Bounce rate may be the most requested (by upper-level executives) metric, along with also being considered one of the most useless metrics in web analytics. If you are unfamiliar with this metric, it’s basically the percentage of people who enter a web site (on any page), and then leave (or bounce) without viewing any other pages.
Now while I will agree that it is over-requested, and not always the best metric to gauge a site’s performance on, I also do not find it entirely useless either. Admittedly the page-level bounce rate is the most useful way to look at bounce-rate. It gives you more data than just which pages are being used as entry pages, but how successful each page is at engaging a visitor to visit other parts of your site (or click on a particular call-to-action on that page).
However I am not entirely convinced that the site-level bounce rate (the over-all bounce rate for an entire site) is a useless metric. I think the usefulness (or uselessness) of this metric is really determined by a number of factors including:
- Web site type
- Web site goal
- Overall marketing efforts
As I’ve advised my clients, if their site is a blog, they can expect a high bounce rate and that this is entirely normal. Most people read blogs to get at specific information and then leave. I often enter a blog on a particular article I am interested in and then go on my way. The blog may be a great blog with tons of great articles in it and I may be back to read others at a later date when Google brings me back, but I rarely seek out additional articles after completing the one I came to read. From folks I’ve talked to, this appears to be fairly typical behavior. Blogs often get higher traffic to their articles than they do their home page.
If your site is not a blog, naturally you would expect a bounce rate much lower – and typically I find a bounce rate of 25% or less is fairly typical. In regular web sites I keep an eye on the site-wide bounce rate primarily to see if it ever spikes up above 25% and really take note if it spikes above 50%. At that point clearly something is wrong and requires some checking into. The problem could be anything from the tagging to a botched online marketing campaign (one of my clients did a massive PPC/Banner ad campaign which doubled their traffic, and their bounce rate soared up past 60% – their campaign targeted the wrong audience).
Now it is a gamble to provide this metric to executives (aka HiPPOs – Highest Paid Person’s Opinion), as many will freak out if the site bounce rate is over 10%. However, I am always prepared if an executive asks for the site-bounce rate and then wants to drill me as to why it’s over 10%.
So, while a site-wide bounce rate may not be the most useful metric to determine overall site performance, I find it’s not completely useless either.