High Bounce Rate? Blame Google Analytics

I've read a lot of stuff online from the experts, especially about having a quality blog.  One of the indicators that gets thrown around a lot is something called bounce rate. According to those in the know, if you have a high bounce rate in tools like Google Analytics, it means readers simply aren't engaged and sticking around.

I have a high bounce rate.

Before I begin whining about my bounce rate, I'll explain.  The bounce rate is a percentage of visitors who visit only 1 page on your site before leaving. This could be for any reason, they click a link, or close the page figuring you don't have anything they want to read about.

I have always had a bounce rate, and to be honest, I have taken it personally.  I spend a lot of time on my other blog trying to build good useful content and building an audience.

I do all the things they say to do to keep visitors around.  I have related posts widgets, popular posts, comments.  I've done everything short of begging people to stay.

According to my Google Analytics reports my bounce rate remained stubbornly in the 70-80% range no matter what I tried.

To add insult to injury, as I drilled further down into my reports, I found a vast majority of visitors hung out for less than 10 seconds!  That made me feel real good about myself, like the whole world was telling me I sucked. 

I decided something must be wrong.

Looking at the stats I noticed a trend.  People either left in under 10 seconds or they stayed for 5 minutes or longer.  This seemed odd.

What was even stranger was that there were a lot of visits that lasted 0 minutes.  Something clearly wasn't right here. 

I decided to search on the elusive 00:00:00 that was showing on my Google Analytics page.  And that's where it hit me.  It might not be something I'm doing wrong at all.

It's all in the technical details

This is going to be a little geeky, so bear with me.  The way most analytics programs work is the put a script on your site. When a user visits, the script is activated phoning the mother ship to begin tracking the session. Each time the visitor clicks a page another signal is hit.

The time between the signals is where the time on page calculation comes from.  So what happens when there's only one page visited? 

The analytics server has no idea how long the visit lasted, so it marks it down as 0. Mystery solved.  Technically, it's still a bounce, just the time on page is inaccurate.

That leaves us with only half the picture.

See, bounce rate isn't exactly the best measure of how well a page is performing.  This is especially true if you want the visitor to take some action that involves leaving your page, such as clicking an ad or reading a news story.

A better way to tell if someone is engaged is to see how long they spent reading the article.  You've spent what, like 2-3 minutes getting to this part of the article?  I'd say you're engaged.

Really, a bounce rate should be a measure of how many readers aren't engaged.  A better metric would be how many people ditched you after 30 seconds.  Perk up because...

I have a solution!

Turns out there's a couple of ways to get a better picture of true bounce rate.  You can either adapt Google Analytics to your will, or try a different analytics program.  I'll explain both briefly.

Solution 1: Correct Google's Oversight

This one is kind of hands-on and geeky, but it works.  

Basically you add a piece of script that tells the browser to contact Google every 10 seconds (I recommend you increase that to 30 or 60). Any visit longer than the threshold you set will not count against your bounce rate.  The time on page will also be registered correctly.

Here's a great article giving you the steps if you want to try it (see you just clicked there and caused a bounce.)

To be honest, I think there are some accuracy issues here.  I set my script to phone home every 60 seconds and had single digit bounce rates.  That seemed too low.  I also ran into an issue where one day the number of visits was higher than page views, which should be impossible.

The first solution seems a little buggy, so I looked for another idea and came across the next solution

Solution 2: Use a Different Service

I ran across another analytics service called Clicky the other day.  It's a slick web 2.0 service that offers some really easy to read metrics.  It also tracks stuff Google Analytics can't, like what outbound links people clicked on (handy if you want to track ads).

But most importantly, Clicky decided it didn't like the way bounce rate was figured either.  Clicky only counts single page views of less than 30 seconds as a bounce by default.  It also counts actions like clicking an outbound link as a positive thing.

I've only used Clicky a short while, but so far I like it.  It comes in a free version with basic features, or paid versions that offer all kinds of fancy upgrades like real time monitoring and Twitter analytics.  You actually get the premium version for 60 days free when you first sign up.

Setting Clicky up is pretty easy, they even have plugins for most major blogging platforms. This might be the easiest way to keep your high bounce rate from driving you crazy.  Click here to sign up.

I'm a much happier guy

I suppose you could also just realize the bounce rate metric is a little junky and a high bounce rate on Google Analytics doesn't actually mean much.  Me, I prefer some kind of number.  Now that I have an analytics program that shows me a much more realistic 20-30% I'm a happy guy.

Until Next Time,

Steve K

Disclosure: Those Clicky links are affiliate links

Image: Ambro / FreeDigitalPhotos.net