Top vs Other Analysis in AdWords

adwords top v other report

Top vs Other analysis has been a feature of AdWords for a couple of years now. However, its still under utlilised by many search professionals, even though it’s an extremely useful tool for digging deep into your average position and performance data.

Why is Average Position Misleading?

Average Position is a statistic that almost all PPC specialists love to use. It shows where your ads are displaying compared to your competitors’ ads. This statistic can be shown at any level: Campaign, Ad Group, Ads, or Keyword.

The first problem with average position is that it only shows the average position of all your impressions. This means it does not include when someone searches one of your keywords, and your ad does not show up. In other words it does not include your lost impression share.

The next problem with average position is due to something called “Aggregation Bias“. The role aggregation bias plays in PPC involves the %chance that your ad gets clicked based on its position.

For example, if an ad is in position 1 on the top of the page it is likely to have a very high CTR (click through rate). If an ad is on top of a page in position two or three it is still likely to be clicked. However, if the ad appears in position four, it is almost the same likelihood to be clicked as if the ad was positioned at position ten.

Therefore, average position can be misleading because it does not show how the average came about. If an ad has an average position of 4 that came about with 100 impressions, it is likely to have a lower CTR than an ad that has the same average position of 4, but got that average position by having 50 impressions at position 1, and 50 impressions at position 8. This is because the impressions at position one are the only impressions likely to have a CTR over one percent (aggregation bias).

This is why using Top vs Other is a useful analysis report. You can find it in the segments tab in AdWords.

Top V Other AdWords

Using Top v. Other To Help Limit Errors in Average Position

In order to use Top vs. Other, go to the tab that you want to see the statistics for, whether it be at Campaign, Ad Group, Keyword or Ad level. Find the “segments” tab and “top vs other” will be listed second from the bottom, right above “annotations.”

This will show you how many of your impressions were on the top of the page on Google, how many were on Google but not on the top of the page (other), how many were on the top of the page on a search partners site, and how many were on a search partners site but not on the top of the page.

You can use these statistics to compare how certain ads did while being on top of the page, rather than just comparing the average position (where aggregation bias is put into play). You can also use these statistics by making an excel spreadsheet to show you what percent of the time your ad appears at the top, what percent it appears somewhere else, and what percent of the time it does not appear at all (lost impression share).

So, is using Top vs Other a perfect example of getting rid of aggregation bias? Absolutely not. Top vs Other does have weaknesses, as it does not show you what position your ad was in. For example, in certain cases, your ad could have been in position 1, but still not show on top of the page because of the lack of interest in the auction for that particular Keyword. In order to completely get rid of aggregation bias, Google would have to supply us with what position the ad was in each time a term was searched for, and where the ad was at on the page.

This would be nearly impossible for Google to show us, and it would be nearly impossible for us to have time to go through all of those statistics. Top vs. Other, however, is very easy for us to digest and gives us a good top level idea of where our ad is located when it gets its impressions and clicks.

Recap

  • Aggregation Bias affects the accuracy of your average position data
  • Use Top v Other segmentation in AdWords to determine how your keywords and ads really perform in different positions