We've all been there. Trying to improve our organic rankings so we can get more traffic from the search engines. And every time we do that, we are left with some big questions in our minds:
- How much traffic would I actually get if I rank on the first page?
- Is it worth my time trying to rank above the fold?
- How much more traffic will I get if I rank first in the organic results?
I've been there, too. I felt overwhelmed and frustrated every time I had finally reached a number one organic ranking in Google only to find out that the traffic coming from the search engine was not making the big difference I was expecting.
So I started searching for a way to find out how much organic traffic I could get for ranking on the top positions in Google.
But I faced a big challenge. These days, with "not provided" being almost 100%, it's very hard to measure how many people reach your website searching for a certain keyword.
So I turned to the best source that I could get this data from, Google Webmaster Tools, which allowed me to see how many people click on my website when searching for the keywords I am interested in. This saved me a lot of time and allowed me to make better choices in the future with the keywords I was targeting.
Sounds like something you would be interested in?
Read on to find out more about how my initial findings turned out into a full fledged organic CTR study and how you can use this data to make better and more informed decisions in the future.
TL;DR: This will be a long post, so for those of you who are anxious to see the results of this study, scroll down to the CTR Study section below. Alternatively, you can download the complete study in PDF format or check out the free Google CTR History tool we have built to aid with this study.
Previous CTR studies
This is not the first study of its kind. There have been a number of studies in the past that have tried to find out the CTR for organic results. It all started when AOL released more than 20 million search queries made by more than a half-million users in 2006.
A number of studies followed after that, including those from Enquiro (now Mediative) in 2007 and later by Chitika and Optify in 2010. More recent studies have been performed by Slingshot in 2011 and then Chitika and Catalyst in 2013 respectively.
Here is a comparison of the Click Through Rate for each study:
It's important to emphasize the major differences in the methodologies applied for each study, as they are the main ingredients responsible for the dissimilarity of the results:
It's worth noting that the studies conducted by Mediative (former Enquiro) and Chitika, have been executed through unique methods that cannot be truly compared to any of the other studies. Mediative's study relies on survey data and eye-tracking research, while Chitika's studies are based on ad impressions served within their network.
Also relevant for a comparison is how CTR is defined for the other three studies previously conducted:
- Optify defines CTR as "the percentage of users that clicked on each position, given that a user clicks on a top 20 organic ranking." Their study makes the assumption that all searches result in a top 20 organic click.
- In the Slingshot SEO study, CTR is calculated as "total visits (via Google Analytics) divided by total searches (via Google AdWords Keyword Tool) for a given keyword over a stable period."
- For the Catalyst study, CTR is defined as "the percentage of impressions that resulted in a click for a website (via Google Webmaster Tools)."
Our study retrieves the CTR data from Google Webmaster Tools so comparing it with the Catalyst study would be the most accurate.
So why a new study?
First of all, the Google search results have evolved significantly since these studies were performed. Besides having a fresh set of data, we also wanted to make this study unique.
- Unique
This study is unique because we have segmented the queries to be able to see how the CTR is affected by different types of searches. For example, we have segmented the keywords by category (industry), search intent, number of words (long tail) and whether the keywords are part of a branded search or not.
Another important section of this study is trying to find out what impact some features that appear in the SERP (such as ads) have on the organic results CTR. - Accurate
To make sure that we get relevant and accurate results, this study is based on search data coming from Google Webmaster Tools for 465.000 keywords and 5.000 websites. - Transparent
This study was intended to be as transparent as possible. Thus, we have included our step-by-step process below so you can see how we arrived at our results.
More than that, we also decided to give away the entire set of data so you can do your own research. To protect our clients, the actual keywords have been anonymized in the data set. - Up to date
As we have seen with previous studies, the organic CTR changes in time due to various factors. It can be affected by the holiday season, or by more features that are constantly being added in the SERPs.
This is why we decided to transform the initial study into a free tool that anyone can use to segment the data and watch how the CTR changes in time.
Read on to see how different types of search results influence users' behavior and what role the user intent has in determining the distribution of clicks.
Our methodology
Here's how we obtained this data in case you want to do a similar analysis for your own websites:
- Download average search query data from GWT
The initial data was obtained from Google Webmaster Tools (GWT) with the default filter: Web. This includes only traffic coming from non mobile devices. Our data set includes only keywords that have at least 50 impressions per month.
