Do Search Engines Use Bounce Rate As A Ranking Factor?

Your web site’s bounce rate may be a significant factor in your search engine rankings. If the bounce rate on your site is high, you could end up with lower rankings in the search engines. Correspondingly, lower bounce rates may actually offer meaningful ranking boosts. (Don’t know what a bounce rate is? Hang on—definitions below.)

Don’t believe that bounce rate is a serious ranking factor? You should. A new study by SEO Black Hat shows some significant impact in a web site’s rankings as a result of significant changes in bounce rate. It is, of course, possible that the data in the study have been affected by other factors taking place at the same time, so this one study is not be any means conclusive.

I speculated on bounce rate as a ranking factor back in August of 2007. The underlying fact is that search engines want quality sites in their search results. High bounce rates may be a very good indicator of a poor site experience, or worse still, a complete mismatch between the content of the site, and the search query entered by the user. This provides heavy motivation to look at bounce rates as a meaningful SEO factor, and how to minimize bounce rates in the hope of increasing your rankings.

Bounce rate defined

What’s bounce rate? Google Analytics defines a bounce as any visit where the visitor views only one page on the site, and then does something else. What happens next? There are several possibilities, including the user clicking on a link to a page on a different web site, closing an open window or tab, typing a new URL, clicking the “back” button to leave the site, or perhaps the user doesn’t do anything and a session timeout occurs. This is still a bit fuzzy because of the nature of how “sessions” are defined in analytics packages. Analytics software that relies on Javascript tags only know when someone loads a page of your web site (so the Javascript runs).

As a result, these analytics packages have difficulty determining what happens in the meantime. In our first scenario, user A comes to your site, views one page, and then goes to a three-martini lunch (well, hopefully not). They then come back and visit 10 other pages on your site. Because of the way that analytics packages work, this will be seen as two different visits, and the first one will be recorded as a bounce (a single page visit).

For search engines there are other possibilities. For example, a bounce could be defined as user A entering a search query, going to your web site, returning to the search engine, and clicking on another result. Another possible definition involves user A entering a search query, going to your site, and returning to the search results page in less than “x” seconds. So there are a few possibilities of these kinds for the search engines to experiment with.

Of course, the major search engines have other data available as well. For example, they have toolbars which can be a rich treasure trove of data for tracking user actions. In addition, search engines license data from major ISPs and collect additional data to track where a user goes. The possibilities go well beyond what an analytics package can do.

There are some issues, of course. What happens if the user is looking for a single bit of information, such as Abe Lincoln’s birthplace. If a search result leads to a good reference site, the user gets the answer and are done. They may still click back to the search results and search on something else, or click on another result from the original search. Even though this is a satisfactory outcome, it’s recorded as a bounce by most analytics packages.

This type of scenario will be prevalent with users who are searching for a simple answer and get their responses from almanac-like information sites. The key here is to also try and factor comparative data in the way that bounce rate is used for influencing rankings. For example, the bounce rate of almanac sites may be higher than the bounce rate of an e-commerce site, which will likely be higher than the bounce rate of a directory site (where users have a high probability of going on to another site).

My guess is that search engines look (or will look) to determine how a site’s bounce rate compares to other sites that are comparable, or other sites that it in considering returning for a user’s search query. In this latter scenario, you could imagine a run-time adjustment where the search engine comes back with the “traditional” results to a user’s query, and then make a bounce rate adjustment.

The bottom line

There is a real possibility that bounce rate is a significant ranking factor right now. Even if it isn’t, it is my opinion that it will be made a factor in the near future. Even if it is not a factor, and even if it does not become one, there are plenty of reasons to look closely a bounce rate anyway.

It speaks to the conversion potential of your site, and this already gives you plenty of reason to look at it. You can read more about this aspect of bounce rate in this excellent post by analytics guru Avinash Kaushik, titled Standard Metrics Revisited: #3: Bounce Rate.

Eric Enge is the president of Stone Temple Consulting, an SEO consultancy outside of Boston. Eric is also co-founder of Moving Traffic Inc., the publisher of Custom Search Guide. The Industrial Strength column appears weekly at Search Engine Land.