17 responses

  1. Lori
    September 13, 2011

    Great article! Thanks a lot! I really appreciate your effort.


  2. Adam Audette
    September 15, 2011

    Great stuff, Ted. Best analysis I’ve seen of CTR by ranking position. Thanks for the sample spreadsheet – will play with this a bit.


  3. Dev Basu
    September 20, 2011

    Thanks for the analysis and offering up the spreadsheet to go along with it. I particularly liked the way you accounted for conservative estimates by implementing constraints.


  4. Rank Pursuit
    September 21, 2011

    Very thorough job here. It’s kind of amazing to me that there isn’t more discussion about abandon rate. I guess it;s not something you really want to talk about with clients specifically but instead just incorporate it into estimated CTR. I think you’re about as spot on as you can be with this estimation process. You kind of follow the saying, “Under promise and over deliver” which I love. One thing with branded keywords, we have to keep an eye on the new Sitelinks in Google search. They’ve popped up in Google searches for keywords that may not necessarily be brand oriented but they match a domain so the results are pushed down because of it. This could change tho so it may not be something to really worry about.


  5. Zunaira Karim
    September 25, 2011

    When you first pinged me about this article, I knew it was going to be an awesome read that I would have to take some time with! Great post Ted!

    Your section on whether clicks go to Paid or Organic search was really interesting. Whether Google provides accurate segmented data or not, we can never be sure. But I think the most we can take from the Google Keyword tool, is regardless of whether volume for a particular keyword mostly comes from Organic or Paid, we can still assume it to be a ‘searched’ term and work towards that.

    I remember us discussing CTR for title tags and meta descriptions a while back on your meta description post…glad you touched on that!


  6. Ted Ives
    October 17, 2011

    Important note to all: I was having lunch with Dan Shure from http://www.evolvingseo.com and when I ran through this with him, I realized – the calibration was done against *broad match* data, not *exact match* data – so, to use the spreadsheet you need to use broad match data from Adwords. I’ve clarified in the posting and uploaded a spreadsheet with corrected instructions.


  7. Will
    October 18, 2011

    Something I’ve been thinking about for a long time. I will use this spreadsheet and report back. Thanks


  8. Peter
    October 21, 2011

    Hey Ted,

    Thanks for this very interesting article. I too have been finding that organic traffic has dropped for the top 3 positions. Now I’m a lot more conservative when estimating traffic for SEO clients. I’d prefer to look at exact match and take a lower 15% as estimated traffic. Anything more would be a bonus to my clients. Will check out the spreadsheet too. Cheers


  9. Grant
    November 7, 2011


    Great job!

    Been working through a few theories myself in an effort to better quantify SEO effort vs. reward.

    Bottom line is that many SEO initiatives may not be short term wins based on cost of effort, but without qualifying the potential opportunity selling *anything* is a tough pitch.

    At least having some basis around opportunity capture possibilities based on the ‘halo effect’ of well-implemented SEO gives better directional metrics on feasible ROI (and gives our sales guys something to propose and our teams some targets to measure against)



  10. Dennis
    February 28, 2012

    Interesting BUT…

    There’s one factor that should be taken in account and that’s the psychological aspect in all of this.

    I DONT click on paid ads. Why? Because if ‘m searching for info I’m not going to the place where anybody with some money can shout it out. I think “correct me if I’m wrong” that if I look at my own behavior and most likely of everybody that left a comment here (professionals) will almost never click on the paid ads for that same reason. So for the B2B market I think the CTR for paid ads are very low even if they are the first 3 results.

    As for the B2C market, I think it will go the same way like with telemarketing. People start to identify what is an add when searching and will leave them for what they are, Commercials! I think it was effective (if I would believe Google) when it was new, but at the end of the road it will be the same as all the advertising received at home, it will end up in the round file cabinet 😉

    I even have comparising for my statement:
    I have a website that ranks nr.1 for 3 keywords resulting in around 1000 unique visitors a month and growing. In a HIGH competive market (tablet pc’s), so there are always 3 paid results above me 😉
    According to Google those 3 keywords would be good for arround 2500 searches (EXACT) a month. So that means that I’m getting about 40% off all search traffic since there are NO other marketing efforts for the domain (No paid, No Social, NO OFF PAGE SEO, NO nothing) I did al this exactly for 1 simple reason. To decide what every marketing effort is worth. I want to know exactly what the ROI will be for everything I do.

    So, for a very competive keyword I harvest 40% for a number 1 position. That’s a little higher then Google wants me to believe 😉 cause they rather see me pay for that kind of result offcourse. But with Adwords I can never get those results and they would cost me a lot more than I’ve paid now 🙂

    I think there can be also other reasons why CTR is dropping. Think about mobile searches, people search more on their mobile now a days. But when searching on a mobile you can be called (which it was made for) and therefor result in abandoned searches. And most likely you can think of a lot of other reasons due to an ever changing search experience. Youtube, shop results, and so on also show up in the search results.

    But what I would like to know now…is if the amount is dropping down for only the “relative” percentages. I mean, the number of searches can be growing while the CTR % drop. Still that can mean that the number of click thru stays the same or even grows. And I use the same example of my website again. 3 months ago the number of searches a month were 2000 for those 3 keywords combined and now 2500, that’s an increase of 25% so even if the CTR would have gone down from 40% to let’s say 33% it would look something like this:
    2000 * 40% = 800
    2500 * 33% = 825

    So even do the CTR drops, my vistor count went up with 3,125% 😉

    As you can see, the numbers are up for interpretation and can be wrongly interpretated 🙁

    I rther use facts (my own) than figures (from Google) 🙂


  11. Cody Sharp
    August 17, 2012


    This is fantastic work and I have started using your spreadsheet for my work. I’m wondering though, as Google gives less and less real estate to organic search terms, what kind of additional drop will we see in click through rates. I’m hoping very little but I feel like it will grow appreciably. Have you seen any change since you first wrote this article?


  12. Ted Ives
    August 17, 2012


    Great thinking – I can take the sites I calibrated it with and re-look at their data and see if the estimate comes in above the actual now, and if so by how much (maybe like an annual calibration, or measurement of what you’re talking about). I think I put this out last September, so I will take a look in a few weeks, great idea!

    This article I did on that topic is definitely up your alley, I did some back-of-the-envelope high-level guessing there:


  13. Mike
    September 7, 2012

    I read one of the posts here commenting on the traffic estimates and whether people look at natural or paid results. I deal with the public on a daily basis and ask almost everyone. My best guess is 60-80% look at natural results.


  14. Jonah Keegan
    October 9, 2012

    Fantastic analysis… my one question is why broad match? Seems like you are jumping through a lot of hoops to try and make a dataset useful (broad match search volumes from the AdWords Keyword Tool) that is not held in high regard by many search marketers. Why not analyze the Keyword Tool predictions of phrase or exact match against your website actuals? I would think the match type with the strongest correlation would be the best one to model with…


    • Ted Ives
      October 25, 2012

      Well, I’ll give two answers, my excuse, and the honest one.

      1.) Since a page will always attract some long tail traffic along with traffic associated with the keyword it’s targeting, broad match makes more sense.

      2.) When I did the analysis, I accidentally had Broad Match on and didn’t realize it until I was well into writing the article! Otherwise I probably would have picked Phrase Match actually for the calibration, I think it would probably be more accurate than a Broad-Match based one. Either way, as long as it’s calibrated, it’s better than the alternative (i.e. nothing).


Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top
mobile desktop