Sorry about the odd title but it's the only way I could make the message fit. I have been averaging 2.75-3.5% CTR for the last 3 years. No matter what I do, my most important keywords are always given a low quality score and I'm no noob. I know what makes a good QS, something is funky here. Check this out: Ad Group 1 Keywords: +Appliance +Repair ... this is my main bread and butter. It lands on my index page which has been optimized for this exact broad match. Ad Group 2 Keywords: +Dryer +Repair ... a side ad that is currently pointing to the same page but will soon point to it's own landing page. #2 has a higher QS than #1 and there's no way on earth that this can be justified when you look at the numbers from SEO Quake for this landing page: Keyword Found In Repeats Density repair...... T,K,D...... 15........ 3.09 service .....K,D....... 14........ 2.89 appliance .T,K,D...... 11........ 2.27 appliances ..............6......... 1.24 dryer ........T........... 2......... 0.41 Looking at the above anyone would feel that +Appliance +Repair would rate far more highly than +Dryer +Repair since they both share +repair but +Appliance has over 5x the repeats, triple the density and repair appears in keywords, description and title vs only the title for +Dryer... Yet +Appliance +Repair has a QS of, get this, 4... in red. +Dryer +Repair somehow gets a 5. I had a QS of 6 when my site was a disaster that SEO quake had red X's everywhere for. Now now Quake is only red for Microformats and Feed. I've changed the page to have real, readable content rather than just a checklist and yet my QS in Ad Words went down instead of up. I really don't undersatand how the QS for a keyword match that the page is not optimized for can be higher than for a keyword match it is optimized for. It's as if the more SEO I do, the worse things get. Good advice from those who have dealt with irrational keyword scores would be appreciated. Thanks.
I found it... Dave Davis' QS improvement guide pointed it out. I had no idea that the longer a keyword was in place the higher the qs went. Edit it one time and, even if you edit it back, you lose all that history. So bouncing between ad groups, ads and keywords resets your history for performance and relevancy every time you do it. Moral of the story... plan your ad groups, your ads and your keywords and then don't make changes all over the place unless you think them through carefully and once you change them, leave them alone and let them build history. Still doesn't explain the dryer doing better though. Sigh...
I do the same with my AdCenter campaigns (I know it's different from AdWords), but once I create ad groups I don't change the keywords for nothing. Every change makes it run differently.