Search engine results pages (SERPs) have changed significantly over the years, and there are lots of new features. Now you have knowledge graphs and carousels. You have answer boxes and local listings in the right rail. Do you know how users behave when rich snippets appear for branded terms and when they do not? We figured that out. What about with mobile? We analyzed the data on that as well.
We gathered data from seven clients, over 248 keywords, 1,339,559 organic impressions, and 168,906 clicks via Google’s Search Console. We also broke this information down between a SERP listing featuring a rich snippet vs. no snippet, branded vs. non-branded, and desktop vs. mobile. (If you are wondering what rich snippets are, why they are important, or how to implement them on your site, we published a recent SERP Optimization article outlining this information in great detail) Our goal was to answer the following question: “Have these features changed the way users interact with SERPs?” We predicted, yes, they have.
Increasing mobile usage has led to speculation that it has trained users to scan more vertically now, as opposed to traditional ‘F’ patterns. The Thrive Agency found that the amount of time it took participants to find a desirable result decreased from 14-15 seconds in 2005 to 8-9 seconds in 2014. Users now have a better sense as to what they are looking for. If they know what they are searching for, does that mean they are more or less inclined to click on a search result with relevant information?
The results of our analysis were quite interesting. Not only did normal desktop SERP listings outperform listings enhanced with snippets at a CTR of 12.9% to 2.2%, but enhanced snippet listings were 23% higher in positioning. That’s pretty huge. If all things were equal, that would be the CTR equivalent of one keyword being in position 2 and the other in position 7. How could this be? Well, I have two theories on this.
One is that I’m wrong about assuming rich snippet enhanced SERP listings would increase CTRs, which could most likely be the case. Conventional thinking follows that rich snippets undoubtedly increase CTRs. Back in 2011, Search Engine Land found that structured data and rich snippets led to a 30% increase in CTR. But that was also 2011; I’m not confident that stat holds up anymore. Their research did lay the foundation for the idea that to increase organic CTRs; rich snippets had to be a part of the recipe.
I revisited the data and scrubbed it of all brand queries. That changed the story. CTRs of non-rich snippet queries dropped from 12.9% to 1.7%, while rich snippet CTRs remained exactly the same and only enjoyed a 7.9% increase in positioning compared against non-rich snippet queries. Now I’m wondering if snippets that appear for brand searches actually do anything at all; are users already inclined to click on your brand/website regardless of how snippety your SERP listing looks?
The second option is this: if a snippet features the information for which a user is searching, then that may eliminate their need to click through to the page. As my colleague Jason stated in his SERP optimization post, 40% of Google searches result in no click at all. If I were a betting man, I’d say some of the keywords I analyzed were a part of that 40%. It’s somewhat of a bittersweet investment as it’s an additive feature: regardless of whether or not you are generating more clicks, you are adding value for the searcher. The issue with this is that you are unable to track those who completed their “goal” without actually clicking to your site.
Oh, I guess there’s a third option too. Since rich snippets are created to enable searchers to make a more educated click decision and improve the likelihood that searchers will choose a page that satisfies their need, maybe my pages aren’t what they are looking for or satisfy their needs . . . blasphemy. What if a user was searching for a reclining sofa and we featured a price snippet. Maybe that price is too high for the user, and they choose to look elsewhere. Would you really want that user clicking through to the site just to bounce anyway?
But What About Mobile?
In our own study we found that in the mobile SERPs, non-branded queries with snippets delivered a CTR of 10.1%, while non-rich snippet queries produced a CTR of 10.3%. There was a difference in average position: rich snippet queries averaged position 5.1, while non-rich snippet queries averaged position 7.4, which may skew the CTR results. We took it a step further and also analyzed scroll depth patterns on mobile sites vs. desktop sites. We found there was no statistically significant difference between scrolling patterns on mobile versus desktop devices; in fact, behavior was almost identical.
I’m not arguing against the use of rich snippets or their benefits; they are extremely useful and have strong benefits for websites. What I did want to do was see for myself if there were obvious benefits for our clients in the use of rich snippets. I had questions about the benefits of rich snippets, and I wanted to answer those questions with data. I found that when branded searches occur, user intent is to view that brand’s website, so rich snippets are not needed or influential. However, when non-branded queries occur, rich snippets do help increase CTRs.
Some questions I was left pondering. I’d love to hear your thoughts about these in the comments or on Twitter:
- What are CTRs if you show up in the knowledge graph?
- What type of uplift do you see when you do appear in the knowledge graph?
- How can we start to measure the benefits of showing up in SERPs when a listing is not clicked?
- Mediative. (2014). The Evolution of Google Search Results Pages and Their Effect on User Behaviour New