Bounce Rate. What does it mean to you?

Bounce rate means different things to different sites.

[Demaestro]

I was reading about some people who were concerned about their bounce rate getting higher. They were talking about all the good SEO things they had done but their bounce rate was going up.

This got me thinking about my bounce rates. I have some information sites. I have some E-Commerce sites.

Wouldn't it be true that if an information site was properly designed and structured and was very SE friendly wouldn't a high bounce indicate success? Someone is looking for something, searched, found your page in the SERP and left because they got the information they were looking for.

Now I would hate to see high bounce rates on my E-Commerce sites, but if those E-Commerce sites have physical brick and mortar stores also and your top page visited is the contact page then wouldn't a high bounce rate from that page again indicate they found what they are looking for?

All of a sudden I am unsure of how to look at this number.

I am curious to know how most of you interpret your bounce rates.

[cgrantski]

You are absolutely right to look at this stat (and many others) in light of what is happening from the visitor point of view. Maybe you want all visitors to be completely satisfied with what they see on the landing page. Almost all site stats can be interpreted in both a good and bad way, which is why it's a shame to see people running around panicking about their stats as a result of applying some kind of black and white interpretation.

[pageoneresults]

I always like these topics...

I am curious to know how most of you interpret your bounce rates.

I take into consideration the Time on Site also. For me, high bounce rates typically mean that the user found what they were looking for (information sites).

High bounce rates for an ecommerce site with no checkout are a concern. There are many things that would affect that one. I think deal shopping is one of the top factors. Maybe those generating the high bounce rates will be back to purchase if they've found you're offering a good deal?

There are other metrics to take into consideration while looking at bounce rates like, how many of those bounces are human? And, could someone program a bot to sit there and generate high bounce rates for your site and wreak havoc on the metrics and the performance of your site?

[EarleyGirl]

There are other metrics to take into consideration while looking at bounce rates like, how many of those bounces are human? And, could someone program a bot to sit there and generate high bounce rates for your site and wreak havoc on the metrics and the performance of your site?

What happens if you discover a competitor is doing this? How does one protect their site from this kind of sabotage?

[onlineleben]

High bounce rates for an ecommerce site with no checkout are a concern

sometimes it sees to be a high bounce rate as the visitor only visits one page on your site, but proceeds to an orderpage on an other site (merchant), where you cannot install your tracking.

... how many of those bounces are human?

When using GA for tracking, shouldn't all logged visits be human as the tracking code is Javascript?

[Demaestro]

shouldn't all logged visits be human as the tracking code is Javascript

That isn't as true as it used to be. I have written code that fetched pages with 'Javascript' turned on. Although it really isn't javascript... what you do is.

Basically what you do is read the js code in and rewrite the same logic in another language however this is something I would only do for specific sites so that I could automate things... this doesn't really work generically across the board.

I have also heard of people writing a javascript interpreter, using the Mozilla javascript engine, but that is something I don't have the time for. To write one for Urchin would be insanity.

Anyway I am finding more and more that bounce rate is very open to interpretation and has more to do with what your site is and what people are using it for more then anything.

[DirvanMolando]

Bounce Rate should be in a context.
The definition of Bounce Rate is visitors that where exposed to just ONE page view before they left the website. I would also consider looking at the dwell, i.e. and staid less than 10 seconds. Please consider the examples below:

1. Campaign - You SEO'd a landing page for a specific campaign and you get a lot of love from SERP. If some one goes to this landing page (just one page, say collect a lead) and then took off. Will you count it as a bounce?

2. Media - You have a website that shows short clips, like you tube. A visitor might watch one clip and leave after "He did what he meant to do which is watch a clip". I would not consider adding that to bounce rate as well.

I hope this make sense.

[seodreamer]

Bounce rate is an interesting topic.

If your site is heavily focused on advertising and PPC schemes then a high bounce rate would be good, most people who bounce would probably have done so through an ad click.

However, think about the long tem affects of bounce rate. A bounced user probably didnt rate your site very highly, meaning they're less likely to blog about it, link to it etc.. A quick win is almost always a short lived one.

To sum it up, bounce rate means different things to different people. Personally i think it counts alot

[JaminFung]

U R right

[seth526]

Most of the people assume bounce rate as visitor leaving from the first page itself.What if the user find whatever he is searching on the first page itself.Would you call this visitor as bounce?

I am definetly not going to consider this visitor as bounce,i will look how much time he stayed on that page.Even for E-comm site if the user stays long definetly he is looking into the every part of that page,may be he might be comparing it with some another page he saw to take a desicion.So definetly your page will leave some impression in the users mind.This time he did not make any purchase from you,but he will definetly consider you in future.

[cgrantski]

Sounds like some people are getting "time on site" even for one-page visits. If you're getting that, how are you doing it?

[Demaestro]

cgrant that is a really good question... the more I think about it the more it hurts... I can see if you click out of the site with an external link that is could catch that but if you just straight urn off your computer or browser or type in a new URL directly how they catch that.

[mfishy]

I find that the more "authoritative" domains I work with actually have a far higher bounce rate. This seems to be largely due to the fact that these sites get way more unrelated traffic than less "powerful" ones.

Of course this is a symptom of bad search algos, but that is an entirely different topic!

[Demaestro]

Good point fishy... just another angle to look at it from.

[markymark]

In and of itself, 'bounce rate' is a pretty meaningless metric. You need to look at what phrase or site has brought the visitor to a particular page in order to judge whether or not that page has failed to deliver what the user is looking for.

