Microsoft adCenter’s Illogical Ad Position Calculations
Ryan Moothart Mar 19 2012
Not too long ago, Bing introduced text ads appearing at the bottom of the search results page in addition to the top and along the right side. However, what you may not know is that these ads can be duplicates with different ad positions. Yes – the same exact ads on a single results page in different positions. Or, as we in the business call it: double serving ads.
Allow me to help you visualize:
Notice that in my simple search for “clothes,” there are four ads at the top, eight on the right, and two at the bottom. The same ads for Oakland Apparel and Clothing at Freddy’s appear at the top and the bottom of the page. In this circumstance, the top ad for Oakland Apparel would be in ad position 1 while the same ad on the bottom would be in ad position 5.
Upon further research, we have been able to determine how this effects average ad position calculations – whichever ad receives the click gets the credit. However, whenever these ads are double-served, it only counts as one impression and not two.
Duplicate Ad Position
But wait, that’s not all. In this instance, the first ad on the bottom AND the first ad on the right-hand side are calculated as being in position 5. The ads at the bottom are considered a continuation of the ads on the top, just as the ads on the right are. Hence, there are two position 5 ads on the same search results page.
Quality Score (QS) in adCenter is based upon keywords, landing pages and user experience. AdCenter explicitly says QS doesn’t impact ad position. But something just doesn’t feel right. Technically speaking, if you were to increase your keyword bids based on an average position, wouldn’t this in some way (while not direct) be influential? We did receive this bit of info on this particular question:
The average position would be calculated for the first ad unless the second ad is receiving clicks and then it would blend the positions. If the second ad doesn’t receive clicks, then your average position would be the average position of the first ad. Therefore, if you are trying to optimize and increase QS by raising bids, your average position would be accurate.
We’ve been finding more than one interesting anomalies in PPC lately. Have you noticed this or any similar issues before? Let us know in the comments.
Ryan is a PPC strategist at Portent, Inc. He started at Portent as an intern soon after graduating from Willamette University with a BA in Rhetoric & Media Studies. Read More