Is quality score (as based on the CTR component) determined on a keyword by keyword basis. Or is it based on an entire adgroup. In other words -- if you have one word which is just about over the delivery threshold which has loads of impression and 10 words which you have a very high CTR but less impressions -- are you affecting the quality score of the 10 words or not? Whats the best way to separate underperforming words ... different adgroups? different campaigns? different account?
I moved a word which was averaging position 3-4 in one adgroup no matter what bid I placed to another established adgroup (with similar creative). Keyword went straight in at 1 and at a reasonable Cost Per Click. Exactly the same keyword. I think the history / performance of a keyword group & ad creative goes a long way to determining position (and in return CPC). I.e if an adgroup suddenly turns bad get your best performing words out of there quick.
No reason to do that...and actually, I would advise against it. If you delete your keywords from 1 ad group and start a new a new one you're essentially "resetting" your quality score and wiping out the history associated with that word.
When I say get the word out - I mean add the word into another adgroup. In my case the word which I couldn't get higher than position 3-4 in one ad group, I paused that ad group. So in effect I could resume it in the future if my test didn't work out. I agree with you, I have always believed that the History of the Keyword is (one of) the main determinants of the quality score. However my recent experience shows that I am now getting more effective performance from the specific keyword being in a different ad group, than I was from having in its original ad group. The keyword was in its original adgroup for 6 months and has been in the "newer" adgroup for 2 weeks. The problem at the moment is that there seems to be so many variables that could possibly wipe out history / quality score.
On this basis it makes sense to split your keywords into as many adgroups as is sensible (based on the type of words). If I remove keywords from one group to another does this mean they will be starting from scratch in terms of quality score history. Does google not recall the performance of the word from a previous adgroup?
I recently got told that keyword history does move with keyword throughout accounts. However having any Quantitive evidence is hard to find. However the reason why I did the above, is because the History of the actual Ad - i.e the creative seems to have more effect than the keyword history - When determining the total quality score The Adgroup I moved the word into has had a really stable, high CTR-generating suite of Ads in it for a long while. And as I said the Keyword I moved could easily run in this group without the need to change the creative. Therefore the Keyword is now running with ads that have longer histories and higher CTR's. The effect was instantaneous - without going into great detail, I could have been bidding £2.50 a click previously and not got higher than position 3-4. Now I am top and paying 25p a click. The difference in CTR's between the keyword in the old and new groups was marginal, which makes me suspect that the ad history etc has a very heavily weighted influence on proceedings
That's a huge difference both in CPC and position....well done! But I think you're right that as G won't give us all the answers we have to play around and see what works. I did move the keywords into another group -- to be honest I should have had them in a separate group ages ago -- it was just laziness. What's unfortunate is that you lose the historical data for those keywords from the adgroup where they were in. Thanks for all your help. Dan
I would re-iterate though, that it is my belief this will only work if you put the keyword into an adgroup that has ad creatives that are very strong (and have along history). I suspect putting the keyword into a completely new adgroup would not generate the improvements. I think it is the effect of the adgroups creative and history that "rubs" off on the keyword.
The flipside of starting a new group is that by creating a more specific creative I might increase my CTR higher than it was on my old campaign because it is that much more specific to the users request. As an afterthought I should probably have paused rather than deleted the words in the old campaign ... doh.... to keep the data but I will monitor the new campaign performance and see how I get on. But I suspect you are right about google having a selective memory about your keyword performance when you start a new campaign.