Bounce rate is a popular metric used by Google Analytics and similar statistics accumulators to measure the “stickiness” of your website. The notion of what constitutes a good bounce rate is rather controversial and so I wanted to add a bit to the discussion, as it is often the source of confusion to newbie webmasters (and often to people that should know better, too).
What is a Bounce Rate?
Google defines Bounce Rate thusly:
Bounce rate is the percentage of single-page visits or visits in which the person left your site from the entrance (landing) page. Use this metric to measure visit quality - a high bounce rate generally indicates that site entrance pages aren't relevant to your visitors...
http://www.google.com/support/analytics/bin/answer.py?answer=81986
While I have no doubt a website’s bounce rate is an important factor in Google’s ranking algorithm, I utterly disagree with the universal applicability of this bit (or even that it's generally true):
“a high bounce rate generally indicates that site entrance pages aren't relevant to your visitors”
I would argue that the opposite could also often be true, it depends greatly on the specifics of one’s target market as well as a website’s architecture.
A high bounce rate could mean as Google indicates, “that site entrance pages aren't relevant to your visitors” or it could well mean that the page was exactly what a visitor is looking for. The bounce rate metric doesn’t tell you.
(Google, of course, knows more than it shares with you because Google knows what visitors do after they “bounce” – if the user keeps searching for the same keywords, that’s a somewhat better indicator that what they found wasn’t what they were looking for (though it could mean they just want more). If they don’t keep searching, that’s a pretty good indicator that the content was spot on, despite the ugly looking “bounce”.
Consider the case of a website that displays local movie times. How many pages should a user looking for Albuquerque movie times have to visit? In my opinion the answer is probably just one.
Certainly, a site could optimize a page for “Albuquerque movie times” and then ask a user to “click here for Albuquerque movie times” – and that would surely lower the bounce rate. On the other hand, a good percentage of visitors would think that the site is some sort of spam and move on.
Here’s another example.
Suppose you have a page called “Images of Rio de Janeiro” that’s full of images of Rio de Janeiro that visitors can view in an image gallery located on the very same page. That’s great for users. No clicking around, no waiting for pages to load. But for your bounce rate? Not so much.
On the other hand, you could make users navigate to a separate page to view each image and watch your bounce rate plunge – and also watch users get irritated with your site and leave after a few photos.
It’s easy to think of examples and counter examples. A site that displays local temperatures vs. dating site. An online thesaurus vs. a classified ads site. A wikipedia page vs. a community forum.
My opinion is that this metric is often meaningless and one is better off using common sense as your guide than blindly seeking to improve the numbers relative to this flawed statistic.
Anyone can get himself or herself an impressively low bounce rate by optimizing a landing page for “your keywords,” then slapping a big flashing red button in the middle of the page that reads, “Click here for “your keywords” and sending them a page with the actual information they’re seeking. Congrats, you’ve got a low bounce rate – but so what??
On the other hand, if your keyword is “Toledo Real Estate listings” and you have a high bounce rate, you either don’t have any Toledo real estate listings on your page for viewers to browse, or your landing page needs a serious redesign. In which case, Google is absolutely correct and bounce rate is a great metric.
Understand that bounce rate “is the percentage of single-page visits or visits in which the person left your site from the entrance (landing) page” and nothing more.
Use your knowledge or your market to determine if it should mean something. If not, forget about it. If yes, then work on it. It’s as simple as that.
How to improve your bounce rate?
There are a couple of reasons for high bounce rates where the bounce rate should be low.
Don't Bait and Switch... or Anything Even Close To It
In those cases where bounce rate should be meaningful, it's often as simple as thinking about what you're advertising and what you're delivering.
Suppose you have page optimized for the terms "weight loss tips" but when visitors arrive, all they see are a bunch of ads for diet pills. Well, in a case like that, a high bounce rate could well be a problem. While both have to do with weight loss, think about what someone who is searching for weight loss tips is looking for... a list of things they can do to lose weight. A better approach would be to provide them with a list of real, well-research weight loss tips and provide little detours with links to what you're selling along the way... Click here for one of the most talked about weight loss books of the decade... Click here to read about various diet supplements that can compliment these weight loss tips, and so on.
In other words, make sure your page gives the visitor EXACTLY what you're advertising with your keywords, meta-tags, etc. - and give it to them well. Be the authority. One stop shopping.
Consider the case of an online marketing campaign. If your ad reads "Buy Classic Cars!" but what you really have is an everyday used car lot, you're going to get a lot of bounces when they see a 1987 Ford Taurus rather than a 1963 Jaguar. While some folks might think of a Taurus as a modern classic, a better title might be simply "Affordable Used Cars".
Don't Bury Your Relevant Content
One reason why people bounce is that they don't see what they're looking for. Make sure the first think a visitor sees is what he would expect to see, don't make him or her have to scroll or click. People have very feeble attention spans nowadays as it is, and even more so online.
As much as search engines like Google might try, people still have to wade through a lot of spam and are used to making judgements about a page in a split second. If your page is optimized as "Weather in Cancún, Mexico" put the weather information right at the top. Don't make them have to wade through a bunch of timeshare promotions (however profitable timeshares may be) to get that weather information. Give then the weather information first, but let them know they could enjoy that lovely weather on a regular basis if they bought a timeshare.
Presentation
That is really just another way of saying credibility. If someone arrives at your site and is bombarded with popups, AdSense all over the place, a lousy design, a gazillion links, misspelled words, sloppy CSS, an ugly design, etc. how much do you think they're going to trust you? Not much. If your site is sloppy, pushy, or too just busy, expect a higher bounce rate.
Performance
Needless to say, people get irritated with slow sites and bounce. Make sure your site is zippy. Nothing is more irritating than a page stuffed so full of stuff that you have to sit and wait for all the images, ads, flash, and other assorted web junk to load. BOUNCE!
Architecture
If your site needs for people to navigate to find what they're looking for, make it as easy for them as possible. Give them clear indication of what they need to do next.
Links
An easy way of lowering your bounce rate for pages that provide information (i.e. words, words, and more words) is to link to other pages that have relevant, related information right there in the body of your text. For example, click the following link to see a page that effectively uses internal links to lower the bounce rate.
These are just a few of the more obvious ways to lower your bounce rate. But again, bounce rate is only relevant where it should be relevant.
Don't try to move mountains or walk on water and certainly don't fret if you have a high bounce rate. If you think you're doing everything you can, chances are you probably are. Different target markets have different bounce rates and some have a naturally "high bounce rate."
Don't let folks who claim that a bounce rate of 20% is a sign of quality and a bounce rate of 50% means you have a low quality site. In most cases, that is just complete nonsense.
Here's an example of a site with a bounce rate of 20% and here's an example of a website with a bounce rate of over 50%.
Guess which one does better?