We then changed this filter to Mobile and downloaded the table again to get CTR data for mobile devices.
The Avg. position column from GWT displays an average of all ranking positions that this keyword has appeared in. This data was used to build the section of the charts.
- Download exact search query data from GWT
In GWT, when you click on a keyword in the Search Queries table, you will be sent to a report called Query Details. This report provides the CTR for each exact ranking position for that keyword.
For example we can see here that every time this keyword was ranked first in the search results, the CTR was 56%. That's because 2,947 people searched for it (Impressions) but only 1,644 people actually clicked on it (Clicks).
- Exclude from exact data the queries with less than 500 impressions per month
This was done to ensure that we get accurate CTR results. A filter was also applied to include only the keywords that had at least 10 impressions per month for each exact position they appeared in.
- Categorise queries based on brand, search intent and number of words
We wanted to see how the CTR changes for searches that contain branded keywords. Most brands rank first for their brand keywords and it is believed that people tend to click on that first result.
For this study we have defined brand searches as searches that contain the entire domain name of the website in the query.
The same thing happens when people include a search intent in their query. It is believed that people act differently when they are interested to buy something as opposed to looking for information about something or when comparing different things.
How can we figure this out? We look for certain words in the search queries, trying to guess what the intent was for that search.
There are three types of search intents that were included in this study:
Informational
This includes searches that contain words like: what, when, where, how, who, restaurant, hotel, flight, definition, define, review, news, weather, time, phone.
Commercial
This includes searches that contain words like: buy, purchase, order, shop, coupon, cheap, cheapest, expensive, pricing.
Location
This includes searches that contain words like: near, nearby, from, directions, how long to, how far away from, how fast, train station, airport, ferry, route, highway, toll, plane tickets, flights, maps, driving directions.We have also tracked long tail queries (more than one word) separately to see how they affect the CTR.
- Find out if the SERP contains ads
We matched the entire set of keywords from Google Webmaster Tools with the ones we track for each client in AWR Cloud. This way we were able to get more information about the features included in the SERP, such as the number of ads and their position and if any Universal features were included in the search results.
- Create graphs for easier data analysis
We first used Excel to display this data in charts but in the end we ended up creating an in house tool because we realized that it would be interesting to see how the CTR changes over time.
Assumptions and limitations
The sample data set that was extracted from GWT belongs to our clients. Their businesses, although variate, may belong to certain industries that are different than the industry you are in. Therefore the results may not be the same for every business.
This study measures the CTR that was observed for a special time frame (within the month of July 2014). That means we cannot predict how the CTR changes for keywords that have higher volumes in different periods of the year.
In this study, we also made the assumption that the data collected from GWT with the above methodology is accurate.
The CTR study
This is the reference chart for the click-through rate (CTR) of organic desktop searches in Google for July, 2014.
It is important to mention that these numbers reflect the CTR across all the searches included in this study. They do not account for the user intent, the features that appear in the SERP, or whether the keywords used in the search included a brand name. These will be addressed later in the study when we segment the data.
On average, 71.33% of searches result in a page one organic click. Page two and three get only 5.59% of the clicks. On the first page alone, the first 5 results account for 67.60% of all the clicks and the results from 6 to 10 account for only 3.73%.
"These numbers serve as a useful reminder of the importance of organic rankings, and reconfirms the importance of the top few positions on Google. Although the first spot is still the most valuable for CTR, it seems to have become less so. I'd guess that part of the reason is that the increased use of ads, universal search results and Google's own comparison and shopping results have reduced the prominence of top slot." Graham Charlton - Econsultancy
In case you wonder where the other 23.08% of the clicks are, here are some possible scenarios:
- Some people may find the ads displayed above the organic results more relevant.
- Some people may not find what they are looking for in the first 10 results so they click on results from the second or third page instead.
- Others may not find what they are looking for at all so they refine the search adding more words to the query to be more explicit.
- With Google providing more and more instant answers people may very well find the answer to what they are looking for in the displayed search results so there is no need for them to click on any of the results.
Mobile
Mobile traffic is getting bigger and bigger day by day. Here we can see the CTR for searches coming from mobile devices compared with the searches from desktop devices.
Given the fact that you can see fewer ranking results above the fold on mobile, people have assumed that the CTR would be higher for the first results on mobile devices. Let's see if that is the case:
Not only is the CTR slightly lower on the first page, but the CTR for mobile searches actually rises on the 2nd and 3rd page, which is opposite to what we would expect and see from mobile searches.
