How Organic Search CTR is Impacted by Google’s Search Features
This study presents never-before-seen information on how Google's search features impact click-through rate (CTR) behavior in the search results. To our knowledge, this will be the first such study that marries actual rankings data with real clickstream data.
Key findings in this report include (but are not limited to):
- As others have reported, no-click queries are a significant part of all searches, but we found that these appear to be much less significant on commercial queries (SERPs that show ads). Clicks go to ads on 28% of all searches with ads, with around 23% of those coming at the expense of ”no-clicks” (clicks that don't result in a click to a website) and the remaining 5% coming at the expense of organic search. As a result, the impact of ads on organic search clicks appears to be fairly small.
- Traditional CTR models that have been published are largely meaningless because they don't separate CTR behavior for branded queries from non-branded queries. For example, CTR for position 1 on branded queries was 69% in our data and only 19% for non-branded queries.
- CTR for queries with featured snippets are slightly higher than they are for queries without, but the difference is small enough as not to be significant. Our conclusion is that there are probably many types of featured snippets that give users that answer they want and CTR drops, but that these are offset by other featured snippets that drive much higher CTR.
We will show a number of detailed scenarios below for many different types of search features and their impact on CTR.
Our partner in this project was AuthorityLabs Data Services. They identified 2M queries to track the spread across five different markets (auto, beauty, finance, retail, and travel) with roughly 400K keywords per market. AuthorityLabs followed the ranking data and provided us the clickstream data for these 2M keywords daily for thirty days.
Perficient Digital then performed a detailed analysis to pull together this study. If you've heard enough of the overview and want to get right to the data, jump ahead to the section titled "The Basic Data."
Also, in the "Detailed Analysis" section of this study, we'll discuss how you can prosper in the brave new world of search.
If you're seeking to understand how Google's SERP features and encroachment into providing content and answers in the results impacts clicks, this report is a must-read.Rand Fishkin, Founder of SparkToro, and previously co-founder of Moz and Inbound.org
AuthorityLabs tracked approximately 2M queries split into five categories – auto, beauty, finance, retail, and travel – with roughly 400K queries each. Rankings for each of these queries were tracked daily for thirty days. The data included traditional ranking data information, as well as a detailed data set showing what search features were present in the Search Engine Results Pages (SERPs) in each daily snapshot.
The tracked search features included:
- Featured Snippets
- Image Carousels
- Job Packs
- Knowledge Graph Results
- Local Packs
- News Carousels
- People Also Ask Boxes
- Related Searches
- Shopping Results
- Video Carousels
AuthorityLabs also pulled clickstream data for each of these queries for the same thirty-day period. AuthorityLabs then provided this data (all 3 terabytes of it!) to the team at Perficient Digital, and we conducted a detailed analysis.
Through our analysis, we were able to amass specific data on the impact of search features on CTR within the search results. We were also able to slice the data by each of the five markets, which allows us to show how search behavior differs across these markets.
Our analysis enabled us to amass specific data on the impact search features have on search results.
The Basic Data
In aggregate, we looked at the outcome of nearly 250M searches. This resulted in 134M organic clicks, 11.3M paid clicks, and 102M no-clicks. Thirty-seven percent of the searches occurred on mobile devices and 63% on desktop devices. Please note that this 37/63% split is not indicative of the market-wide split between desktop and mobile, but instead only indicates the distribution resulting from the data collection methodology used. See detailed data on mobile vs. desktop usage here.
Digging a bit deeper, we can see the distribution of the organic clicks, paid clicks, and no-clicks for desktop devices:
These are the results for mobile devices:
Interestingly, the CTR for paid search is nearly the same for both desktop and mobile, with an average of 4.58%. No-clicks on desktop accounted for about one-third of the total, and 54.5% for mobile. Also of interest is the ratio of organic clicks vs. paid clicks in both environments.
Given that there are 13.4 times as many clicks to organic results as paid results on the desktop and 9.1 times as many on mobile, it's clear that organic search continues to matter significantly. In light of this data, the ongoing disparity between what brands invest in paid search vs. organic search is stunning. There's no doubt that the greater certainty that paid search brings is a factor here, but there is substantially more value in investments in organic search.
There's no doubt that the greater certainty that paid search brings is a factor here, but there is substantially more value to be had from investments in organic search.
How Big a Problem Are No-Clicks?
There has been much industry discussion on the impact of the rise of no-click searches (clicks that don't result in a click to a website). This is indeed an area of concern, but it's interesting to note that the volume of no-click searches is not nearly as high on commercial queries. Here's a look at the CTR for all the queries that contained ads:
Now compare that to the CTR for all queries that do not contain ads:
What we see here is that the vast majority of the clicks that go to paid results come at the expense of no-click results. This means that the volume of no-clicks on commercial queries (those with ads) is relatively small, roughly only 14%.
