What is the Ideal Bounce Rate for a website?
I’ve been asked this question a few times in my career and it always confuses me. It’s a bit like “how long is a piece of string?”.
Bounce rate will differ by page, by search term and by search intent. There is no universal “good” or “bad” number.
That’s not to say bounce rate is a useless metric and you can’t benchmark between, for example, different product or service pages to identify areas for potential improvement.
What really is a bounce?
To confirm we are on the right page (before you bounce!) let’s look at what the true definition of bounce rate is. I will focus on Google Analytics in this article as this tracking solution is used by the majority of businesses we work with.
“the percentage of all sessions on your site in which users viewed only a single page and triggered only a single request to the Analytics server.”
Assuming your tracking is working correctly a user who bounces may do so for one of two reasons:
- They successfully found the content they were looking for and have now moved on to some other activity.
- The landing page did not offer what the user expected and they were not enticed to explore your site further.
A user who visits any additional page on your site has not “bounced” but has not necessarily found what they were looking for or had a good experience.
Consider these two scenarios:
- A user searches for a factual query such as “iphone 13 price”
- They click through to the product page on the Apple website.
- The webpage provides the answer the searcher needed and the user closes the browser tab.
- The data says: This is a bounce yet the specific user need has been satisfactorily met.
- A user searches for “size 12 red dress”
- They land on a product page from Retailer A.
- The webpage shows the product as being out of stock.
- “Related products” links appear on the product page and the user clicks through to one of these viewing a second product.
- The 2nd product is not available in the searchers size.
- The user is disappointed and decides to try another retailer. They navigate back to the Google results page and click through to a competitor, retailer B.
- The data says: This is not a bounce as the user viewed multiple pages during their session, yet the user need was not satisfactorily met.
In short – a high bounce rate does not mean your content is poor.
Google wants to kill factual searches with high bounce rates
For user intents such as those outlined in Scenario 1 above Google tries to provide answers in-SERP without the need for a user to click through to a third-party website. It does this by scraping information from publisher websites and presenting this information itself in an answer box or featured snippet or answer box in order to improve user experience.
If your business relies on such terms for a significant amount of traffic at the moment you need to consider your SEO strategy as this is not a sustainable approach long term. Where simple factual simple answers Google is fast working to automate their provision of these negating the need for a searcher to ever visit your webpage.
Is bounce rate a Google ranking factor?
No. John Mueller, Webmaster Trends Analyst at Google, has stated that Google does not use Google Analytics bounce rate as a ranking factor.
Reporting on bounce rate
So, if it can often be misleading and is not a ranking factor is bounce rate a useful metric at all?
Absolutely – it all depends on the context. Bounce rate can be an extremely useful metric when comparing the performance of an individual page over time or if when undertaking a content audit you notice a bounce rate is not in line with your expectations for the page and the search queries currently sending traffic. This may highlight issues with the quality, relevance or timeliness of particular articles or even technical difficulties or obtrusive page elements which may be causing an increased number of visitors to leave after a single page view.
You can also use bounce rate to prioritise webpages with which to improve internal linking to ensure visitors do have a “next step” in their journey on your site.
Modified bounce rate
Use of a modified bounce rate metric enables you to change how bounce rate is measured by your own parameters. This is possible within Google Analytics for example you may wish to exclude any single-page view session of over 30 seconds in length from being counted as a bounce. Sometimes referred to as an “adjusted bounce rate” such customisation enables you to filter out sessions of a duration which show that a user is likely engaged with your content.
Better measures of engagement
Performance of content should not be measured on a single metric alone. Other indications of the performance of individual articles and sections of a website on aggregate can also provide insights into the experience offered to users:
- Time on page – similar to modified bounce rate as discussed above.
- Scroll depth – are users viewing all of your content or only a small section of the page?
- To identify drop off points use heat-mapping or session-recording software to visualise how your users interact with your content.
- User feedback – this may be as simple as asking users to give an article a thumbs up or down, as MoneyHelper does at the bottom of each article or promoting users to provide written qualitative feedback on a particular webpage.
How to lower bounce rate on non “instant answer” queries:
- For ecommerce sites make sure your product pages contain links to relevant in-stock products to capture users who land on an out of stock product.
- Use geo-ip detection to show a message to users who have landed on content targeted at a different market to direct them to their local version.
- On editorial or blog content ensure you link to related articles to encourage users to explore your site further.
- Strong internal linking to reference other articles or cite statistics hosted elsewhere on site can increase average pageviews per session.
- Check the Performance report in Google Search Console to identify the specific search terms sending traffic to a page, does your page provide an answer to this term?
Experiencing a very high or very low bounce rate?
Here are some tips to help determine the cause of extreme bounce rates:
- How big is the sample size? With only 1 session landing on a page your bounce rate will either be 0% or 100%. Small sample sizes give very little insight into true content performance. In these cases consider an aggregate across all pages of a similar type e.g. all blog posts, all product pages, all dress sub-category pages.
- Monitor sudden spikes or drops in bounce rate for signals that your content is providing a poor experience, for example is out of date, or that a tracking issue has occurred.
- High bounce rates commonly occur if a large proportion of sessions are from outside your target geographic area – for example searchers from India are unlikely to want to buy the services of a UK company. Drill down in Analytics to see the bounce rate for your target markets.
- A really low bounce rate across a decent sample size is usually an indication of a tracking problem. Most commonly this is seen when your Google Analytics code is triggered twice on a page reporting two page-views almost simultaneously. This can occur for example if your code appears both directly in the page source code and is also implemented through Google Tag Manager.
In summary, there is no single criteria to demonstrate in isolation if a particular bounce rate is good or bad and bounce rate is only one of a range of metrics that can be used to measure content performance. This does not however mean that you should entirely ignore bounce rate when measuring the success of a piece of content. Bounce rate can be useful when comparing performance over time or between sections of a website and useful in order to identify opportunities to improve content provision.