Here are several suggestions for ways to keep visitors from leaving your site and enticing them to stick around and view more pages.
Bounce rates tell you what percentage of people left a given page on your website without viewing any other pages. It’s not to be confused with exit rates, which simply tell you the percentage of visitors that left the site from a page (i.e. they may have viewed other pages first).
Also, it’s important to be aware that users could spend 10 minutes on your page before they leave the site.
In this scenario, it could well be that the page has fulfilled its purpose (or that the user has just forgotten to close it).
It’s generally used, along with other metrics, as a measure of a site’s ‘stickiness’.
For example, on SEW, I’d like people to click on a link from search, Twitter or some other referral source, find a useful article, then decide to browse further and view all of our other lovely content.
If bounce rates are high, it could mean that our content isn’t doing its job properly, though there are plenty of other possible explanations.
As a rule, I’d generally look at trends over time, and use bounce rates as one of several metrics for measuring the success of a particular page.
For example, this Google Analytics custom segment looking at the percentage of visitors viewing multiple pages provides a measure of a site’s ability to retain users’ interest beyond the page they land on.
Other measures, such as average time on page or using event tracking to see how many people read to the bottom of your posts (as described here by Justin Cutroni) can also help.
Still, the principle is important. If you’re the kind of site that wants people to stick around for a while, bounce rates provide a good general guide.
What exactly a good bounce rate is will depend on the type of site you’re running.
Working in online publishing, my experience is that bounce rates for articles can be as low as 40% and as high as 98% for individual articles. The average would vary between 70% and 85%; obviously I’m aiming for nearer 70%.
The Google Analytics screenshot below shows some of the more ‘evergreen' articles we've published. As we can see, the bounce rates for such articles are lower than the average, which is nearer 80%.
In a word, no. It can depend very much on the purpose of the website.
For example, people may want to quickly find a contact number or check facts. If the site enables them to find this information easily, they’ll leave quickly, thus pushing up the bounce rates.
I may need to know how old Al Pacino is (as you do). I can Google his name, click on Wikipedia, and the information is instantly available on the right of the page. Then I hit the back button.
Of course, I could linger longer, read more and click some of the links, but if that’s all I want to know, I’m playing a small part in increasing the site’s bounce rates.
For publishers like Search Engine Watch, we’d rather keep people on the site longer so, if someone clicks on the page, decides they’d rather not read the article in question and leaves, then that may mean we haven’t delivered on their expectations.
In the latter case, high bounce rates are a bad thing.
Of course, Google doesn’t know your bounce rates, though it theoretically can find this information from the millions of sites that use Google Analytics.
In theory it would be a useful ranking factor, as it is an indication of how relevant your landing page is to the user’s search query, though allowances would have to be made for the type of site and query.
If someone wants a guide to landing page design and bounces within seconds we can assume the page hasn’t delivered. However, if they just wanted to quickly check the weather for today, then maybe it has served its purpose.
The concept of dwell time, or the ‘long click' (as explained here by Bill Slawski) is important. It’s similar to – but not the same as – bounce rates. It’s essentially a measure of how long a user spends on a page before returning to the search results page.
Whether this is a ranking factor or not is open to debate, but it certainly makes sense in the light of Google's search for quality signals.
In essence, it works like this:
Of course, this is a simplified version, and there are variables. For example, what if the site answered the query immediately (as in my Al Pacino example)?
I would assume that Google would be able to find different metrics for different types of search query so that it could take account of this.
The following factors should help to reduce bounce rates, but also should serve to keep users on site for a longer period. Or at least remove factors which will make them leave the site.
Here we are, in no particular order… Continue Reading