"I would've expected mobile to drop off much, much faster than desktop. These rates seem to imply that the first positions on a mobile results page are less significant than we thought. Does that mean people are scrolling more?" Ian Lurie - Portent
Branded vs. unbranded
One might assume that when users are making generic searches on Google, they end up making a brand selection from the results retrieved. They choose from the handful of options received, the source of information or provider to trust in for satisfying their need.
But what happens when branded searches are made? If the users are clearly looking for information related to a specific brand, will they follow the same behavioural pattern as for generic searches?
For branded searches the first result is almost always associated with the brand's website, which makes it the obvious choice for most users and very hard to miss. This would justify the big CTR difference between the first position and the rest of the SERP.
This big difference in CTR may also be affected by the fact that brand searches usually display a pack of 6 site links just below the first result, making it more prominent in the search results.
"People will seek click on a brand in the first position for a search on that brand way out of proportion to all other positions." Danny Sullivan - Search Engine Land
"The CTR data coming straight from Google suggests that we should be even more conservative when estimating potential search traffic. Most of our keyword research is going to revolve around non-branded terms. If you study the data, you'll see a dramatic difference between CTR for the #1 position of branded vs. non-branded search. Our views of how many clicks you will get with an average position of 1 may be skewed because of this. But now with this segmentation data, I know I will be viewing traffic potential even more conservatively based upon CTR of only non-branded keywords." Dan Shure - Evolving SEO
Search intent
Most of us have some sort of intent when we search for something. We may need to find the location of a restaurant or a better price for that big TV we always wanted to get in the living room.
It is believed that people who search for keywords with high commercial intent ("buy 4k LCD TV") are more likely to click on the first results than people who perform basic informational searches ("where is the nearest thai restaurant").
Let's see if search intent does indeed affect how people click on the results.
This chart reveals that people tend to click more on the first results when their search has a specific intent. So we wanted to dig deeper and see which of the search intents affect the CTR and how.
"Google uses a lot of context cues beyond the keyword so if I type 'restaurant' the intent isn't there, until you realise it is midday and I'm on the street searching on my iPhone. This might explain the significant uptick in clicks on positions 1-3 for searches with intent." Tom Anthony - Distilled
The "Specific Intent" in the chart above is the set of all keywords found in the Informational, Commercial and Location sections and the "Other Intent" means all the other keywords.
The following chart compares these three search intents and how they affect the CTR:
Google is getting better and better at figuring out search intent. Nowadays, many of the search results contain instant answers so people no longer need to click on a website to find out what they're looking for. The answer is already there.
Commercial intent searches usually trigger ads that have colorful pictures of the products we search. It's usually a lot more tempting to click on these pictures than on the first organic results.
"Search results for commercial intent keywords usually contain more features (eg: pricing, ratings, shopping results) which might dilute the CTR across the page." Richard Baxter - Builtvisible
"It's interesting that commercial intent searches have a lower organic CTR than informational searches. We've seen the opposite hold true for paid CTRs. This may be because commercial intent KWs are more likely to trigger ads, which lower the organic CTR." Mark Irvine - WordStream
Estimating organic traffic based on CTR
Remember the initial goal of this study? To find out how many organic visits one could receive for ranking in the top results on Google. We are now closer to reaching our goal.
By knowing the CTR for each position in the organic search, we can now calculate the organic traffic potential of a website. Depending on the ranking of a keyword and how many people click on that website, we can easily calculate how many people would reach that website from organic search.
Theoretically, by taking into account all these factors, one could easily estimate the amount of organic traffic. The formula is quite simple:
Traffic = Search Volume * CTR
But things get a little complicated when taking into account that each keyword is different.
As this study showed, searches for branded keywords have a higher CTR. Search intent also affects organic CTR significantly and long tail keyword searches show higher CTRs for first page listings.
Let's see an example for an unbranded keyword with a volume of 1,000 searches per month where you rank first in the organic results with no ads above you:
1,000 x 24.8 / 100 = 248 (visits per month)
where 24.8 is the CTR for the 1st position for unbranded keywords.
Applying this formula for each keyword, enables you to estimate the amount of organic search traffic for any website.
Where can you get this study from?
This post contains only parts of the actual study. To find out how ads affect the CTR of organic results and more, download the complete Google Organic CTR Study in PDF format.
