Numerous discoveries have been made because of accidents, like penicillin. A while back I made an Excel error (messed up a CONCATENATE function because of relative cell references) that uploaded the following keyword/ad group pairs:
What I MEANT to upload was the ad group name as the keyword. Oops! However, it was amazing to see the difference in quality scores:
The campaign had 60 keywords and I loaded each one as broad, phrase, and exact. As you can see I uploaded my keywords as “keyword + campaign name” and the keywords matched the ad group name exactly. What was interesting was how this affected the QS. For the initial upload the average QS was 9.27 and for the corrected upload the average QS was 6.81. So 2.5 points of QS were lost when I removed the campaign name from the keyword. No other changes were made.
How have you seen account structure influence QS?