Note: As alluded to by the title for the images, we used the presence of ads as a proxy for identifying commercial queries. While not a perfect measure, it's an effective method of identifying that class of queries.
The vast majority of the clicks that go to paid results come at the expense of no-click results.
What Is the Real Story on CTR by SERP Position?
Capturing data on the click-through rate by ranking position proved to be an interesting exercise, as it showed us how different CTR is between branded and non-branded queries:
As you can see, the CTR for branded queries is far higher than it is for non-branded queries, and the skew toward position 1 is far greater as well. Based on this, we believe that classical viewpoints on how CTR by SERP position works need to be updated. Past studies have not separated brand queries from non-branded queries, and these have significantly been skewed as a result. CTR for the first position for non-branded queries is actually under 20% (19.23%).
Note: Branded queries were identified here as all queries that have sitelinks associated with them.
How Search Features Influence Search Behavior
One of the more remarkable aspects of the data we're sharing here is a detailed look at how search features impact user CTR behavior. Here's a look at the popularity of search features in our data:
Please note that this distribution of search features is not necessarily representative of the search landscape in general. The high density of results with sitelinks (47%) suggests that the queries in our dataset skew toward being navigational. Nonetheless, the large percentage of knowledge graph results, “People Also Ask” boxes, and video carousels are quite interesting.
Because of the high density of results with sitelinks, we were careful to perform our analyses by separating brand versus non-brand behavior, where appropriate. In aggregate, here is a look at the total CTR of the top 10 results:
As you can see, the aggregate CTR for brand queries is quite high, whereas the aggregate CTR for non-branded results is under 38%. Next in our review is the data on how Featured Snippets impact aggregate organic CTR behavior for non-branded queries:
As you can see, there is a marginal increase in overall CTR. For clarity, since the URL for featured snippets also shows up in the top 10 search results, clicks to links within the featured snippets are counted here. Further, this data set shows the results for all results with featured snippets, as compared to all results without featured snippets.
For comparison purposes, we also looked at all queries that showed featured snippets for some, but not all, of the thirty-day period. For these queries, we compared the CTR for those queries on the days when they had featured snippets with the data for which they did not have featured snippets. For these queries, there was virtually no difference in the CTR.
Together, these data points tell us that, in aggregate, featured snippets don't have much impact on CTR. However, this is only a measurement of aggregate behavior. There are many types of featured snippets that provide the complete answer that many users desire, resulting in a significant drop in CTR.
Further, other queries have the opposite effect, with the impact of driving substantially higher CTR to the sites from which the featured snippet was derived. What this means is that Google likely has multiple ways of measuring the impact of featured snippets. In some cases, lower CTR is good, and in other instances, a higher CTR is an indicator of a better result.
What about People Also Ask boxes? Let's take a look:
These seem to lower aggregate CTR. However, unlike featured snippets, these URLs are not in the clickstream data, so it makes sense that the overall CTR is somewhat lower. This tells us that there are definitely a material number of clicks going to the People Also Ask boxes themselves, and an unknown number of clicks going to the web sites listed in them.
The People Also Ask results based on the ranking position are interesting:
Here you see that positions 1 and 2 have notably lower click-through rates than the average of 35.74%, but the remaining positions, 3 through 10, are all higher. Please note that placements of People Ask boxes in position 1 are comparatively rare, and we had only one case of this in our data set.
For the Knowledge Graph results on non-branded queries, we also see that they lower the organic click-through rate for non-branded queries:
Next, let's look at the impact of image carousels on organic CTR behavior:
This data suggests that image carousels increase overall organic CTR on non-branded queries by more than 12%. Note that URLs for image carousel links are tracked in our clickstream data, so this indicates that they are effective in increasing overall CTR. This is even more impressive is when you look at this based on the placement of the image carousel in the search results:
We see a slightly lower CTR when the image carousel is in positions 1 through 3, but a materially higher CTR in positions 4 through 6. It's almost as if the presence of the image carousel increases the CTR to the higher position results because it acts as a divider to the search results page.
Continuing this trend, we see that news carousels also tend to cause a decline in the overall organic click-through results:
Related searches appear to have a dramatic impact on organic CTR as well:
In this case, there seems to be a different mechanism underway. It's likely Google doesn't show related searches where click-through rates are already quite high.
Video carousels also appear to result in a drop in overall organic CTR:
When we look at the data from the point of view of the ranking position of the video carousel, we see that this appears to matter quite a bit in the overall CTR:
Non-Branded CTR by Ranking Position by Market
Let's look at the data split out by market category for non-branded queries:
The auto category had the highest organic CTR, followed by the beauty category. The lowest overall CTR was for retail, with travel showing only a slightly higher CTR.
Image carousels increase overall organic CTR on non-branded queries by more than 12%.
