Dueling Local Carousel CTR Studies: 48% vs. 30% vs. 10%

Mike Ramsey published a study (n=83) performed by search marketer Matthew Hunt on the effect of Google’s new Local Carousel on clicks and user attention. The study found that 48% of clicks were on the Carousel results.

At the bottom of his post Mike cites another study (n=102) from Ethical SEO Consulting where 30% of clicks were on the Carousel but a greater percentage were on the map. I received the data from yet a third study, which had the largest number of participants and saw the fewest clicks on the Carousel (10%).

Here’s the CTR map for the Hunt study:

Google-Carousel-study-overall-clicks

The Carousel gets most of the attention and clicks in the Hunt study. However, immediately below, in the Ethical SEO study, the first Carousel listing sees a lot of action but there’s more attention to the map:

Screen-shot-2013-06-21-at-11.01.40-AM

Finally, the third Carousel study, which was provided to me by someone who didn’t want to be named, involved 273 individuals — the largest sample of the three. Here roughly 80% of respondents were from the US. The rest were from a variety of other countries.

As mentioned, only 10% of clicks in this third study were on the Carousel vs. 30% and 48% in the other studies. A higher percentage of clicks came from the top organic results below the Carousel.

The prompt for the third study was “You’re in New York City and craving sushi for dinner. You don’t know where to eat, so you type “sushi NYC” into Google. The following page appears. What do you click first?”

Screen Shot 2013-06-26 at 4.18.20 PM

What these three studies indicate to me are the following:

  • The sample size matters (% of Carousel clicks went down as sample populations increased in these studies)
  • The question/prompt matters (the initial study above asked ““Click on the part of this google search results page that most interests you” vs. the practical scenario of the third: “sushi NYC”)
  • Further study is needed to definitively ascertain the impact of the Carousel

Having said all that there is consistency among the three studies: the Carousel gets plenty of attention as do the top organic results. And so does the map.

What would be your takeaways looking at all three of these?

You can follow any responses to this entry through the comments feed.

8 Responses to “Dueling Local Carousel CTR Studies: 48% vs. 30% vs. 10%”

  1. Mike Blumenthal says at

    This issues with the studies go well beyond just sample size.

    First and foremost is that most of these studies are “first click” looks. The carousel lends itself to multiple clicks (and I am reasonably sure that is one of the reasons that Google likes it). The issue is what does the searcher do/what action do they ultimately take. 

    Sample size would not be a problem if the sample is a representative one. I don’t think many of these studies were that concerned with sample make up. 

    They studies did not control for image quality and image make up. I am sure that the carousel will perform differently with different images. 

    Lastly is that none of the studies control for information below the carousel. Was it an ad or an organic result would affect outcome.

    In a very limited test of real users in front of a real screen,  I found that most users scanned the images, hovered over several and likely clicked several. In addition if there was a Yelp or Tripadvisor result present that seemed to seal the deal. 

    So while these results are useful and fast first looks they do, as you point out, require more analysis and further study. 

  2. Greg Sterling says at

    Agree. 

  3. Brad Simonis says at

    I’m surprised that the top organic results didn’t do better in the 3rd study.

  4. Greg Sterling says at

    These are just preliminary informal studies.

  5. Dave says at

    Those sample sizes are very small and limited.  We need large data to get a sense of this.   Additionally are there differences between hotel actions versus restaurant actions?   They will probably get the most activity.

    ….and lastly…there is a group that already has good data on this–>Google.  Why don’t you ask them to share with us?   I’m sure :D  they’ll be happy to do so.  LOL

  6. Greg Sterling says at

    It’s very true these are very small samples, which is what I say. However I shared these because they were interesting and showed some slightly different outcomes. 

  7. Service Central says at

    Greg, it might be too early to tell, but have their been an examples of local aggregators (YP publishers, directories, lead gen co’s) being able to get into these local carousels?

  8. Greg Sterling says at

    I think it’s pure local business listings. No third party “aggregators” that I have seen, which makes sense from the POV of what Google is trying to accomplish. It’s really the old “X pack” translated into a richer, horizontal display

Leave a Reply