Let’s talk about bounce rates

Useful for: Not much by themselves!

Be careful of: The fact that there are both good bounces and bad bounces

Our preferred tool: Google Analytics

Bounce rates have long been thought of as the metric for understanding whether a landing page is performing well. We’re just going to come out and say it though – this is a false assumption.

Indeed, there can such a thing as a ‘good bounce’, where a user visits a webpage, satisfies their search query, and carries on with their life. Think of a time when you entered a query into Google such as ‘when does the next series of Game of Thrones start?’. You likely either satisfied your query with a Knowledge Panel result or came across a webpage with the answer in bright lights. You didn’t necessarily visit any other pages on that website, but the website did its job regardless.

This then leads to another question; how can you tell whether your landing pages are satisfying the users’ search intent? There are a few ways you can do this, including adapting your GA tracking code to fire off a non-interaction event every few seconds, thus allowing you to get an accurate picture of how long people spent on your landing pages. Do not rely on average session data alone however, as out-of-the-box this is only calculated by using a sample of visitors who visited more than one page on your website. It will not take into account users who bounced.

You could also use a third party tool such as Hotjar or Mixpanel to see how long spent on your landing pages, as well as working out the specific interactions they took (e.g. scrolls, clicks, etc).

Finally, Hotjar also has the option of triggering a pop-up survey when a user attempts to navigate away from your site. You could use this to ask the question of whether that page has been of assistance, and indeed whether there’s anything you could do improve.

So all of this makes sense for ‘good bounces’, but what about bad bounces? Let’s take an e-commerce store as an example.

Imagine this e-commerce store undergoes a website redesign and their bounce rates increase. On the face of it this is a poor result, and indeed this may be the case. However, perhaps these new bounces were from users who realised they were on the wrong site much earlier than they would have done previously. In this case looking at bounce rates alone would provide a blinkered view on overall performance, and other metrics (e.g. conversion rates, average time on site, etc) should also be taken into consideration.

In summary, before you report on bounce rates as a KPI, be sure to think about what type of site you’re running and whether it might be benefiting from good bounces. If this is the case then be sure to include this caveat in the reporting to your stakeholders.