Should we pay attention to bounce rate, page per visit and time-on-site?
When is it the right time to use metrics such as bounce rate, pages per visit and even time-on-site? Are they of value or can we just completely ignore them? Some marketers believe they are valuable, some think we shouldn’t pay any attention to them, whilst others sit in the middle and think their value is dependent on the context of what is being measured. The answer is probably with the latter; these metrics rely on many more data points to enable them to provide insightful learnings of on-site engagement.
Firstly, what are these metrics (most commonly reported on via Google Analytics)?
- Bounce rate – Average percentage of users who have landed on a page and then left the website without continuing onto another page on that website
- Pages per visit – Users who came through a particular page and then visited another page(s)
- Time on site – Average time a user spends on the site
A couple of factors that these metrics rely on to make them valid:
- Traffic source
If a user lands on a landing page via Twitter (or other social channels, and on mobile), there is high chance that they are going to have high bounce rates, low pages per visit and time-on-site. This is because this is the type of behaviour most commonly found from those traffic sources – A user finds a Tweet or Facebook post with a link to an article, they are only clicking through for that article and not for the website which would result in high bounce rate and low pages per visit. If the content did not meet their requirements, that will then result in low time-on-site (high time-on-site if it does).
However, if a user comes through the Google search results page via an organic result (and an informational search), the reverse should happen.
- Purpose of the landing page (and content)
On-site engagement metrics are continuing to have a bigger impact on rankings with bounce rate, pages per visit and time-on-site being part of those on-site metrics. The extent of their impact on rankings can be a bit of grey area as much as their value. This leads to how the purpose of a landing page and the content can be a factor.
For example, if a landing page is well optimised for SEO, audience-led content which answers the user’s search query, when the user lands on the page, their search query is fulfilled and exit the website, that landing page will have a 100% (or very high) bounce rate. Does that mean the landing page is poor? It’s done its job by providing the information the user was searching for.
Another instance is that some websites want their users to have a low time-on-site. Websites that have a purchase funnel want their users to complete the purchase as quickly as possible. If they are spending 10, 15 minutes instead of 5 minutes, it’s likely they won’t complete their purchase and drop off – But the website has a high time-on-site, is that a good thing?
All Response Media viewpoint
Ultimately, these metrics cannot be depicted on their own without considering other factors to give more context if these metrics are providing useful insights or not. There are so many moving parts that can determine the use of these metrics if they are useful and not useful metrics and therefore pay attention to them.
If you must report on these metrics to internal and external stakeholders, ensure there is context behind the metrics to give a fully rounded view on what they are reporting in terms of performance.
Here at All Response Media, we utilise additional metrics/tools to analyse and evaluate on-site engagement such as heat maps and click maps to see where users have interacted with the page. Another is setting up Events to fire when certain actions are completed on a page, for example, scroll depth to see if users do scroll down and read all the content, etc. The insights these methods provide enable us to make the right UX and optimisation recommendations (and test recommendations) to improve performance.