There has been a lot of talk about Google’s Quality Score lately. In a quick write-up of the current discussion, I want to extract some interesting facts and figures about it. I argue that you shouldn’t ignore good old Quality Score.
A recent announcement by Google about changes in the way Quality Score will be reported has sparked a renewed discussion around this infamous metric.
While Quality-Score-haters gleefully reiterated that Quality Score doesn’t matter anyway, and won’t matter much more after the announced changes in reporting, others took the chance to bravely defend Quality Score’s honor.
The big G itself emphasizes that the changes won’t affect how Quality Score is actually calculated, but the way it’s displayed in the interface. Weighing in on the conversation, Google also put up 10 facts about Quality Score.
Quality Score is just a fancy word for CTR
Here are a couple of nuggets regarding Quality Score (QS) that I extracted from the current discussion that I find noteworthy and deserve to be repeated:
- Even though there are many factors that determine Quality Score such as landing page quality and geographical relevance, it’s mostly about Click-Through-Rate (CTR).
- There are three types of CTR performances that affect QS: performance of a keyword with a specific ad text, performance of an ad within the account and performance of a keyword within the account.
- CTR is normalized by ad positions. This means that boasting a CTR of 10% on position 1 could actually lower your Quality Score if your competitors normally score a higher CTR. On the other hand, don’t feel bad if your ad scores a 2% CTR on position 6. Your QS might get a boost, because ads on position 6 aren’t expected to score a very high CTR by Google.
- Not only is QS directly related to your actual costs per click, but also to your Cost-per-Order (CPO). Larry Kim analyzed Quality Scores of keywords that are attributed with about $ 100 Mio. in annual revenue. He found a correlation between QS and revenue, even when branded terms are excluded from the analysis (branded terms traditionally have both high QS and revenue).
- If a client asks, “why do I have to pay high CPCs even though there is very little competition on a keyword?”, the answer is: the QS stinks. If your keyword has a QS of 3 and your only competitor boasts a solid 7, you will have to pay close to your maximum bid.
- If your performance temporarily drops due to new keywords, restructuring, pausing keywords, weak demand or similar changes with a short-term impact, your QS should not be ruined. Frederick Vallaeys, a former Googler who helped build AdWords, recently shared a script for calculating an account-level QS. Even though Google doesn’t actually create a singular number that reflects “account QS”, Vallaeys suggests monitoring a weighted average QS per account. Interestingly, the code weighs the Quality Scores of keywords based on the impressions generated during the past 30 days. This suggests that adding new keywords with low QS won’t affect your overall QS much, because they haven’t yet generated enough impressions in the past 30 days to merit a penalty on account level. It’s therefore plausible that Google employs a similar “moving average”-technique when evaluating historical account Quality Score data.
So, Quality Score is complex. The six facts above are just some of the many aspects of QS to keep in mind. It’s not an arbitrary number that unfairly hinders the performance of your new keywords. It actually makes sense. QS indicates how your keywords, ads and account performs relative to your peer group with similar keywords and ad positions. Google thinks of it as a predictor of the likelihood that a user will click on your ad, based on a wealth of performance data about that keyword, your ads and the general awesomeness of you as an advertiser.
What’s it to search marketers, then?
Try filtering keywords that have low quality scores (your 2s and 3s) but a boatload of impressions and figure out why Google thinks those keywords are irrelevant. Remember, your keywords could have seemingly high CTR and still a poor QS. Also, check if your best converting keywords really have above average QS (this isn’t necessarily the case since conversion rate doesn’t factor into QS). If not, you might try getting lower CPCs out of these keywords by testing copy changes.
Monitoring and optimizing for QS will ultimately lower your CPC, and most probably your CPO. Moreover, having keywords with healthy Quality Scores will not only improve your current performance but also the performance of your future keywords in money-making ways.
In conclusion, don’t hate on Quality Score. Reconsider your pride and prejudices!