Bounce Rate
When a visitor looks at only 1 page on a site, and leaves without looking at other pages, that’s called a “bounce”. The visitor found no reason to continue looking around on the site.
This could mean they found exactly what they were looking for, or that the page did not meet expectations.
Generally, a high bounce rate is an indication of content problems, but you need to check the traffic sources* to be sure. If the traffic sources are not relevant to the page, then the bounce rate is justified – the inbound traffic is poor quality.
However, if the traffic sources are good, and the the page bounces anyway, then look for problems on the page that are turning the visitors away – it’s usually due to poor, irrelevant or uninteresting content – the page content is not resonating with visitors for some reason.
* A traffic source can be a link on a website, a link in an email, a search engine result, a pay per click campaign – anything that sends traffic to the website is a traffic source
% Exit
The page on which the last trackable event occured is called the exit page. This metric indicates how often the page was the last page viewed by a visitor. Everyone leaves, eventually, but a high exit rate might mean there is a page content problem.
The % exit is calculated as the ratio of exits per page visit. For example, a % exit of 25% means that 25 out of 100 visitors leave the website after visiting the page.
Unique Visitors
The number of unique people that visited a site. Technically, it’s not possible to link a person to a website visit, but browser sessions are tracked to get reasonably accurate data.
Visits
The entire time that a visitor spends on the site is called a visit. When a visitor leaves the website for more than 30 minutes, and returns at some later time or date, that’s counted as a new visit.
When is the data relevant?
It’s important to view these metrics in the correct context. Are you looking at the metric for the whole site, or only one page? Did enough events occur to make the metric statistically significant? For example, if there were only 10 visitors to a page, and the exit rate for that page is 75%, is enough data available to make a judgment based on one month of data? Probably not. But if that metric remains unchanged month after month, then there is probably something to it.