You will also get access to the entire data set that we used for this study if you want to do your own research.
Future developments of the study
We will be constantly adding new features to this study, such as more ways to segment the data or insights on how different features that may appear in the SERP affect the CTR. These new additions will be featured first in the free Google Organic CTR History tool, so make sure you check it out.
The first thing we want to tackle next is how the features that appear in the Universal results (such as news, videos, places, etc.) affect the CTR. We will then dig deeper to see how the CTR is affected by carousels, answer boxes and other knowledge graph features that appear in the SERP.
Your turn
Is there something in particular you would like to see in further updates of this study?
Post your comments below and let's find out how we can improve this tool to benefit the entire community.
Interesting stuff, and I really like the breakdowns. I know CTR curves have gotten a lot of flak recently, and many of the criticisms are fair, but I'm more and more haunted by a question: "Compared to what?"
Yes, there are many limitations to generalizing a CTR curve, BUT (and it's a big one) the problem is that we all already have a CTR curve in our heads, and for many SEOs, that "curve" is a straight line. When we count rankings as 1, 2, 3, etc. we treat them as falling on a CTR line. To me, even a best guess at a curve is closer to reality than the way we've counted for years. Is it imperfect? Oh, yes - absolutely, but it's much less imperfect than the thing we've all just come to accept and live by professionally.
Philip, thank you for doing this - it's very much appreciated. The one question I had was if you had used "all web search data" and "all mobile data" without selecting country/region? If so, do you think separating by countries would have produced different results?
Hi David, thanks for your comment. Right now we have two segments: all data (international) and US only. All data here means all searches that were perfomed regardless of their location. The graphs in the study are mainly made from all data but you can check out the free tool that allows you to segment the data to see only the searches in US. We can also add UK searches next month if people are interested.
I'll definitely check it out, and provide credit to you when I use the information in your post for forecasts. Also, try not to pay attention to the negativity, behind closed doors this post is being praised by professionals, just trust me on that ;) I am very curious to see if CTR changes per country, although we'd never know why....
Dear Philip,
I like your post very much. It's a subject which i studied and read a lot about and I truly do think your graphs and explanations add value to my own research. I use my own research mainly and only for convincing my clients do make some though choices.
Have you ever thought about looking into the amount of effort it would take to increase rankings versus the growth in potential visitors versus them turning into clients and earning your client some money? I would love to see an in depth research about that.
I have some clients who rank in the Top 5 but would love to get to that 2nd or 1st position but when looking at the competition it would take a lot of effort getting there. Looking at your general statement about 31,4% of searches click the first result it could possibly increase traffic big time, however as stated before, the amount of effort you have to put into it would also be significant.
Lets say you have to put in about 20/25 hours of time into getting those rankings up. At a cost of around $ 100,00 / hour you would be looking at an investment of $2000/$2500 for getting the result up from position 2 to 1. That's an increase from 14,04% to 31,4% = 17,36%. Lets say the keyword specified gets around 8000 searchers each month. Than the theoretical increase would be around 1390 new visitors.
To be safe I recommend clients to take 1% conversion rate on there website (sometimes even 0,5%). That would mean that the new visitors would possible generate almost 14 new orders. If you sell some products on your website at an average price of $40,00 with a profit of maybe $8/10 per product then your total new sales would be around $560 with a profit of $112/140. Then it would defiantly take some months before you earn back your investment for just 1 keyword. Still the changes I would make could also effect some more keywords and thus have a lot more increase in traffic and thus in sales and profit.
I am truly wondering about your vision on this.
Have you also taken into effect that a website could rank with multiple places for a single keyword? I have client who rank 1 to 4 for 1 particular keyword. Could i then be safe and tell them that an approximate 50% of searches find his website? Or have you diversified these finding into the results?
Hope you completely understand what it is that I am saying.
Hope to hear from you.
Regards
Jarno
Hi Jarno,
Ultimately, it all goes down to how much of the search engine traffic you actually convert into money.
It does not really matter if you're number one if nobody clicks on your website. Also, it doesn't really matter if you get 1.000 visitors a day from a number 1 spot as long as none of these visitors actually convert.
So you're right, we do need to do these calculations ourselves. In fact I believe most of us do. We keep track of conversion rates, average order value and match those with keyword volume and traffic data.
