Occasionally, web analysts come across this question about the role of bounce rate in web optimization and analysis. We cannot understand the role of bounce rate until we thoroughly understand what bounce rate is. In this post, I wanted to discuss the different methodology used for calculating the bounce rate.
To start with, bounce rate is the coolest metrics because it can give you a good understanding of the website performance. It shows how many people visited only one page on site. Here is the exact definition from Wikipedia – “Bounce rate represents the average percentage of initial visitors to a site who “bounce” away to a different site, rather than continue on to other pages within the same site.”
There are different ways to calculate bounce rate. Some use the time spent on the site data and some use the single page visit data. Let’s discuss all the techniques in detail.
Time on Site Methodology – Let’s imagine a person is searching for holiday gifts and he/she types “holiday gifts ideas” on Google. Fortunately, your website shows up at the top position out of 928,000 search listings . The visitor immediately clicks your site listing and lands on the home page. After few seconds on the site (less than 10) the visitor is not impressed with the site layout and shuts the browser window . This is considered as a bounce because the time spent on the site was too little to encourage clicks or engagement.
In our example, the visit was a single page (home page) visit. However, there is a chance that someone lands on the site and quickly clicks an internal link and then exits the site from that page. This will not be considered as a bounce using the single page visit methodology (coming up next). It is still considered as bounce using time on site method because the time spent on site is less than 10 seconds (some use 5 seconds).
Time on site is a good indicator of the site bounce rate but unfortunately most of the analytical solutions cannot spit out this data directly. You have to use the time on site report and calculate the bounce rate manually.
Single page visits – Just like the name suggests, single page visit bounce rate calculation is based on each single page site visit. It’s the percent of the time a visit is a single page visit. This is the least recommended of all methods. The reason single page visit is not considered as accurate bounce rate is it does not count the second and the third visits to the site by the same visitor.
To understand why second and the third visit by same visitor is important in the bounce rate calculation let me give an example– John, an imaginary web surfer, visits a page A on my site and exits my site form the same page A. This visit is counted as a single page visit. Ideally, this is a bounce rate of 100% but the story does not end here. John decides he should give my site another chance (lucky me ;). He revisits the same day the same page A on my site but this time he also visits pages B and C. Using the single page visit formula, John’s visit is still a bounce, but in reality it is not due to his second visit.
Single page visit bounce can only be accurate when every visit on the site is a first visit.
Single page/Entry page – This is one of the technique I recommend to calculate the bounce rate and it is also the most popular bounce rate calculation. This method incorporates the second or more visits data by the same visitor in the calculation and eliminates the single page visit discrepancy.
According to this technique, a bounce rate is a percentage of total number of single page visits divided by total number of entry page visits for a given time period. Notice that the denominator is the entry page visits.
Going back to our imaginary surfer John’s example, we see that page A was the entry page two times and so the new bounce rate for the site will be –
Single page visit / Entry page visit = ½ = 50% (compared to 100% using single page method)
Most web analytics tool, include this bounce rate report by default.
True bounce – I promise this is the last one . Both, the time on site and the single page/entry page method are good and each has its own advantage and disadvantage. I decided to use the combination of both, and I call it as a true bounce. There is no direct method or analytics reporting hack to spit out the combination bounce rate, but I calculate true bounce by using averages of both bounce rates.
I think the true bounce is as close as one can get calculating the accurate bounce rate. This method not only considers the time spent but also incorporates the single page visits and multiple visit criteria.
Segmentation – Site bounce rate is an important metric but without segmentation you can never know the reason behind the reason (why visitor bounced off from your site?). It is important that you calculate the page level, campaign level and search keyword level bounce rate.
Please provide your comments, feedback, or tweets here.
Here is a list of other helpful blog posts –