CTR and Quality Score Decoded

The more time I spend in the forums, the more confusion I see about Google AdWords’ Ad Rank (CPC bid × Quality Score). I’m going to try to break this down, not by what Google says it is, but what it actually is.

According to Google:

A keyword-targeted ad is ranked on a search result page based on the matched keyword’s cost-per-click (CPC) bid* and Quality Score.

Ad Rank = CPC bid × Quality Score

The Quality Score for Ad Rank on the search network is determined by:

• The historical clickthrough rate (CTR) of the ad and of the matched keyword on Google; CTR on the Google Network is not considered
• The relevance of the keyword and ad to the search query
• Your account history, which is measured by the CTR of all the ads and keywords in your account
• Other relevance factors

http://adwords.google.com/support/bin/answer.py?answer=6111

The more you think about this, the more your head hurts. In reality, the formula is quite simple: Ad Rank = How much revenue your ad generates for Google. The more revenue your ad generates, the higher your Ad Rank.

So wait, Is Google lying?

Not exactly. What Google states is 100% true. They’re just telling you the factors that determine the revenue potential of your ad and not the practical application.

Here’s an example of what I mean.

Let’s say:

• Pos#2 ad is bidding $1.00 with a 2% CTR
• Pos#3 ad is bidding $0.75 with a 4% CTR

Since, per Mil (1,000 impressions):

• Pos#2 ad generates $20 per Mil ((1,000 x 2%) x $1.00)
• Pos#3 ad generates $30 per Mil ((1,000 x 4%) x $0.75)

In other words, the ad in position #3 generates 50% more RPM (Revenue per Mil) than the ad in position #2.

Let’s state another premise we know is true: Some ads are viewed more than others (typically the higher the position, the more it is viewed). Therefore, by moving the lower positioned ad to a higher position, Google can increase the CTR …and its own RPM in the process.

So… By swapping ad#2 & ad#3, Google can boost the CTR of the #3 ad to 5% or 6% turning it into a $38-$45 RPM ad and still make $17 per Mil off the ad that was previously in its place. In other words, by moving up the higher RPM ad, Google has turned the average RPM of those to ads from $25 to $30 …increasing their RPM by 8%.

I know I’ve over simplified this but, for the most part, it really is this simple.

What I did not explain is the predictive modeling I believe Google can do (“If ad#2 has a $20 RPM & ad#3 has a $19 RPM, what is the predicted RPM if the two ads are swapped?”) or the ads that lead to a poor user experience, like arbitrage sites. Obviously Google has to balance revenue with great user experience (which keeps the user clicking on their ads). They can’t allow just any ad to be displayed, if that ad leads to and adverse experience. My theories around user experience would better be explained in another post …if i can ever get around to posting more than once a month.

Disclaimer: I have no documentation to back this up, but my experience (and common sense) tell me this is true.