Services and insight for building
effective, user-oriented web sites.
Understanding web site traffic reports
Web traffic reports can be hard to understand. If you don't know the difference between a hit, a page, or a visit, the report is of little use.
However, by learning the meanings of a few words, these reports can become much more useful, and not nearly so hard to understand.
In this article, I will cover the following:
- Definitions of common web traffic report terms
- Where the information in the reports comes from
- Some web reporting software
- And some additional reading
Some definitions
- Hit
- A request from a web client (Internet Explorer, Netscape, etc.) for a file on a web server.
- File
- Indicates a Hit that resulted in a successfully downloaded file. The number of Files will always be lower than the number of Hits. In a perfect world, they would be equal.
- Page
- A page is one complete web page, most likely composed of a number of Files. For instance, if your home page has five images and some text, there will be six Hits (and hopefully Files) that make up a single Page.
- Visit
- When a user visits your site, the user often clicks to a few more pages on
your site. The period during which the user looks at those pages is
called
a Visit. And, if the user leaves your site, but comes back within a few minutes,
it still usually counts as just one visit.
Note: With a little math, you can identify some interesting numbers like Pages per Visit and Files per Page. - User agent
- User agent, or client, is geek-speak for a browser or other program that accesses your web site. This information can help you identify what browsers people are using to view your site, and it also identifies many search engine agents that visit your site.
- Site or Host
- This term is not to be confused with the a web host. Rather counter-intuitively, this term actually refers to the computers that access the web site. These are often identified by IP numbers (these look like 64.23.144.5) or the name of the computer itself. So, if you are using a computer named JOSIAH and you are logged into a domain called IMADOMAIN.COM, your computer might show up in the log file as JOSIAH.IMADOMAIN.COM. If your visitors are using proxy servers, the proxy server might show up here instead.
- Search terms
- Often, when someone gets to your site through a search engine result, the log file will record what words they searched for to find your site. This can be very helpful information, especially if you are trying to measure how effective certain keywords are for you.
- Usernames
- If you are running an Intranet, or you have a section of your site that clients log in to, the Usernames list will actually show which usernames were used to access parts of your web site. If you don’t have any such areas, this section should remain blank.
- Errors
- Web logs should also record errors, and these are identified by different number codes. We have all seen the 404 Error, File Not Found on some web sites we’ve tried to get to. That information is in the log as well. And, yes, if you have too many of these, you should be concerned.
So where do these reports come from?
Web pages are kept on web servers. When a person wants to look at a web page, they type in the web address, or URL, into the address bar of their web browser (Internet Explorer, for instance). That URL points to a specific web server, and then to the specific web page. If all is well, the web server sends the requested page back to the person’s web browser.
That exchange is called a "client-server transaction." As a general rule, web servers record every transaction in text files known as log files.
Now, log files are really ugly to look at, and are guaranteed to cause headaches in all but the geekist of geeks. Fortunately, there is software that looks at the log files for us, and shows it to use in a much more user-friendly fashion, complete with tables and colorful charts.
Where do you get web reporting software?
There are many options for log analysis programs, and if you pay for web hosting, your web host will most likely provide a program for little or no extra charge.
Some common products are:
- WebTrends -- Good, but sometimes pricey. Usually installed by your web host.
- Webalizer -- Also good, and actually free. Usually installed by your web host.
- ClickTracks -- Shows a different type of information: the path that users take through your site. This is desktop software.
Additional reading
If you are looking at your web logs, read these articles by Wil Reynolds posted at MarketingProfs.com:
Send me a note about this web tip or suggest a new topic.
Contact me to discuss how I can help you produce a more valuable user-experience for your web audience.
--
Davin Granroth