Does Hal Varian Disprove “Tire Kickers” In AdWords Position 1?
I believe that Hal Varian, the Chief Economist at Google, is a very smart man. (If you haven’t seen his introduction to Google’s ad auction on You Tube, I highly recommend it.) He recently released a study showing that conversion rates in AdWords don’t vary much by position. This seemed a bit counter-intuitive to me, and didn’t exactly dovetail with my personal experience, but he addressed the issue with a well reasoned sense of its complexity, and I can buy most of the thinking.
But there’s one element of his reasoning that’s really got me. If we accept what he says as true, this also applies to position 1, where most AdWords professionals believe a lot of unprofitable clicks exist because of a disproportionate amount of errant clicks from searchers just “kicking the tires” and looking without an interest to buy. I really have seen in my personal experience the CPA soar at position 1 in AdWords, and I don’t think it’s just because of the math of the ad auction. I really think it’s because conversion rates have dropped a bit in that position, although going back to check the data is a bit difficult (in part because accounts I regularly review these days are rarely averaging position 1).
Has Hal Varian proven that the tire kickers premise is false, with hard data? Well, for one thing he says that conversion rates don’t vary much. I’d venture a guess that where they do vary a little, it’s position 1. Also, even if conversion rates are the same in position 1, the algorithm of the ad auction may boost the price paid for position 1 to the point that the CPA is unprofitable even if the conversion rate stays the same. And when your CPA is up, that’s the metric that’s most likely going to stick in your mind, even if conversion rates remain constant.
Push Bids Up And Average Position Goes Down, Now How Did That Happen?
Another observation that Hal Varian had in his notes was this counter-intuitive AdWords nugget: If you raise your bids, your average position can go down. Now, how can that happen? If the ad auction is essentially based on a formula where “quality score x bid = position” and your bid is raised, simple math says average position will necessarily go up, which is what you do normally see in real life applications in AdWords.
However, impression share is also tied closely to your bid in AdWords. Raise your bid and you increase your impression share. If your bid increase qualifies you for additional ad auctions you wouldn’t have been in before, and in those additional auctions you place lower on the age (which Hal explains is indeed possible), then yes, you’re lowering your ad position. You’re not, however, lowering it for similar auctions as you were in before. Instead you are diluting the average position data with additional impressions. Apples to apples, you’re in the same shape or better in terms of ad position.
by Rob Sieracki
Director of Paid Search
Sources:
http://adwords.blogspot.com/2009/08/conversion-rates-dont-vary-much-with-ad.html
http://adwords.google.com/support/aw/bin/answer.py?hl=en&answer=146080