I believe, on a broad level like in this study, these calculations would be hard to do, because there would be a lot of speculation in the actual conversion rate or average order value. But on per client level, these calculations not only can be done, but they are a must in order to figure out how well you are doing.
In your example, ranking number one for that keyword would definitely bring some value, but you also need to take into account the stage that the visitor is in. This is closely related to the keyword intent. Is the visitor looking for information about your products, is he comparing products or is he ready to buy?
I find this data even harder to use sometime because the keywords that people search for are different in all these stages. They will probably search for "the best product" for what they are looking for at one stage and then search for your "brand name" once they are ready to buy because they already know what they are looking for.
When ranking in multiple places for the same keyword, it's usually safe to assume that you would get the traffic that is associated with the highest position. That is because people will only click once on the search results.
These are some very good comments, BTW. Thanks for asking.
Hi Jarno,
I think you move the discussion into a very interesting direction here. As we will never be able to optimize for all keywords, we have to choose which ones to prioritize. And then the question is, how do we prioritize?
The answer IMO is a function of 5 factors -
a) the amount of work(=investment) it takes to achieve an increase 1 position
b) the increase in traffic by improving 1 position
c) the amount of searches per month for the keyword
d) the conversion rate for that traffic and
e) the value per conversion for that traffic
Sounds almost too complex to be feasible - but let's try anyway. So to make a quick example, lets assume we have the same conversion rate and Value/Conversion (d and e) for all SEO-traffic.
Example: We are on #5 for "Japanese knives", which has 1.000 impressions/month and on #4 for "Asian knives" which has 2.000 monthly impressions. According to the list of CTRs in your table, improving from position 5 to 4 would bring an increase in CTR of 5.5% to 6.97 % = that is 26,7%. Now we get the calculation for
1. "Japanese knives":
-Position improvement from 5->4
-Impressions/month: 1.000
-CTR before: 5.5%, CTR after 6.97%
->Traffic before (5,5%*1.000)=55 visitors, now (6.97% *1000)=70 visitors
->Increase: 15 visitors
2. "Asian knives"
-Position improvement from 4->3
-Impressions/month: 2.000
-CTR before: 6.97%, CTR after 9.85%
->Traffic before (6.97%*2.000)=139 visitors, now (9.85% *2.000)=197 visitors
->Increase: 58 visitors
If, as we said, conversion rate and conversion value is the same for both, we can invest up to (58/15)-1 = 287 % more time in optimizing "Asian knives".
If, secondly, we an index that tells us how difficult it is for each keyword to move up 1 position (and that index is a crucial part of the whole system), we can put that number against the percentage above (287%) and can actually decide which keyword to go for first. At last..
First, I want to thank you for posting this in-depth study. It's always fascinating to see how users are behaving in the search results. I think we all recognize that any study like this will have limitations.
That being said, I don't think you can reach this conclusion without knowledge of what happens the other 23% of the time: "On the first page alone, the first 5 results account for 67.60% of all the clicks and the results from 6 to 10 account for only 3.73%."
67.60% represents the sum of the CTRs of the first five organic results. That doesn't mean these results are getting 67.60% of the clicks. That means that one of the first five results is getting a click 67.6% of the time. The only way we can make the statement "the first 5 results account for 67.60% of all the clicks" is if we are certain that every Google search results in a click (which isn't true).
According to this data, approximately 77% of Google searches result in an organic click (71% 1st page, 4% 2nd page, 1.6% for 3rd page and beyond). If organic CTR is 77%, and the first 5 results have a 67.6% CTR, then it seems that the first five results would actually be getting 87% of the organic clicks. What percentage of all clicks are they getting? We have no idea (but we do know it is between 67.6% and 87%).
You go on to say this:
"In case you wonder where the other 23.08% of the clicks are, here are some possible scenarios:
This second bullet point here is already factored into the data, so it does not apply to the "other 23.08%."
In reply to : Michael Martinez
Interesting point of view, but somewhere (especially this type of case studies) is moreover reverse engineering (while Google didn't show us everything - but Science always be like that :) )
In fact, if this studies is not "scientificaly proved", it shows an interesting method that should win, be updated with some big data requests... The point that bigdata could, (not every time) using to show tendancies very closely to a "scientific proved final result").
Since English isn't my native languages (and as i havn't worked with english people those last years) sorry for my mistakes languages :)
Really interesting test Philip. It's next to impossible to get accurate data on this these days.
