Skip to content

Understanding Website Statistics

Management loves statistics and will often rely on their website statistical program to give them the needed information on how well their website is performing. The data collected from these programs are assumed to be accurate. However, the reality is that no statistical program is perfect. Each tool makes different assumptions and will produce different results.

Let’s take a moment to learn what it all means.

Clarifying the Terminology

While many loosely equate hits with unique users, a “hit” really represents a single file request. For example, one page view may generate more than 10 hits.  The real focus should be on the number of visitors, and more importantly, the number of unique visitors. This will give you better insight as to how much traffic your site is seeing and if there is some stickiness (i.e., return visitors) to your traffic.

Types of Statistical Programs

Server-based Statistical Programs (Webalizer)

Every time something visits your website and requests a file (an HTML file, CSS file, JavaScript file, graphic file, Flash file, PDF document, etc), the web server records this information in your account’s log files. Things to keep in mind:

  • All requests are counted as traffic including search engine bots.
  • User agent strings offer minimal information and can be faked (to ensure sites do not block the browser).
  • Server-based programs cannot normally assess a visitor’s screen resolution or whether they had JavaScript and Flash enabled.
  • If a visitor is sharing the same IP address and or user agent string, the program will show that activity as the same visitor.
  • Server log files cannot record cached files from either your browser or ISP’s proxy servers. When you enter a web address, you may see files returned from that proxy server rather than the originating website. As your site increases in popularity, you could even experience a drop in file access as more proxy servers cache your site. This is important as it will lower your reported visitor traffic.
  • IP/user agents set a “visitor session” at different time intervals. For example:  access from the same IP/user agent with a 10 minute interval would reflect two sessions if a user who visits a page then waits 11 minutes before clicking a link elsewhere. But an application which assumed a 15 minute interval would only record only one visitor.

Client-side Data Collection and Analysis (Google Analytics)

Client-side data collection requires JavaScript code to be inserted on every page you want to analyze. Although many client-side programs produce informative reports, the data collection process is more volatile than server-side methods. Things to keep in mind:

  • Client-side programs collect more detailed information such as the time spent on a page, clicked links, the screen resolution, etc.
  • If users have JavaScript and cookies disabled or blocked, you will not be able to collect data from their activity. In general, you can expect around 5% of your traffic to not run JavaScript. Keep in mind that this will differ from site to site.
  • If another script on your page causes an error it could prevent data collection either in general or with just one browser.