This metric can be important to certain websites as it gives you an idea about the first impression your site is making. It tells you if your landing page meets the expectation of the visitors who landed there. If your bounce rate is high, you can suspect that something is wrong with either your landing page or your sources of traffic.
You can calculate the bounce rate with this formula:
A concrete example
- Let’s say that you have a website for dog owners. You are selling eco-friendly dog food and you have created a specific landing page with an offer for a specific promotion.
- You reach out with your promotion to your affiliate partners and on the next day you see 1000 new visits, but with 99% bounce rate. (That means every 99 out of 100 visitors left your website without browsing further.)
- You are really depressed. Then it turns out that the 1000 visitors were coming from a cat forum – where you were linked by one of your partners accidentally.
Obviously, your bounce rate is high because people on that forum expected cat-related things, but they got dog-related things instead when they came to your webpage. The source of the traffic was wrong.
- Going forward in time, another of your affiliate partners starts to promote your website. This time correctly: for dog owners. The ad says “Eco-friendly dog food! Special discount for dog lovers! This week only.”
- On the next day you see another 1000 new visitors, with 70% bounce rate. (That means every 7 out of 10 visitors bounced.)
- 70% bounce rate still seems to be high… So you start to investigate and it turns out that you only gave 20% discount – and at the same time one of your competitors gave 40% for the same product.
In this case, the bounce rate is high because the potential customers haven’t found your offer on your landing page good enough. To be more general: this time there was something wrong with the landing page.
To summarize this:
High bounce rate means you have an issue that’s worth investigating. How to investigate? User testing, heat mapping, further data analyses, common sense and many other ways.
But that leads us to the next big question:
What is a good/bad bounce rate?
To answer this question, we need to demystify two common misconceptions about bounce rate.
Misconception #1 – Bounce rate is defined as the time spent on page
Many people think that the definition of bounce rate is somehow correlated with the time spent on the entrance page. That’s not true. It doesn’t matter if your visitors spend 3 seconds or 3 hours on your entrance page. If they don’t visit another page, then it’s counted as a bounce. If they do, then it’s not a bounce! Simple as that.
Misconception #2 – There is a good or bad bounce rate
Your optimal bounce rate depends on many things.
For instance, if you have a blog, most likely your visitors will come to read one article at a time. Understanding the above, you will know that there will be a lot of bounces – even if your goal (read your blog post) was fulfilled. So on a blog, even an 80% bounce rate could be considered an acceptable bounce rate.
On the other hand, if you have an ecommerce shop and you want people to browse through your products, you would expect them to click through your product listings. So for an ecommerce shop, an 80% bounce rate can be considered an unacceptable bounce rate.
These are only 2 small examples, but it shows perfectly that your optimal bounce rate depends on many things:
- the type of your website
- the type of your business
- brand awareness
- visitor types
- product type
You have to define your optimal bounce rate based on your historical data (e.g., average bounce rate in the last year) and not on misleading benchmarks. Also: evaluate the different landing pages, comparing them to each other, not to other businesses.
How to measure bounce rate?
There are many analytics suites out there. Google Analytics is one of the most well known. It’s worth mentioning though that you can measure bounce rate in your own database (e.g., SQL) if you log pageviews and clicks. It’s up to you which way you choose!