Website stats. They have all the charm of 12th grade algebra. But they are very important in the world of online marketing. In this post, we’re going to look at a couple of web stats and what they might mean to you.
Annually, monthly, weekly, by the day or hour are the most common ways of looking at them. Any date range can be specified so you can have a look at your websites performance in large chunks or tiny slivers of time.
Two of the most important web stats are “sessions” and “pageviews.” Other very important stats are “bounce rate” and “length of session.” Most web statistics programs such as Google Analytics and the Urchin software it is based on can display web stats in a variety of ways. There are a variety of other web stats analysis applications that may already be on your web server.
A “session” is recorded every time a visitor arrives at your site for the first time since the expiration of their last session. Session expiration occurs when activity ceases for a time specified in your server configuration or when the visitor leaves your site.
The “sessions” graph shows the number of visitors your website had in a given time period. In the example graph below (from the logs of a server I use for testing), you can see I had a range of session numbers from 2 to 13.
A “pageview” occurs when a visitor enters a page on your website. A visitor can visit multiple pages during a single session. Pageviews are often mistakenly called “hits” which are a different stat entirely. Calling a “pageview” a “hit” is a lot like calling a Manufactured Home by the “T” word.
The pageviews stat tells you how many of your web pages are being looked at in the specified time. Now we’re ranging from 2 to 126. I like to see 3 or more pageviews per visit on my site – more is great, but less could simply be the result of having a site with few pages to look at. Here is the pageviews graph from my test server.
“Hits” are the most misleading and worthless stats on a web log. A “hit” is recorded every time a page element is loaded. So if you have a web page with 5 images, a link to a JavaScript and a link to a CSS stylesheet, the log will record 8 hits. Anyone can increase their “hit” count by inserting more graphics, even single pixel transparent images.
When you hear someone describe how many “hits” their website is getting, it’s a good bet you can divide that by a factor of 10 to 25. Ask them how many pageviews their site gets. If they can’t answer that, as Lyndon Johnson used to say, “I put my hand on my wallet.” NEVER give advertising money to a website based on the number of “hits” they get. If they use the improper term, it could be an honest mistake (some manufactured home pros still use ‘trailer’), but they should know the difference and should be willing to share the correct data.
Here is a “hits” graph of my test server for the same period as the other two charts. Notice the difference, especially in the Saturday and Monday stats. Do you see how inflated the “hits” stats are? 1,171 hits for 85 pageviews? That’s a ratio of 14 hits per pageview on a site with minimal graphics.
“Bounce rate” is determined by the number of visitors who leave your site after viewing only one page. They can “bounce” for any number of reasons (for more, see the blog post What is Bounce Rate? What does it mean to you? on BobStovall.com. A decent bounce rate could be anywhere from 33% to 67%. Some one-page sales sites have a bounce rate of 100% and do OK. But if you have a multi-page informational site and you have an unusually high bounce rate (67%-95%) you need to have a look at your site and see what is driving visitors away so quickly.
“Length of Session” is another “tell-tale” stat. It tells you how long the average visitor is spending looking at your web pages per session. It might seem odd, but 2-3 minutes is a pretty good session length. if 66% of your visitors are leaving immediately, the other 33% are staying for 6-9 minutes and that is quite respectable.
There are a lot of other things your web logs can tell you, such as what percentage of your visitors are using Macs or Windows, how many are visiting using Firefox, Google Chrome, Internet Explorer or Safari, what pages they are entering on, what pages they are leaving from and more.
We’ll delve into more web stats in the future, but I think you now have a good basic understanding of which stats to keep an eye on and what to look for.