Wendy Schollum: Analysing web statistics can be tricky

Analysing web statistics can be a dangerous game.  While they can be a fantastic tool to drive positive change in websites, if misinterpreted, they can also mislead a business into thinking that their website is doing well, when it fact it is failing, or inspire unnecessary alterations and functionality.
"Unique visitors" and "bounce rates" tracking are examples of such dangerous statistics.  Many people mistakenly interrupt their monthly "unique visitors" statistics as a measure of how well (or not) their website is doing - assuming more visitors mean a better performing website.  What they forget is that this statistic also tracks those users who accidentally happen upon their website and leave immediately.
In order to measure your website's overall success, you need to become acutely aware of the "bounce rate" your website has and how to translate this statistical information into positive changes to your website design, content and layout.
A "bounce rate" is the percentage of visitors who enter your website on a specific page and then leave without looking at any other web pages - they "bounce" off your website.
However, it is important that "bounce rates" not be confused as being a site wide statistic (like "unique visitors").  You need to analyse content and keyword "bounce rates" separately (rather than looking at the overall "bounce rate" for a website) to reflect the levels of commitment visitors have to your website, based on where they came from.
For example, users who end up at your website by using a Search Engine or paid links are considered committed visitors.  They have actively searched for a keyword that relates to what you do/offer and have intentionally chosen to click a link that led them to your website.  If such users leave your website immediately (or "bounce") then this is a clear indication that your website isn't providing your users with what they need and should be revised. 
This said, website tourists can make up as much as 90% of a website's "bounce rate" statistics and shouldn't be confused with your website doing poorly.  These are users who have found your website through a directory you may be listed on, have searched for an obscure keyword and accidentally found your website (which is unrelated to what they wanted to find), or were just idly surfing the web.  They are completely uncommitted to exploring your website and (in most cases) will leave almost as soon as your page loads.  Your website and its content can hardly be blamed for these "bounces". 
Likewise, there is some argument as to whether or not you should have a high number of "bounces" from return visitors to your website. For example, if you send out regular e-mail marketing campaigns to your contact database, you will usually find an increase in your "bounce rate" on the days you send these.  This anomaly is caused by loyal users clicking through to your website to read an article, reading what they came for and then leaving.  This "bounce rate" is caused because the user has already investigated your website and only wanted to see what was new, so shouldn't necessarily be considered as negative statistics.
It is my opinion that a website should always try to engage a user.  So, to make the most out of your loyal (returning) visitors, you should try to include links to other new or related information in your website at the bottom of the page.  Make sure they know there is more to be read and you will see a significant reduction in your "bounce rate".
Gone are the days when your "unique visitor" statistics were proof enough of your website's success.  With advancements in web analytics, it is now important that website owners drill down and expose where your website is and is not working - providing you with the ability to make educated alterations to your online marketing.
Wendy Schollum is a web strategist and managing director of Xplore - your web agency (www.xplore.net).  If you would like more information on deciphering your web statistics, follow Xplore on Twitter (www.twitter.com/xploreNET), join us on Facebook (www.facebook.com/xploreNET) or call 0800 100 900.