You also need to make a judgement on whether that page fulfilled it's requirement for particular visitors precisely ; causing them to find what they are looking for and then going away again.

The essential problem here is the words 'judgement' and 'judge' - basically 'bounce rate' tells you that people have left your site but everything else is just interpretation, guesswork and judgement.

Forget 'bounce rate' as a metric for all but the most keenly focussed lead gen/sales page type sites.

[conor]

Like every metric Bounce rates should be taken in context and cross referenced with other metrics if any meaningful conclusions are to be drawn.

[johnnie]

What about the people printing information pages? That too will lead to high bounce rates but a satisfactory user experience.

[incrediBILL]

Bounce rates can mean many things:

- Bad traffic from overly aggressive SEO bringing in the wrong visitors

- Crappy web site with the wrong message for the right traffic

- Site required technology like Flash or Javascript that the visitor a) didn't have installed or b) had disabled or c) was incapable of enabling due to corporate restrictions or device restrictions

- The product or service didn't meet the requirements at first glance

- The prices were simply too high

Instead of grasping for straws in the dark, one way to attempt to find out is trying to run an exit poll for anyone that doesn't stay at least 30 seconds or doesn't visit a second page to see the reason.

Additionally, offering an instant discount for "bouncers" can sometimes bring them back into the fold and convert them to customers.

At any rate, try asking them before bouncing to conclusions.

Sorry, couldn't resist ;)

[BillyS]

Our bounce rate is high but we're hoping that this isn't counted against us. We provide information on a relatively LONG page - sometimes as many as 2,500 words.

We could probably reduce our bounce rate by breaking these large pages into sections, but that's only gaming the system...

[EarleyGirl]

A high bounce rate can also mean a competitor has targeted you. We are experiencing this and from my research so far, there's not that can be done about it.

[incrediBILL]

Our bounce rate is high but we're hoping that this isn't counted against us

It's not a problem unless you have advertisers on your site because the bounce rate could, and should, drive down the price of your advertising.

Other than that, why would anyone else care?

[annej]

how many of those bounces are human?

How can you tell using GA?

I do think you get a higher bounce rate when the page is ranked well especially on general terms. You get more people just surfing around. I have no way to measure this and wonder if anyone has.

[koan]

It's not a problem unless you have advertisers on your site because the bounce rate could, and should, drive down the price of your advertising.

Doesn't having an ad on a page increase your bounce rate since a percentage of visitors will exit the site when they click on that same ad? (I'm not even kidding, I could see the effect on GA when I removed a big square ad)

It just seems kind of silly if the presence of an ad on a page drive down the price of that ad.

I have a low quality site (from an advertisers perspective, it's a great site with good content, it's just not converting) with very low bounce rates, and high quality sites with high bounce rates. It really depend on the subject. I'm sure your average celeb wallpaper site has a very low bounce rate since most visitors would probably check a few pictures of a celeb they just searched for, and a long and thorough technical page on widget X would have a high bounce rate since everything imaginable would already be communicated on that page. Yet the celeb site is probably low quality for advertisers and the technical one would be high quality.

I just don't agree with you that bounce rate should be used for an ad price, unless it is taken into consideration among many, many other factors before it defines the quality of a site.

[Receptional]

Pageone:

I take into consideration the Time on Site also.

Well - what Crganski and DeMaestro said. Bounce rate means really that the user only hit the site once. So in general you can't measure how long the user engaged with the site. That's a real problem, but couldn't someone write a script that refreshed every now and then? I'm no programmer, but I do see pages that do this for other reasons. Presumably this could be done, but it would really need to be built into the analytics system being used otherwise it's really not helping man nor beast.

[gdawg]

One thing to keep in mind is that you cannot accurately measure avg time on page for those pages that do have high bounce rates. The reason is that most anlytics programs calculate avg time on page by tracking the time of the initial page request, after which they wait for a second page request and then substract the 2 request to get the avg time on page. Since pages that have high bounce rates, usually only have 1 request you will only be able to get avg time on that page for visitors who had at least 2 page request.

[Mohamed]

Websites that are browser's default homepage would have high bounce rate. so you can't deduce anything from high bounce rate alone.

[honestman]

As with advertising on an information site, one of the supreme paradoxes is that content that is TOO GOOD--or too rich with links to other useful information--often results in an audience member likely to leave the page to take up one of the suggestions or links from your content.

I personally have no problem with this whatsoever on an information site. I think that Google and even Yahoo originally built itself as a "pass-through" and thereby gained loyalty. It was and remains the honest thing to do.

Some information sites--where you have to navigate down several levels to get any information at all--get on my nerves and are just not honest. I avoid them as much as possible. Likewise, sites which put in "frames" external links use a level of control which I find counter to the spirit of the Internet.

I think it is best to be honest and people will come back--whether they use you as a pass-through or not is irrelevant; you can always sell them something along the way if they come en masse.

This does not apply to forums and e-commerce sites, of course.

[dealguien]

To measure time on site on a single page, the GA (or other) javascript just needs to regularly send a ping to the server via ajax. Maybe every 10 seconds or so...

[Virtue]

As you said that you have e commerce sites then I don't think you should worry about your bounce rate if it is increasing. As people getting something that which they don't want. But if it is about your information sites then information should be beneficial for the people

[karamkshetra]

Basically in bounce rate you should consider the number of exits, pages consumed and time spent on each page.

As if your website is an information site a user will stay there for sometime read your content and than move on... so basically here time spent can be a good factor to consider