Does Adwords Ad Rotation Still Work?
Tried to run a campaign to test domain names for a new website. A new campaign was created with 1 adgroup containing 9 ad text ad variations. All 9 variations were the same except for the display url. Ad rotation was set to rotate indefinably.
The problem is the ad that got the most clicks all so received almost 3 time more impressions than ads that received lesser or no clicks.
This presents a problem since I don't know if the ad that got the most clicks was due to having 3 times more impressions or people preferring the domain name.
Is there a way to force a more even ad rotation and more evenly distribute impressions?
Re: Does Adwords Ad Rotation Still Work?
Hi, what exactly was the difference between the Display URLs? Since an Ad Group can only contain Ads pointed at the same domain, and the Destination domain must match the Display domain, I presume the differences were in subdomains or text in the URL?
Ad Rotation does not show your Ads evenly; in other words it does not guarantee that if your 9 Ads are in total shown 90 times in an hour, that each will be shown 10 times. Even Ad Rotation enters your Ads evenly into the auction. This is a subtle but important difference when you take into account that Quality Score is calculated for each auction as well. It is possible that some of your Ads have a better performance than others because of their Display text (hence my question above) - in fact it's quite likely, this is the whole point of your test, no? So, those with better performance may, when entered into the auction, have a higher Quality Score and therefore a better Ad Rank which of course, will almost certainly mean more clicks (because CTR rises with Ad Rank) and may mean more impressions (when the poorer performing Ads fail to appear on the first page).
If the above is what's happening, your test may actually have worked, perhaps just not in the way you expected. It probably is that the Ads receiving most impressions/clicks were the most popular and you have your answer. I'd look to confirm this by pausing the low performance Ads in that Group and running the test for a similar period. If the impression share for each Ad is much closer, then it suggests there is a similarity of QS for these Ads.