Brad Geddes / PPC Geek
Official AdWords Seminar Leader.
Author of Advanced Google AdWords.
Co-Founder, Adalysis.
(312) 884-9017
Brad Geddes's Theories on Marketing Quality Score is Based on Precise Match Not Exact Match

Quality Score is Based on Precise Match Not Exact Match

I made a comment on Twitter that stirred up some confusion; however, as tweets are limited to 140 characters; it seemed easiest to write out the explanation as a blog post and then let people read the rationale in its entirety.

The common assumption is that the CTR used in the search quality score algo is based upon the search query matching your exact matched keywords. Sometimes this is stated as ‘the search query exactly matches your keywords’. That statement is actually true, but sometimes confusing as Google has an exact match type; and the match type itself does not play a factor into determining quality score.

What if you do not have exact match keywords? How are those keywords assigned a quality score? You do not have to have exact match versions of broad matched words. Please note, this is not a debate about what is best practice, but about what is possible – there’s a difference.

The truth is that the CTR that determines your keyword’s quality score is based upon the user’s search query precisely matching the keyword in your ad group, regardless of the match type you use.

For instance, if you have the keyword ‘Google AdWords’ broad matched, then your ad could show for:

  1. Google ad words
  2. Google adwords help
  3. Google adwords alternative
  4. my adwords ad is not running on Google
  5. Google AdWords
  6. etc…

In this case, only the CTR of number 5, Google AdWords, would be used to calculate your quality score.

This same exercise can be applied to phrase and exact match.

This is also why if you have the same keyword as an exact, phrase, and broad match in the same ad group, they will have the same quality score. As quality score uses precise match regardless of match type, all of those keywords are assigned CTR information under the same conditions.

<Please note this next section is for advanced PPC marketers and could require paragraphs of explanation. It might not make sense to many of you – and if it doesn’t – that’s OK>

There is a difference between all three match types showing the same quality score and being displayed in the same positions.

For instance, one of the quality score factors is, “Relevance of keyword and ad to search query”. This factor is calculated after dynamic insertion is applied to an ad copy. Therefore, if for some reason a broad matched variation was doing fantastic with DKI, yet the exact match was doing average with that ad, it might appear that the broad matched word had a higher CTR than the exact match, which might result in a higher average position. Yet since that broad matched variation is not a keyword in your account, your broad matched word would still maintain the same quality score as the exact matched word that has a lower average position. This example is a rare case; however, it could happen.

<End advanced section – next section for everyone>

The same principles that apply to “Your Broad Match Keywords Are Not Converting Higher than Your Exact Match Keywords” apply to quality score. However, instead of examining conversion rates, just substitute quality score relevancy factors.

I hope that helps explain the confusion. If not, please add a comment and I’ll do a long blog post using the search query report and actual numbers to showcase how match types and quality score interact with the actual search query.

12 Comments

  1. Brad Geddes aka eWhisper
    October 1, 2009 at 10:01 am · Reply

    Negative keywords do not affect the CTR algo used in determining quality score.

    However, the statement that only exact match keywords are used to determine CTR is incorrect, as your broad and phrase matched words have to have a quality score to determine ad rank (QS & max CPC).

    Adding negative keywords so your broad and phrase matched words may increase the CTRs of your phrase and broad matched words, which will help them show for lesser searched variations.

    This statement is also why in the bid management interview I did with David Szetela the one slide on negative keywords is slighlty incorrect and you need to listed to the audio to hear the caveats.

    Bottom line:
    Negative keyword research is very important.
    Quality score is only calculated when the user’s query matches your keyword regardless of match type.

  2. Dan PPCPROZ
    October 1, 2009 at 9:49 am · Reply

    The source of the tweet thread began with Glenn’s post at…

    http://www.payperclicksearchmarketing.com/negative-keyword-slides/

    Thanks Brad for jumping into the conversation.

    So then, the Myth is not myth at all, correct?

  3. Dan
    October 2, 2009 at 3:50 am · Reply

    So…..at the end of the day, you would recommend always using all 3 match types everywhere?

  4. Brad Geddes aka eWhisper
    October 2, 2009 at 4:53 am · Reply

    That question goes to budget and your goals.

    If you are starting with a large budget, want to throw everyting at the wall, and then pair down what is and is not working; then a lot of broad/phrase is the way to start.

    If you have a smaller budget, and are trying to grow your spend based upon returns, then a lot of exact, with some phrase, and maybe a little broad is useful.

    Overtime, you will want mostly exact, with some phrase & broad and then refine using the search query report.

    As 20% of all queries have not been seen on google in the past 90 days; you can never have every possible keyword. If you’re hitting your budget everyday; then use more exact/phrase. However, if you want to reach everyone – you will want to utilize some broad match.

  5. Steve
    October 2, 2009 at 10:02 am · Reply

    Good stuff – really interesting read, all makes sense however I would love to read your post outlining how to use the Search Query Report to prove this.

  6. Cameron Wilson
    October 7, 2009 at 3:00 pm · Reply

    “If not, please add a comment and I’ll do a long blog post using the search query report and actual numbers to showcase how match types and quality score interact with the actual search query.”

    I think that would be a great post to read.

  7. Debbie Formica
    May 12, 2010 at 9:40 am · Reply

    Why is it that I get one quality score before I up the bid past my minimum amount, and usually a LOWER score after I have raised the bid ( from gray amount to black amount… past the minimum established amount).

    • Brad Geddes aka eWhisper
      May 14, 2010 at 8:46 am · Reply

      What’s common is that when your keywords are first added to your account, you have a default quality score which is based upon other advertisers.

      When you start to accumulate data, that quality score is replaced with your own.

      Often when you raise your bid past the first page amount, the keyword is shown more often and can accumulate enough stats to receive its own quality score.

  8. Jaap
    June 1, 2010 at 3:47 am · Reply

    @Brad,

    Thanks for the article I also read your comment on the article on search engine land of april 23rd : http://searchengineland.com/broad-vs-exact-match-types-the-hard-data-40053

    As you say there once in a while we know see different QS for the same words in broad vs exact match types. Have you got an explanation for this fact which is contradictory to what Google’s saying.

    • Brad Geddes aka eWhisper
      June 2, 2010 at 3:35 am · Reply

      Jaap,

      Sometimes it happens because the keywords have so few impressions that the ‘default’ quality score is used; yet a different match type might have enough data to have earned its own quality score.

      However, for the most part I have no idea about this. I’ve only seen it happen in a handful of accounts; and there doesn’t seem to be a matching pattern between the accounts.

      brad

  9. toni
    May 24, 2011 at 8:22 am · Reply

    Hi Brad, any updated comments on this?

    “This is also why if you have the same keyword as an exact, phrase, and broad match in the same ad group, they will have the same quality score.”

    I’m asking because recently I have seen different QS on the same search term on broad vs exact – exact usually higher by one digit.

    • brad
      May 26, 2011 at 12:33 pm · Reply

      The most common (but not the only reason) your QS varies by match type is the data available to Google to determine if you have enough date to determine CTR.

      I’ve seen a few other cases of QS varying, sometimes it has been how often QS was updated in an account – and other times no one at Google seems to be able to explain it.

Leave a Reply