Imagine for a moment that you are buying ads in a magazine.
Full page, full colour ads. Nice.
The salesman at the magazine tells you that you’re getting a great deal. For only $10,000 dollars your ad is going to be in 150,000 magazines. You think that’s a great price. You book the ad, give them your money and run back to your office to wait for the phone to start ringing off the hook.
But the results are less than spectacular, and you start to question why. So you get back in touch with the salesman. “How many magazines does my ad appear in?”, you ask. “150,000,” is the happy reply “but we actually only sell 50,000 of them, the rest we just recycle and pulp.”
Would you be upset? You betcha.
Welcome to the Google AdWords display platform.
Before you panic and cancel all your display campaigns – don’t worry. It’s not actually as bad as I made it sound.
But it does need some explanation.
Google initially counts all impressions of your ad. But many of these are not visible in the viewport when a user opens up a webpage. Think about when you open up a newspaper website, for example. These pages can be long – very long. Long enough to give your mouse finger cramp. An ad at the bottom of that page is not going to be visible when the page is opened. In fact, it may be a very small number of people who ever get to see the ad in that position. But it’s there. The ad was in the code. and if you had scrolled down to the bottom of the page it would have been there in all it’s glory.
Advertisers understandably become more and more upset at paying for ads that were simply never visible (please remember that this only applies to campaigns using the CPM – or pay per thousand ad impressions model). So Google now use active viewable impressions as a measure against which you can – and should – choose to pay.
The screenshot below is from a recent campaign we ran for a client. Notice how the number of impressions is comparatively high at 159,671. And these generated 209 clicks. A CTR of 0.13%.
However, the active viewable impressions column (which you’ll need to activate to see under the columns tab) shows that only 59,040 of these impressions were ever actually seen by users. About a third.
Since this campaign is using the PPC model where we only pay for clicks and not impressions, this is not really a problem. Except there is one thing that is misleading… the CTR. The 0.13% CTR is for all impressions. But a user who never sees an ad cannot click on it. So the true CTR is actually 0.35% – significantly healthier.
If we are judging the success of our campaigns by (in part at least) their relative CTR’s, then we could possibly reject this campaign as an underperformer whereas, it may have performed fairly well.
The takeaway from this is that we must be very careful when we look at the headline numbers presented in the default reports in Google AdWords – the story they tell is often distorted and requires a little reworking before we can base our optimisations on it.
