Google PPC Quality Score (adventures in)

[EugeneF]

We have one keyword that brings our site the most visitors.  This keyword is the brand name we carry.  We have several years of tracking it in Adwords.  For some extended time, this keyword [exact match] has averaged 19 cents per click, 2.7 average position, 4.5% click through, and a quality score of 7/10.

We wanted more clicks.   We could think of what was needed to increase the quality score.  Sure, we could change the meta tag title and the adwords title to be the same as the single word keyword, but this would be less informative.  We decided to keep these titles as phrases which include the brand name.

First change we made: we increased the bid.  After all, it was profitable for the two ads above us, right?  We increased our bid from .50 to $1.50.  Effect?  Average position increased to 2.3 from 2.7.  Click through increased from 4.5% to 4.9%.  Cost per click went from .19 to .51.   The incremental cost for each sale was......well really really high.....this didn't work. (oh, we rank #2 organically.  Our organic CTR dropped from 3.2% to 2.9% with this change as well)

Reversed back to where we were and decided to focus on the quality score.  We realized that the keyword was part of an add group with about 20 other keywords.  This word was important.....lets put it in it's own ad group.  We then made an "exact" copy of the ad and started up a new ad group.  Paused the old keyword.   We very quickly realized that the quality score on this "same" keyword was now 4/10.   That was odd....lets give it a few days......quality score drops to 3/10 and no longer qualifies for first page.   What was different we wondered?   AH!   We capitalized the first letter of the word.  Changing this took the quality score up to 6/10 instantly.  hmmm, we thought capitalization didn't matter?  Seems it did.  We now wait to see where the quality score goes.   Saga to continue....

[Michael-Goode]

Hey Eugene,

I actually had to contact Adwords myself asking for advice on how to gain a better quality score. I received quite a lengthy response that I will paste into here, I hope it can be of use.

• WHAT EXACTLY DO WE MEAN BY RELEVANCY OF A KEYWORD:

RELEVANCY OF A KEYWORD IS THE AVERAGE PERFORMANCE OF THE KEYWORD ACROSS ALL ACCOUNTS:

We look at the average CTR of the keyword in question across all accounts. CTR is used to determine the "relevance" that you see in the ads diagnostic tool. Why? The logic is that if people have clicked ads for this keyword, they must find it relevant. Otherwise, they wouldn't click. If a KW (keyword) has a bad overall CTR, it will be hard for the client to get a good QS number unless they have extremely good individual performance. Relevance, in this case, is NOT simply having the keyword on your landing page. This is a common misconception about relevance.

Relevancy is about how much users find it appropriate. In order words, how many users click on an advertiser's ad when they search for that keyword.

After thorough analysis of your account with our specialists, I have the following findings and recommendations for you:

1. CTR OF THE KEYWORD

As you can see from above points while defining the “Keyword Relevance” for a keyword we not only try to provide you with information on how “closely related” the keyword, ad and webpage are to each other/your service (defined by points 2,3 and 4) but we also try to provide signals about how good this keyword has been (in terms of performance) as a marketing keyword on google.com in general.

Please note, this keyword, I will like to confirm, has historically not performed that well in general on google.com and hence the same is reflected in the Quality Score of this keyword. However as we get more impressions and are able to keep a high CTR for this keyword in our account, this keyword’s performance in our account will start playing more significant role in terms of defining the Quality Score of this keyword and the same will improve.

2. EXACT MATCH CTR FOR KEYWORD IS LOW:

As you can see from above points (point 2) that the exact match CTR of the keyword also plays a significant role in determining the Quality Score.

I will like to inform you that besides the performance (read CTR) it is also the quantum of performance (read number of impressions) that impacts quality Score. The higher is the number of impressions you have the bigger role the CTR of that keyword will play in defining the Quality Score for that keyword. In this case, we can see that the keyword has received around 62 exact match impressions for the past 30 days, compared to around 1900 monthly exact match queries for this term, hence the CTR of the keyword is playing an important role here.

NOTE: It works both ways: that is to say a bad CTR with high impressions will impact Quality Score that badly and also a good CTR with high impressions will impact the Quality that well.

As this keyword receives around 0% of the traffic for this keyword in a month, the CTR at keyword level is not impacting the Quality Score of this keyword. However as the impressions will increase & if the CTR at keyword level is high it will have +ve impact on Quality Score.

As your Exact Match Impressions would increase and the number of clicks increases the Exact Match CTR would improve and will start impacting the quality score of the keyword positively.

3. GENERIC KEYWORD:

With broad matches, the Google AdWords system automatically runs your ads on relevant variations of your keywords, even if these terms aren't in your keyword lists. Keyword variations can include synonyms, singular/plural forms, relevant variants of your keywords, and phrases containing your keywords.

[EugeneF]

A follow up and update to my adwords adventure.

I the end, it seems capitalization did not matter.  The quality score of our keyword settled over those first few days at 3/10.  In trying to figure out more about quality score I found two incredible resources which answered many questions.

First was this blog post

http://www.epiphanysearch.co.uk/blog/decoding-the-quality-score-2/

This has a graph for each QS and charts position vs CTR.   This data gave me bench marks to understand what I needed to achieve with CTR.   "Broadbeach Media" replied to this post and reinforced the idea that CTR was the leading indicator of relevance.

I also read all of the blog posts at

http://www.clickequations.com/blog/

This was an invaluable resource.    So many time I spend hours doing searches look for answers to questions; Then I find a gem like this blog that answers so many questions it kept me up at night reading.

I didn't give up on my keyword.  I kept my bid high.   MY CPC was at 450% of where I started.   I split test my add like a madman.  Originally couldn't imagine a better ad than I had.   I made 6 different versions of my ad title.   I used a statistical significance spreadsheet (wish I could tell you where I got it, but the author didn't provide that information on the spreadsheet.  There are several of these available online).

As I gain significance each day, I was able to eliminate ads.  Just with title change, my best performer was able to get a 5.6% CTR (from 4.5% before).  Through out this process I watched the average position to make sure I had comparable data.  (I manipulated my bid to maintain a constant average position.)  Then I started testing a few change in the body of the ad copy.  Goldmine!  My best performer was now hitting 7.9%.  QS was increasing slowly over the 3 weeks. (QS is now where I was pre-test QS=7).  A few days ago, I let loose on the bid.  My average CPC is now the same as it was pre-test.  My average position is now 1.2 (from 2.7).  CTR is between 8% and 10%.   If my QS goes up again, I will be paying less per click than originally.  total clicks for PPC have just about doubled!