Why is Google Implementing These Features?
Put simply, Google is a corporate entity motivated by profit. However, with the exception of Google Ads and Google Shopping, the search features reviewed in this study don't directly drive revenue. This means that there are two major ways for Google to increase its revenue:
- 1. Grow market share
- 2. Increase usage
Search features such as those covered in this study help them do both of these things. For example, featured snippets and knowledge graph results are about providing instant answers, and in some cases, these will keep users on the Google platform. People Also Ask boxes provide a mechanism to get rapid answers to closely related questions.
Image carousels and video carousels help users who are looking for visual responses. This makes sense. Per Google's Aparna Chennapragada, 30% of the neurons in the cortex of our brain are for vision. Local Packs and Job Packs clearly help users get information related to those types of needs.
As long as these features bring value to users and help them get what they want as quickly as possible, they will help Google grow its market share and increase overall usage. This ultimately leads to more ad revenue. For that reason, you can expect that they will continue to test and optimize these types of features on an ongoing basis.
There are other types of search features that weren't covered in this study. Some of these include their weather, mortgage, and color picker widgets. These are features that are inferior to those provided by third-party providers, but if they offer enough functionality for users, they may settle for that and stay on the Google platform.
Expanding on these types of features, which encroach on opportunities for third-party publishers, may also represent a core part of the ongoing opportunity for growth for Google. Hopefully that is not the direction Google chooses because an ecosystem that brings benefits to all of its participants is best for all, including Google, in the long run.
Google constantly changes their SERP features as a means to grow market share and increase usage.
Detailed Analysis and Recommendations
The split between branded and non-branded queries was a fascinating aspect of this study.
Aggregate CTR to the top 10 results for branded queries was quite high at 71.36%, but for non-branded queries, it was only 37.88%. Please note that this does not include most clicks to local search results, knowledge graph results, or People Also Ask boxes, as these were not trackable by our clickstream data.
This shows that as more and more search features are implemented, the chance to obtain organic search traffic from Google may well be declining. So, what can you do to continue to maximize your opportunity with Google? Focus your efforts on ranking within Google's search features.
Those opportunities include:
- Featured Snippets: Learn how to modify your content to earn featured snippets. While you're at it, teach your content creators how to write content that's more likely to earn these snippets. The more systematic you are with this process, the better.
- People Also Ask Boxes: A secondary opportunity for your content is to get into People Also Ask boxes. The good news here is that the process for modifying your content is the same used to earn featured snippets. In fact, if you show up in the People Also Ask boxes, but not for any featured snippets, it's a sign that your pursuit of the featured snippet is on the right track, but not quite there yet.
- Image Carousels: Google is making moves toward becoming more visual. In this year's version of our annual study of Google's rich answers, we saw their use of image carousels more than double. Of course, basic image optimization practices apply, but it's also important to post high-quality, original, and relevant images high up on the page. Also, implement image site maps for those images.
- Job Packs: If you're trying to get job listings to show up in Google's results, implement Job Postings Schema, which makes you eligible to show up in those results. You can read more about the entire process here.
- Local Packs: If you have physical locations, optimizing for local search is a must. The first key is to get your name, address, and phone number ("NAP") information to properly show across the entire web. You can use services such as Yext or Moz Local to do this. Learn more about other aspects of local SEO optimization here.
- News Carousels: You're only eligible for this if you're in Google News. If you are, you can show up in these results. To appear in the news carousel on Google's mobile search results, you need to implement AMP. In fact, in the mobile results, AMP is a requirement. You can read more about implementing AMP here.
- Shopping Results: While this is an advertising-based experience, it does occupy prominent space in the search results. You can start the process of learning how to rank well for these here.
- Video Carousels: In the data we used for this study, video carousels showed up in more than two-thirds of all results. That spells lots of opportunity. Of course, YouTube is a sizable search engine on its own, but if you're looking to get your video into Google, it uses a different ranking algorithm.
Ranking in search features, as outlined above, is a significant area of opportunity, but there are additional steps you can take. Increase your focus on producing content that addresses the long tail of search. The long tail of search still represents 70% or more of all searches, and Google will be very limited in how they can provide instant answers to users for these types of queries.
One business that is doing this extremely well is REI, which is known for the outdoor-focused content it publishes. Hundreds of new pieces of content are published each year, and they're driving strong traffic results:
Driving organic search growth is not the only reason to publish informational content. Data from a Conductor study suggests that consumers who read informational content are much more likely to buy from you.
This idea of publishing a significant amount of educational content is critical. It remains one of the areas in search where the intrusion from Google remains relatively light. Given the scope of the challenges created by Google's increasing use of search features, focusing on areas where Google won't pay much attention remains one of your top opportunities.
The pursuit of this project is a result of the great work Rand Fishkin of SparkToro has done.