I think we all know that those top positions dominate the amount of clicks.
The only concern for me is that differing industries have different CTRs relative to positions. Amazon, as an example, will distort the CTR of any SERPS they appear in thanks to the brand loyalty effect.
I think it's underestimated just how much of an impact that can have.
Thanks for the post!
Thanks Sam, I'm glad you find it interesting.
I totally agree with you here. Different industries do have different CTRs. If you look at the free tool that we built to complement this study you can see a breakdown by keyword category or industry. Hope you find it useful.
Philip
Your a superstar, thanks will reviewing.
Hi Philip - thanks for posting. I always love data focused Moz posts.
The second image under the section, 'Previous CTR studies', is inaccurate. It looks like you mislabeled the 'Catalyst 2013' column. The Catalyst CTR study should be the column with 17,500 keywords, 59 websites, etc.
Hi Andrew,
Thanks for the heads up. I will update the picture ASAP.
I think it's pretty safe to say that studies like this aren't trying to revolutionize SEO or prove that ranking #1 is the best place to rank. We already know that. It's not earth shattering. I think it's also safe to say that SEOs understand that an SEO strategy should include data regarding the individual site they are optimizing.
If you are looking to optimize your site and need this study to convince you that you should rank #1 for important terms, you're doing it wrong.
I think this study is good for adding to the data the industry has already compiled regarding CTR. In SEO you should never look at one study on one factor and rely solely on it for your strategy. However, researching many of these studies across many of the important factors will begin to give you a lens through which you can look at your site and begin to optimize based on the specific and unique needs and performance of that site.
If young SEOs take anything from this comment thread, I hope it is that SEO is still the Wild West and we are still figuring it out (and likely always will be). Keep researching. Keep testing. Don't rest on laurels. And there is no magic formula.
Michael - I have to say that I don't totally disagree with your points (even if I'm not a fan of their presentation). Google has gotten so good at personalizing search that it seems like a study like this is irrelevant to any single site and it's SEO efforts. I agree with that. But that doesn't mean this data is useless. I will take the assumptions and conclusions made by this study and test those against what I am seeing with the sites I'm working on.
I though you made a good point regarding SEOs looking at organic rankings like a Paid Search pro looks at CTR vs. Conversion Rates on ads. That is interesting and important. A lot of times it's not worth the time, effort, and money we put into optimizing a page for a specific keyword because we likely won't be able to crack the top 3 or #1 spot and the increase in conversions doesn't outweigh the input.
But, making decisions like that is why SEOs are paid (among a lot of other reasons). Just because we decide a certain keyword isn't worth optimizing for doesn't mean the CTR still isn't better in the #1 spot for that keyword. It comes down to compiling data and using it to make informed decisions. If you don't want to use this particular data or think it unwise for others to do so, you've made that perfectly clear. I happen to disagree.
Ok but what about CTR related to number of words? I saw once that longer keywords give better CTR compared to one-two keywords. Could You do that comparison too?
This has already been addressed in the study. Check out the free tool that we have built which allows you to segment the data by the number of words that a query contains. Let me know what you think.
Saw it, thanks:) Just like I thought:)
This is a way to use Google Webmaster tools wisely. I want to know how the new study differs from the old one?
Good work !
I have several keywords which rank organically either #1 or #2 according google's webmaster tools. However, my CTR is 1%! FYI, the search term is very relevant to my services. Even if my meta description which was horrible (which it is not), it should get closer to 50%, right? Can any of you knowledgeable people advise?
We are facing a strange challenge from last few days.
One of our website used to get 12% CTR as an average (from GWT). Suddenly we started getting Impressions spikes (in the range of 60-120%) but there was no change in clicks.
We started investigating rank increase, keyword inclusions etc but everything was normal. During one of hit & trial methods I filtered parent country in GWT (in our case it is India) and found everything was normal for India(no spikes at all) but spikes start appearing if I select all countries. So we were getting 150000+ impressions on Google Search from some unknown country and nobody was clicking.
Can anybody help me in understanding this?
Great study. I have also found that CTR on mobile is lower for number one results for some reason! It would be interesting to delve deeper into why.
