Monthly Metric: Bounce Rate

February 21, 2008 at 6:32 am | Posted in Advertising, Communication, Decision making, General, Marketing, Online marketing, ROI, Search, SEM, SEO, Usability, Web 2.0 | 2 Comments
Tags: , , , , , , , , , , , ,

Someone lied to you if they told you statistics were boring. Website metrics show just how your audience is using your site and you ignore this data at your own peril.

A bounce rate is when someone comes to your site and immediately leaves. They bounce off of your website for whatever reason. A bounce is undesirable – you want people to come and stay on your website! Bounce is the opposite of sticky.

Time vs. Pages

I had always understood bounce determined by time – that this figure was measured from people leaving a site in a certain increment (usually 2, 5, or 10 seconds). So I was surprised when I read in Website Magazine that they asserted that bounce rate “is calculated by dividing the number of total page visits by those visits that did not result in an additional page view.”

This seems insignificant, but really there is a huge gap. True, both methods measure engagement – how much your visitors care and want to get involved with your website (in the case of bounce, the answer is “not much”). However, the magazine’s definition is critically flawed (and I don’t mean to pick on Website Magazine – many others define bounce rate the same way).

If bounce is determined by page – measuring the amount a given page is the last page viewed against the aggregate number of times that given page was the first page viewed – this does not mean the user was instantly turned off. Co-worker Mia described this page-derived metric as indicating “not an unsuccessful visit, but unsuccessful engagement.”

For instance, I can search a site, take several minutes to read and digest a page’s information, be satisfied, and leave. This wasn’t an unpleasant visit – I got the information I was looking for – but it was unsuccessful in getting me interested in what else the site had to offer. Shame on the marketer for not including links or other content that might entice me to another page, or offering some reason to further explore that given topic. But that does not mean the content or design was flawed in some way. In this manner, I think it’s clear that determining bounce rate by last page is inherently flawed.

However, the time-derived metric is incredibly handy. Robust metrics engines like WebTrends can be programmed to measure any increment of time on a given page (though you’ll probably have to ask your IT guy for help – the usual default setting only measures in minutes which isn’t small enough for our purposes).

So, what can I do (and what does Jesus have to do with SEO)?

This uber-geeky battle of the bounce rates might not mean much to you – you’re probably interested in ways to fix high bounce rates. Let’s look at some common problems that might be plaguing your website.

  1. Bad design: If your website design looks cheap, people will logically think the content is cheap as well. Make sure your design is professional – check with an expert and ask others’ opinions (if you’re on a tight budget, use SurveyMonkey or another free tool to send around to friends and family). One important note: there is often confusion over cheap vs. simple when it comes to design. Google is simple and it works. Sites trying to sell you pirated DVDs from China often look cheap and it doesn’t work. Even useful sites like Jakob Nielson’s walk the line (though he’s likely proving a point about usability, so he gets a pass). Use your best judgment.
  2. Bad content: People can tell when they’re on the wrong site. Maybe it wasn’t what they were looking for. Or maybe your misspellings or wacky fonts or wrong information instantly turned them off.
  3. Bad site architecture: Think about this from the user’s perspective. I worked with a client whose homepage (i.e. http://www.company.com) was where user’s had to click whether they wanted the American or European versions of the site. They lost a fifth of all traffic due to this terrible landing page. Enable your visitors to get the right information quickly and you will have a better bounce rate.
  4. SEO: Search engines crawl certain information – title, header tags, first couple paragraphs. If you are overly clever or leave the crux of your argument at the bottom, the right audience isn’t going to find it. If Jesus had been a blogger, he would have had SEO woes. The parable of the mustard seed would have brought in gardeners and horticulturists – not exactly the point of the story.

Homework

For more detailed directions, check out Avinash Kaushik’s wonderful post at Occam’s Razor. He skims over the time vs. page argument that I wrote about, but he delves into some great suggestions for finding out more about your traffic and improving bounce rates. I especially recommend #2 (Measure the bounce rate for your traffic sources) and #4 (Measure the bounce rate of your AdWords, AdCenter, YSM (PPC) campaigns).

Many companies, especially small business owners, think they don’t have the time to properly measure metrics. The truth is that they can’t afford not to.

It’s logical that you need the right audience at the right time in the right place. Bounce rates are a perfect way to determine whether that’s occurring. Potential business could be floating to other websites instead of yours. Don’t be the metrics fool.

Advertisements

2 Comments »

RSS feed for comments on this post. TrackBack URI

  1. Jakob Nielen’s web sites offends my aesthetic sensibilities and I don’t want to spend any time there. Yuck.

  2. DJ – Good stuff. I agree that measuring the “time” element is a crucial way for understanding how effective your landing pages are delivering your message. However, I would suggest that there is a difference between ecommerce vs. informational sites. On an informational site, your reader may be getting everything they need on that one page – and, to your point, their visit may be successful. On an ecommerce site, where you live or die from making the sale, a high 1-and-Out rate may be sufficient cause to investigate.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com.
Entries and comments feeds.

%d bloggers like this: