Don’t judge all bounce rates alike

We’ve told you about bouncers, those visitors who land on a website and leave without clicking through to other pages. We’ve explained how some websites may have a low bounce rate and some may have a high one, where more people bounce than click through; we’ve explained how, if a website’s bounce rate is on the high side, we consider it worth trying to find out why.

But high bounce rates aren’t always bad, and in some cases are to be expected. Take, for example, Wikipedia, which is an example of a website that should expect a high bounce rate. Web surfers may land on a Wikipedia page as a result of a specific search, find all the information they were seeking from that page, and exit the site.

But that’s not how things work for an e-commerce website. The ideal result here would be that a visitor who lands on a product page on an e-commerce site clicks the “Buy” button, moving ahead in the purchase process. Alternately, the customer could be drawn to visit additional pages and add more products to the shopping cart.

But neither of these favorable outcomes can occur if visitors don’t click through a website. That is why a “high” bounce rate is bad for e-commerce websites.

When we see a high bounce rate, we ask ourselves: Is the website unwelcoming visually? Is it difficult to navigate? When we address these issues, we’ll also monitor whether the changes are affecting the bounce rate. Rather than focusing on the bounce rate in relation to an arbitrary ideal – what the “experts” say it should be – we may find it more instructive to focus on improving the bounce rate in comparison to the historical bounce rates of the website itself.

It’s a case of interpreting the data, rather than just seeing the numbers. We’ll talk about some causes and cures in future blogs.