« Paid Links Debate - The Experts Weigh In | Main | Sponsored Reviews - No Strings Attached Interview »

May 08, 2007

Google Quality Score Myths & Truths

By Account Deleted

In today's world, most SEM Gurus play the quality score game with Google and to some extent Yahoo. However, in order to impress a client or get kudos for a really innovative idea to improve CTR% or conversion, we are forced to some sort of A/B test, Ad multivariate tests and sometimes Day-parting (with the help of analytics). Typically speaking, one would think that all of this testing would wreak havoc on our goals for high Google Quality Score. Well, after a bit of thought and a quick call into Google, I can confirm that it's not as bad as we all may think it is. Let's discuss.

Day-parting DOES NOT AFFECT overall Google Quality Score because Google keeps the CTR% history. However, day-parting could backfire on you if the CTR% for that specific time is lower than it's original full-time running. So it's important to use analytics to verify when is the best time to segment the day parting.

Pausing & Deleting of Keywords:
At first, I assumed this was a no brainer that it would affect quality score, but oddly enough according to Google the deletion and/or pausing of a particular keyword/search term DOES NOT AFFECT Quality Score, because there is no loss of performance history and hence keeps it's own Quality Score.

Quality Score & Ads/Creatives:
When I attended the SES NY conference in April, 2007, one of the presenters mentioned that when testing Ads on Google, it's important to not change the Ads too much because you will lose history and eventually Quality Score. However, it's importance to note what really effects quality score at the Ad/Creative level is any change made to the landing page of a Destination URL field. (Note: This is different than losing performance stats that get recycled  when making a change to any of the Ad fields) To back up my claim, here's a real example from Google regarding an inquiry I had made about a past client of mine. "As long as the landing page of the Destination URL is not changed, merely adding additional tags to the URL will not affect your Quality Score or average cost."

So, in a SEM-nutshell, the skepticism of on-going testing (day parting, deleting/pausing of keywords and Ad Variations) does not hurt the ongoing efforts of Google quality score. Frankly, I was a bit surprised that Google continues to capture of history on all of these areas. I have to admit that I am impressed with Google's dedication to allowing it's customers test and improve performance of their keywords/campaigns, etc... Let's just hope that neither of these testing elements change in the near future, especially since it's getting more expensive everyday to compete in the space.


TrackBack URL for this entry:

Listed below are links to weblogs that reference Google Quality Score Myths & Truths:


> ONLY THE Destination URL which triggers
> the loss of History and performance.
> This was also confirmed by Google.

Are you talking about this from a quality score perspective? The ads definitely lose their history in reporting and every report you can run on ads shows it as a unique ad even if you've only changed 1 character.

Of course, the performance history of the Ad group and keywords are still there and you will likely not experience many problems updating ads because of this, but Google told *me* that changing ad creative does reset your performance history for that ad (basically creates a new ad).

Everything I've seen in experiments confirms this. I'm just curious what data you have to back this up.

David, thanks for the comments. With regard to the Ads/Creatives you are correct about seeing the resetting of the performance data of the Ad. The point I am trying to make is that according to multiple Google reps. that I have spoken with in both the past and present, have told me the only field in an Ad/creatives that would affect quality score/history is the destination url field only and not title, line #1, line#2 and display url.

With regard to how I back up this claim, I was advised by Google. (as this was an issue with a client) It appears that this issue has created another level of confusion by Google to differentiate how Quality score is measured at the keyword level vs. Ad Level and that would make a great blog post in the future. We can definitely dig deeper into this with Google to see why there's a discrepancy in what we are talking about.

Strange that we're getting different information from Google. This is actually an important issue for me as I change my ad creative across the board quite often. I'm going to contact my Google rep again and see if I can get clarification. It's been a while since I tested this so I might give that a go as well. I'll let you know what I find out.

Thanks for clarifying the pause/delete question. I had always wondered that.

While changing the creative may not technically reset the whole history, my experience is that any changes, even 1 character force the ads back through editorial, at which point something that was approved earlier, may not be approved this time. Also, at one point, it was my understanding, that ads in the editorial phase were not syndicated across the search network. Not sure whether this is still the case. One time, a bug forced the creative to get hung up in editorial for a month and cost my client $$$,$$$ in revenue. Google did admit it and compensated us $,$$$. My point simply being, that I always have serious hesitation changing previously successful creative for a number of reasons, especially in adgroups that drive most of my revenue.

Guess I'm eating a little crow now. Per a conversation I just had with Google, quality score history is of course just tied to keywords and not to ad creative.

Changing your ad creative may change your quality score on the keyword level as click through rates, etc. change with the new ad creative.

Any changes in quality score will effect your ad's ranking (min bid) in a short amount of time, but any historical factors used to determine quality score are tied solely to keywords. In short, changing ad creative will not change any historical factors related to your quality score.

Thanks for posting on this Greg. I think I was working off of some old information so it was good to revisit this issue and clear things up.

> even 1 character force the ads back
> through editorial, at which point
> something that was approved earlier,
> may not be approved this time.

Tell me about it. Yahoo is the worst. I always leave at least one static backup ad running in each Adgroup in case I have editorial issues changing my variable ads.

Google definitely does provide god information on Quality Score factors however, we need to analyze the factors that they state and understand the underlying factors that can actually improve the Quality Score ( http://seo-kolkata.blogspot.com/2007/05/how-to-improve-your-google-adwords.html
) . I have tried to create a documentation on how to improve Adwords Quality score which covers most of the points but am sure there is much more to it. Would appreciate your comments on this.

You may be interested in a 10-page report (PDF file) that Jeremy Palmer has just published, called the Unofficial Guide to Google's Quality Score - http://www.quityourdayjob.com/qualityscore.pdf

He goes into a lot more detail.

(I'm an affiliate of Jeremy's, but that isn't an affiliate link.)

It's good Google provided a little more details on the Quality Score, but even thou QS will remain a big mistery for all of us. Some reverse engineering can be done anyway :). My conclusion is that a highly optimized page in terms of SEO, will have a great quality score. One of my colleagues wrote some articles about Google's QS. You can find them at http://www.pitstopmedia.com/sem/google-adwords-quality-score-part-1 Google Adwords Quality Score - Part 1 and Part 2 - http://www.pitstopmedia.com/sem/google-adwords-quality-score-2

Thanks for this article! In regards to the pause/delete question I always thought it was best practice to remove keywords that performed poorly. The idea is that the removal of keywords with a low CTR would improve quality score. I thought by reducing irrelevant impressions you would see an increase in Quality Score. Am I mistaken with this assumption?

The comments to this entry are closed.

Get SMG Today - Free!

Get SMG by RSS What Is RSS?
Get Search Marketing Gurus Today via RSS! Add to Google Reader or Homepage
Add to netvibes
Get SMG in Your Bloglines
Get SMG in Your NewsGator Online

Get SMG by E-Mail
Subscribe to SMG via Email
Enter your email address:

Delivered by FeedBurner

SMG Conversations

If You Like SMG Favorite Us on Technorati!
Add to Technorati Favorites
If You Like What SMG Has To Say, Joins Us At These Places!
Subscribe on YouTube to SMG's Videos
follow Li on Twitter
Follow Li on FriendFeed

Copyright 2006, 2007, 2008 SearchMarketingGurus.com