decisions – even writing sophisticated bidding scripts – that manage to this metric.
However, ignoring the full picture can kill an account’s performance.
Impression Share Was Right All Along
Let’s say a PPC manager is managing to an Average Position of 1.0 for a key campaign. And let’s also say that this manager isn’t paying close attention to impression share or the “Limited by Budget” status warning.
They’ve pushed their bids to where they’re paying big dollars per click, and have achieved their goal – a 1.0 Avg Position. Nice! Right?
But wait! If their daily budget is low, they’re going to get only a few clicks before slamming into their daily budget cap, (we’ve seen accounts that hit their budget cap before 8 am). For the rest of the day, their ads are not showing at all – but their reporting for that day will still display the glorious 1.0 position that they achieved.
You know what metric could have alerted them that this was happening? Impression share. Specifically, IS Lost due to Budget. Seeing a good Average Position but a low Impression Share has always been a red flag that bids and budgets are misaligned in an account. And fortunately, as automated bidding gets better and better, this problem will become rarer.
Manage To Your Actual KPIs
What is sometimes lost in discussions about these competitive metrics is that these are fundamentally not our end goals. The point of PPC is to drive more revenue – either directly via e-commerce sales or by generating high-quality leads, and these metrics matter far more than market coverage or ad position.
Rather than fussing over small changes in ad position or impression share, look at ROAS (conv value / cost) or cost per conversion. If they are good for a given keyword, only look at the competitive metrics to make sure that there’s more headroom to increase the ad spend.
Except for extreme niche market businesses, no company has the budget to rank #1 for 100% of the relevant search volume all the time, especially not if they want the campaigns to add to their bottom line.
Let’s Embrace Better Metrics
Impression share has always been the more helpful metric, but it never got the same attention as Average Position. I get it: it’s harder to communicate to a client or your boss “how the ads are doing” with a percentage, especially when they might be pointing directly on their screen at a competitor’s ad above your own. It’s hard to talk about bid adjustments, auction-time factors, machine-learning, budget constraints, and the dozens of other things that go into an ad serving in a real user SERP to all the stakeholders.
But I think switching our language to talking about impression share – especially at the top (and absolute “tippy” top) – will be a good thing. By changing the metrics that we track, and managing to those metrics, we’ll all be better for it.
Clay McDonald, Director of Operations