I’ve often cautioned against worrying unduly about “bounce rate” in your website analytics. Although it seems to be a Bad Thing on the surface, when you think about it, perhaps a high bounce rate is not always undesirable. If a “bounce” is just someone who comes to your site, takes a brief look and leaves, that might be simply a sign that your page did the job it was meant to do efficiently. So when people hear that the search engines take “bounce rate” into account when determining their rankings, it’s often a cue for some alarm and a little confusion. Should we design our sites so that visitors rarely get all the information they need at a glance? Should we force them to hang around unnecessarily, and in doing so “not bounce”? Would such unhelpfulness give us higher rankings?
The answer is to think about how bounces are measured, and to realise that what your analytics application identifies as a bounce is different to what a search engine can see. Your analytics application will identify that a visitor came to a page and didn’t interact with the site within a certain period (and therefore presumably left), recording that as a bounce. A search engine, however, doesn’t know what someone who’s clicked through to your site then goes on to do, internally. It can only assume that the visitor has bounced if that person comes back to the search engine within a certain time and clicks on another result. That’s a quite different thing.
Now, obviously some searches will result in people returning to the search engine much more frequently than others. If I’m looking for the best price for something, I’m likely to be clicking on a search engine result, seeing the price at the website, returning to the search engine and quickly repeating the exercise on the next few results. The search engines know all this. They have enormous quantities of data to work with. What matters is how your site performs in comparison to others for a certain search. Let’s say you’re number one for a search on “blue widgets”. The search engines see that on average, 90% of the people who click on your link come back and click on the number two result within a few seconds. They then see that of those people, 50% of them come back and click on the number three result. And of those, 50% come back and click on the number four result. It would appear that the sites at numbers two and three are retaining more visitors than you are. If an element of your site’s ranking is down to relevance, maybe you’ll be marked down in comparison to those other sites. You might be moved down one or two positions, where the search engine can then repeat the experiment the other way around. Maybe having a lot of these “poor relevance” page and search term combinations will actually affect your site’s overall strength in the search engine results. We don’t know, but it can’t be a good thing.
So, what’s the takeaway from all of this? Perhaps you need to identify the pages and search term combinations which are resulting in a high “bounce rate” (as identified by your analytics application), and see if you can do something about that. In the screenshot below, from Google Analytics on our website, I’ve looked at keywords which sent more than 10 people and ordered them in terms of bounce rate from highest to lowest. So we see that the search term “feedblitz vs mailchimp” brought in 21 visitors, but 100% of them “bounced”. I can see that the page they landed on didn’t answer their question at a glance, so clearly they weren’t bouncing because it did its job incredibly quickly and well – they were bouncing because they didn’t like the look of the page. So maybe we should build a page which does serve that query better. It’s not a search term I’d ever thought of focusing on, but there’s obviously traffic out there, it would be a relevant topic for this site, and Google has shown it’s prepared to send people our way. It would be an easy win …and there might be some similar ones out there for you.