Also - many of the top websites that appear in the top searches - take for example in the law industry - are not very relevant or are advertising websites. For example, if someone searches 'Lawyers' in a given city, they are likely to find findlaw.com and thumbtack.com and yellowpages.com and so on, in which they are not directly seeing the websites of the lawyers themselves. I would like to see a study that examines these types of sites and their impact on organic searches for actual business websites. Not necessarily in the legal industry, but across the board these websites are popping up - are people valuing them the same as an actual website or more so? How do click through rates of these advertisement/directory websites compare to actual websites of businesses? Why does Google seem to value them more than an actual lawyer website for certain searches?
Philip, do you have a new study in the works? I'd like to see the trend from 2014 through the end of 2016 for mobile CTRs.
Hey Douglas, we have a tool that we update with recent data. Check it out here:
https://www.advancedwebranking.com/cloud/ctrstudy/
Hope this helps.
i want to see how local listings effect ctr especially now that Google has changed the first page real estate to 3 local restults.
I just wanted to add that I came across this article after doing something very similar for some of my clients. However, one thing I noticed that was not mentioned (and so I thought I would add) is that I actually tend to see a slight bump in the CTR at position 10. Although this changes, I was thinking that this could be from people scrolling down to the bottom of the results and looking in reverse order.
SEO + PPC = CTR Improvement. Especially on mobile!
i am very impress about this blog.
Your blog provides the complete Process to get good traffic for our websites.
How to do each and every steps about calculate Avg. traffic form stat to end.
Thanks
Hi Philip,
I, for one, appreciate the time and effort it took to realize this study and post it for all to read. I also feel some of the comments are excessively critical -- typical of pure statisticians who are weary of any analysis results.
That said, it was time for another study considering the outdatedness of the previous ones. I find your methodology acceptable for its purpose. Previous such studies all have their own flaws, yet when more than half a dozen studies show similar results, that's a valuable take-away. I think it would be very fair to say that "overall" the #1 result accounts for a quarter to a third of all CTRs, the TOP 5 results for about one half to two thirds of all CTRs and the Top 10 for two thirds to three quarters of all click-throughs. I don't think anyone on this thread would argue that ranking on the first page will drive more traffic than a lesser ranking.
One of your posts states that we can use this data to justify a client investment in raising a ranking from, say, a number #3 position to a #1 position. The thought is interesting and has crossed my mind many times. However, it would be very difficult to justify considering all the variables and assumptions: we assume GWT tools is correct, we assume Google's keyword tool gives us accurate monthly keyword estimates, we make assumptions about "not provided" results, further assumptions about the average CTR depending on the type of site or industry, assumptions about the client's conversion rate (would it be safe to assume the same conversion for a site that sells The Clapper as for one that sells luxury yachts? Or for that matter the conversion on these same two sites during the month of July versus December?). And it doesn't stop there. So in this case, I do agree with GIGO.
Finally, I work in Spain where the general understanding of any aspect of online marketing is very limited (compared to the U.S.). Few clients understand that increased traffic by itself means absolutely nothing -- in some instances it only results in a higher bounce rate. They simply want to drive LOTS of traffic to a website often with a vague offering, long signup forms and a limited payment gateway that fails one third of the time!
Thanks again for posting the study.
Henrik
Its really something interesting, Learned lot of things from this data. Thank You Philip.
Thanks for this study Philip! It could be interesting to know the magnitude of the new sitelinks search box of Google directly in SERPs and the impact of this in the CTR. What do you think?
Thanks!
Hi Rubén,
These are usually associated with a brand search and they appear after the 6 site links that Google shows after the first organic result for these types of queries.
I believe it's safe to assume that people would either click on the first result, one of the 6 site links or search within that site, so the CTR would be the same as for a branded keyword in this case.
Unfortunately there is no way to measure this right now. Not with the data that Google makes available.
What really surprises me is that there is not a significant drip off between CTR between the send and thirs pages. Also really surprised that in desktop searches CTR for position 6-10. I can not for the life of me.figure out what would be the cause of that. I mean when you go to page 3 you are going really deep into search results.
I like the questions, and it is something you need process before starting on an SEO campaign.
How much traffic would I actually get if I rank on the first page? Is it worth my time trying to rank above the fold? << This is where a competitive analysis is super important.
How much more traffic will I get if I rank first in the organic results?
How will that traffic convert?
The value of the conversion to the company?
Obviously these questions are important too.
