For as long as SEO has been around, search optimizers have debated how much “user experience” factors into a search rank. According to some data, qualitative factors like how long a user spends on a page can influence how that page ranks—but you could also make for a case of correlation influencing this relationship, rather than causation. On the other hand, you have classic “standbys” as ranking influencers, such as inbound link quality, with all other measurable ranking factors being secondary, correlational, or purely coincidental.
Now, thanks to some insights from Google engineer Paul Haahr, we may have a clue as to whether one of the most hotly debated topics in the user experience debate (click-through rates) is just a myth, or if it truly does influence how your site ranks in Google.
The concept behind CTR influence is pretty simple, and it’s likely the reason so many search optimizers have found it easy to believe that it’s a verifiable ranking signal.
Google has an anticipated spread of CTRs for its various search results ranks. For example, let’s say it expects 1,000 click for the top query, 200 for the second, and 100 for the third. Now, let’s say after a while, the three sites in these positions offer a major discrepancy; the first site is only getting 400 clicks, the second site gets its expected 200, and the third gets 700. That’s an anomaly, and Google might come to the conclusion that this third entry is way more relevant than the other two. Accordingly, it may boost its rank.
(Image Source: SearchEngineLand)
Unfortunately, there’s no easy way to account for this pattern in a controlled experiment, keeping all other ranking factors consistent.
There have been some interesting and, admittedly, persuasive studies in the past that have seemingly disproven, or at least suggested evidence to the contrary of the idea that organic CTRs influence search rank potential. One in particular used a “click bot” to automatically click on certain results for a small range of keywords as a controlled experiment to see if additional clicks from searches alone were enough to move the rankings of a particular entry. The results, as you might imagine, were nonexistent. There was no upward momentum whatsoever.
(Image Source: SearchEngineLand)
However, as others have pointed out, there’s a serious flaw in this study: it used click bots. Google is no stranger to the use of bots in manipulation of search ranks (and other online advantages), and it has precautions in place to guard against these negative techniques. While the experiment is interesting, it doesn’t offer conclusive proof that organic CTR isn’t a ranking signal.
A recent experiment done by Wordstream (and published by Moz) illsustrates a very interesting relationship between CTRs and search, and it goes a step further by drawing in possible effects from RankBrain, which helps Google decipher and understand semantically complex user queries.
Here’s the basic rundown of the experiment. Wordstream examined the relationship between CTRs for given search queries and how they relate to a given search position. The key here is that basic keyword “head terms” are plotted separately from long-tail keywords, which is a major focus of RankBrain.
(Image Source: Moz)
As you can see, long-tail keywords tend to carry a higher CTR, on average, than their basic counterparts. The same keyword niche was used to attempt to isolate variables that may have otherwise influenced the difference—so what could account for this?
You could make the argument that the big difference here is the fact that long-tail keywords have a higher likelihood of premeditated user intent, which in turn could influence higher CTRs in general. However, note that in high-position ranks, long-tail terms greatly outperform basic keyword phrases, while in lower organic ranks (10 and lower), the difference is almost negligible.
Keep that in mind when looking at this graph of similar keyword terms in paid results:
(Image Source: Moz)
The same pattern is not visible here. In the top ranks, the differences between shorter and longer keyword phrases is much tighter together, following a much more linear path as the ranks get lower.
What’s the key takeaway from this study? There’s something interesting going on with CTRs and specifically organic search ranks. There’s just one thing stopping us from certifying this as evidence that CTRs positively influence search rank.
The big problem is that CTRs and search ranks are co-dependent variables. Assuming that CTR does influence search rank, the two become mutually inseparable. Did a search rank increase because it got a higher CTR, or did its CTR grow higher because it got a higher search rank? It’s almost impossible to isolate the factors here.
As there’s no direct proof of causation between CTRs and organic search ranks, and because even if there was, there are dozens of factors that are more important (including site structure, content, and external links), this shouldn’t affect your strategy too much. Click-through rates are still a good thing, and you should still aim to optimize for them with compelling title tags and accurate meta descriptions, but they may not directly affect your search ranks. Until we have more information, keep user experience optimization as a strategy separate from your SEO, and improve both for the best possible results for your site.