A Little AdWords Quality Score Sluething
July 17, 2007
Bill Slawski over at SEO by the Sea recently unearthed 3 patents filed by Google that map out the methods AdWords uses to determine the quality of an advertisement. As expected, the patents are mind-numbingly dry, but still we found them illuminating since the PPC Hero Team is constantly trying to breakdown the AdWords quality score to the most minute detail. As we are always looking to be of service to you, I read through these patents so that you wouldn’t have to. Here are the summaries and major points (from what I could decipher):
The first patent discusses predicting ad quality: An advertisement’s relevancy to a user’s search is determined by its performance and displayed accordingly. However, I find this statement of interest, “applying an algorithm to the retrieved past quality scores to provide a value that predicts the future quality of the advertisement.” This means that AdWords determines the likelihood that an advertisement is good by blending your past, present and predicted performance.
We all know that AdWords filters advertisements by the keywords in any given query. And we know that if your advertisement has performed well in the past, then you’ll be rewarded with a lower cost-per-click (CPC) and a higher ad rank. However, this patent states that AdWords also takes into account your advertisement’s future performance as well.
The second patent discusses ad filtering, ranking and promotion: How does CTR*CPC+K.sub.1*QP.sub.1*CTR-K.sub.2*QP.sub.2*CTR affect your AdWords performance? Good question! But I can’t answer it. This equation details how an advertisement is ranked without factoring in click-through rate.
This patent describes, “a method, comprising: obtaining a first parameter (QP.sub.1) associated with a quality of an advertisement among a plurality of advertisements with at least one other parameter to filter, rank or promote the advertisement among the plurality of advertisements.” Basically this means that AdWords uses another parameter other than CTR to determine an advertisements’ rank and promotion. But the first parameter (QP.sub.1) is never defined so it is difficult to dissect this ranking method effectively.
However, minus the definition of the illusive QP.sub.1, I can tell you that rank is determined by your bid, click-through rate and overall quality score. This information is nothing new, but I wonder what the QP.sub. 1 could be? Could this be the attributes of the quality score minus the CTR? Perhaps an actual human review that ranks the relevancy of each advertisement?
The third patent discusses how user behavior may influence your quality score: There are a number of actions factored into a user’s interaction with paid listings, and here they are:
- The time an advertisement selection is displayed
- How many ads are selected before & after a given advertisement
- How many search results are selected before & after a given advertisement
- How many other types of results are selected before & after a given advertisement (ex: map)
- How many documents are viewed before & after a given advertisement
- How many different search queries were performed by the user before & after the given advertisement was selected
- How many times a user selects the same given advertisement
- Was the given advertisement the last one selected for the given query
- Was the given advertisement the last one selected during the user’s session
I don’t believe a score is given to each of these actions, rather an aggregate score is given after all or any of these actions have taken place.
These are certainly the “murkier” attributes of the quality score. When managing your PPC campaigns it is difficult to pander to every attribute to the quality score (especially if there are going to be so many!). How much weight does each attribute of the quality score carry? You’ll be able to throw in your two cents on this topic! Watch the blog tomorrow because we’ll be taking a survey to see what you think.
Update: We want to know your opinion! Complete our quick Quality Score Survey.
Browse By Category
How to Drive Calls that Actually Convert
Our experts are here to explain how gaining call tracking data can help you close the gap in your customer journey and tailor their experience so they actually convert.
New Facebook Analytics Updates from F8
Learn about journeys and automated insights, new tools announced at Facebook's F8 conference for the Facebook Analytics tool.
Spotify Advertising: The Why and How
This blog post gives you the rundown on Spotify, including why and how you might advertise on Spotify's new self-serve ad platform, Ad Studio.
A bi-weekly newsletter packed full of resources and strategies that will help make you a better PPC expert.
Hanapin Marketing | The PPC Agency of Experts Behind PPC Hero
5 Facebook Ad Settings to Audit In Your Account
5 Easy settings to audit in your Facebook Ads account to improve performance and understand website traffic.
Analyzing the AdWords Distance Report
The distance report provides a unique view of your geographic data. The report helps users understand how distance from a location impacts search ad performance.
How to Master Attribution for Easy Budget Planning
In this webinar, Hanapin’s Cassie Oumedian and Google’s Julia Meter will provide tips on using AdWords attribution to directionally decide where you should allocate your budget, and how you can use the changing industry to your advantage.