You can discuss this organic CTR topic seven ways till Sunday and you will NEVER be able to conclude 100% that it is X. It is fun to try and calculate it but there are so many variables. Like many things your conclusion rests on the quality of the data and the conclusions you draw from the data. And sure if there was a 100% transparent source you could draw some accurate conclusions. Will a search engine provide these results? Doubtful
Changing the subject a little I remember years ago trying to use the estimated traffic tool in Adwords. Google supplied that data. Look how inaccurate that tool was and still is despite the "changes" made to it to become more accurate. I would rank #1 for a keyword or phrase and get about 35% of the estimated traffic.
Bottomline: the organic CTR will decline steadily over the years because Google is always looking for a way to increase their ADwords revenue. Look at the changes of ranking Algorithms and the changes in organic search results. Look at PLA's and tell me that they did not take away from organic clicks. The answerbox is another one. If there is a way to increase Adwords revenue they will do it and at the expense of free organic traffic.
In the 90's man could survive on SEO alone. I did not mind putting all my eggs in one basket, because it was easy to own the "basket" err Google. However in 2014 and beyond, Marketing Diversification and knowing how each marketing method can be used to support/promote each other is the key to success. IMO
"First Page or Die Trying" Keep up the good work!
Philip, Thanks for the article. I also read the Advanced Web ranking, Google CTR study, you linked in. I am surprised actual Ad copy does not get a mention in relation to CTR. It is the easiest uplift on CTR I get for clients. Did you come across any statistics on CTR relating to good copy on Ads as distinct from bad copy? ie we have a client ranked at average position 7 on a globally highly competitive keyword and we have tripled their CTR simply by making the Ad more enticing. They average above 6% now. Is there any studies you have covered in this area?
Hi Andrew,
This study has been focused only on organic results as the data is coming from GWT. However, there is a section in the study that tackles how ads affect the CTR of organic results. You might want to read more about this in the study PDF or in this post from Search Engine Land:
https://searchengineland.com/different-types-ads-influence-organic-ctr-google-204676
Great article! Once, just days after the new Panda developement, I wrote an article based on my own penalization.. It was called "All about CTR" .. It's not about what's in the article, but it may interest you. It's in Italian, buy Google may translate it for you :)
Amazing, It will help's search people to with SEO in better manner.
Dear Philip Petrescu,
In my competition field in Germany I have almost 3 Ads slots at Top above the organic search results and also at right side many times 8 Ad slots like for example https://www.google.de/?gws_rd=ssl#q=ferienhaus+d%C3%A4nemark
You wrote "Let's see an example for an unbranded keyword with a volume of 1,000 searches per month where you rank first in the organic results with no ads above you:
1,000 x 24.8 / 100 = 248 (visits per month)"
How much you suggest to discount the 24.8 CTR for position 1 or in general an approximate factor for the diagram you put under Search intent for unbranded keywords?
Thank you very much, I will download your study and read it later.
Greetings from Hamburg,
Hans
This is a great Case Study. Philip. I am sorry to see so many negative comments from M.M. Nothing in SEO industry is 100% accurate except you get ranking increase with strong authoritative links from reputable website.I will implement your study in future SEO experiments.
The flaw is that the studies are all "post-click", i.e. they measure the following:
Given that someone clicked on an organic result, which position got that click?
The problem is, people don't click on organic results 100% of the time when they do a search, primarily due to two behaviors: clicking on paid search results instead, and "abandoned searches" - where they do a second search without clicking on the first set of results at all.
This is why people are always frustrated applying these types of curves to real-world data, because the estimates come out way too high.
I covered the whole thing in this article and included a spreadsheet for estimating organic traffic, which I calibrated against real-world results - it's worked pretty well for me, although I'm sure it could use some updating.
https://www.coconutheadphones.com/estimating-organic-search-opportunity-part-2-of-2/
The lower numbers also roughly correspond with the paid search curves Richard Stokes shows in his "Ultimate Guide to Pay Per Click Advertising" - which makes sense since Paid Search and Organic are basically the same thing - ordered lists of headings and descriptions.
This study only addresses the organic CTR, but you're right, there are features in the SERPs such as ads, answer boxes, carousels, etc. that do affect the CTR a lot. If you download the PDF with the study, you will find there more information about how ads affect the CTR of organic results and an interesting CTR curve for ads too provided by a research study conducted by Wordstream.
Did you miss this part?
In case you wonder where the other 23.08% of the clicks are, here are some possible scenarios:
In fact - yes! ;-). Thx Andy. Will definitely check out the full PDF too, thanks Philip!
I am satisfied of this graph..................