The more PPC I manage, the more I’ve come to believe that Quality Score is mostly a function of click thru rate (CTR) and account history.
Quality Score is mostly a click probability calculation.
If a keyword-ad combination has a high CTR, it usually receives a perfect quality score. However, it takes time to earn that perfect score because Google is performing probability calculations using an ever growing set of data. If your ad has 1,000 impressions and earns a 6% CTR (for example), that’s a good sign that your ad is a strong performer…but 1,000 impressions isn’t a large enough sample (statistically speaking) to determine if your ad is worthy of a ’10’. SO, you get a Quality Score of 7 or 8 and you wait.
After 10,000 impressions, holding a 6% CTR is a stronger signal that your ad “works”. After 50,000 impressions, the data set is probably large enough for Google to give you the exact quality score you deserve.
However, what about keywords with low search volume? What about new ads? What about a new account? There is no way to calculate quality score with any accuracy in these situations, so Google calculates the click probability of your ad-keyword combination and uses that information to determine your Quality Score.
Most likely, Google weights account history heavily when calculating click probability on all your ads. This is why it’s best to start with a brand keyword, as it will enjoy a very high CTR and help all your future ad-keyword combinations earn a higher click probability calculation.
Of course, things like tightly focused ad groups, ads and keywords and landing pages that all “match”, the use of negative keywords, etc. help Quality Score too. This is all fairly basic information that any PPC manager should have, but mostly these things are important because they boost CTR.
So, assuming that you know what you’re doing, the strongest factor for raising or lowering quality score outside of CTR would seem to be account history. This matches my experience.