So far, I’ve assumed the quality of an ad to be only its CTR. That was good enough to demonstrate the basics of how Adwords works. The reality is a little more complex.
As mentioned, Google invented what it calls the Quality Score to measure the quality of an ad. In fact, saying the quality of an ad is technically wrong. While we do talk about the quality of an ad, it is really the quality of the combination of the ad and keyword. The QS is calculated for each keyword-ad combination. The same keyword will therefore have a different QS for each ad in that group although that is not shown to you at that level. What happens when you have more than one ad? How is QS calculated if there are many ads and QS is calculated for each ad and keyword? Google simply averages out all the quality scores for all active ads. Of course, that means the QS for a keyword cannot be better than your best ad but neither can it be worse than your worst ad. It is also a reason you should test only two ads at a time. Having four or five ads with two bad ones will drag down your overall QS on a keyword. Having too many ads active at a time also does not provide a good testing environment.
While Google has never said exactly how the QS is calculated, they have made some information public, keep most of it secret but have also dropped some hints.
Quality Score is made up of three major components: the keyword’s CTR, the relevancy of the keyword to the ad and landing page and finally, other factors. What we don’t know for sure is the exact weight of each component. Again, Google has dropped some hints and has said that the CTR component is by far the largest.
We all know by now what the CTR is. We know it has the biggest weight of the three components and we figure it to be at least 60% and almost surely no more than 66%, two thirds. The exact percentage doesn’t really matter; the QS calculation can change and the weight of each component can be adjusted as new components are added or they decide to change the weight of one component. What matters is that by having the most weight, it makes sense to pay more attention to CTR than other components.
Not to say that the other components should be ignored. Relevancy of the keywords to the landing page is extremely important too. It represents about 25% of the QS. If you’re not relevant, there is no way you can get a QS of ten or even eight for that matter. By not being relevant, you already put yourself at a huge disadvantage.
Relevancy should be the easiest to achieve a high score on. All an advertiser has to do is choose keywords that describe the product or service they are selling.
Unfortunately, time and time again I see advertisers messing it up and get a poorer QS as a result. Every day I see ads advertising a product or service which have nothing to do with the search term I typed. Other times, if I happen to click on an ad, out of curiosity and therefore costing that advertiser, I land on a page which, again, either has nothing to do with my terms or little of what I expected based on what the ad said and the keywords I used. I see this with a large percentage of my new clients as well.
Your keyword is either relevant or it is not. This of course as judged by the Adwords software. Therefore, since keyword relevancy is 25% of the QS, if your keyword is judged relevant, you have one quarter of your QS taken care of. So if your page is selling dog food, use “dog food” as a keyword and not cat food. If you are not relevant, you are wasting a quarter of your QS and likely 25% more expensive. And as it is not relevant, people don’t click on the ad taking your QS further down and costs up. Although I know better, I assume we all pass the relevancy test with flying colors thereby leaving only CTR. That is why I say QS is in effect your click rate and if you increase your CTR, you will increase your QS.
One must remember however that, at least partially, the QS is calculated by a computer. Computers don’t think like us and there’s always the possibility of it encountering something out of the ordinary. This happened to me recently with the lowest of possible quality scores on a new campaign. Being an expert, I scratched my head trying to figure out why and called Google. They admitted the software did not properly figure out the relevancy of the keyword and it was soon fixed. However, this kind of episode happens very rarely. It was in fact the first time it happened to me after having managed dozens of campaigns involving thousands of ads and tens of thousands of keywords. The point is, don’t go off the deep end and thinking that Google doesn’t know what it’s doing. It does know for the vast majority of cases. Your job as an advertiser is to figure out what you are doing wrong. Check your keyword-landing page relevancy. If that looks good to you and Adwords agrees, concentrate on increasing your click rate. The third and final major QS component is the landing page itself. This is not having the keywords you used on your page, that part is handled by the relevancy component. This component is other on-page factors and it represents only about 10% of the QS.
The only thing Google has made public about it so far is how fast your page loads. This may be a large portion, up to 50% or even all, of the landing page component, hard to say. If so, your page’s loading time would represent no more than 5% of the total QS. In other words, if your page loads more slowly than the average, your QS will not suffer tremendously. It may mean a QS of nine instead of ten.
As for the rest of the page component, many assume you need a policy page of some kind and that could indeed be true. It’s not a bad idea to have such a page whether the QS takes those into consideration or not. Others say you need to think in SEO terms and have lots of unique content and backlinks. Of this, I am not convinced. PPC is not SEO and your QS being affected by backlinks is ridiculous and illogical. Again, think of QS as your click rate.