Fog Creek Software
Discussion Board

Web application usage tracking tool

I don't know what's the appropriate title for this thread. But here is what I am looking for.
I have a web application which is used by around 1000 users on a daily basis. I am looking for a tool which will help me track the way in which the users use this web app. I would like to track, which links are being clicked by a user once he logs on, which pages he visits, how long does he stay at any given page etc. Basically the entire path until he logs out. Are there any tools which do this or do I have to develop one? I am running ASP application on MS IIS. Thanks for the help in advance.

Friday, August 20, 2004

Forgot to add, I would prefer if there are any free tools :)

Friday, August 20, 2004

standard webserver logs?  Especially if you add a token so you can link users up if several appear behind the same ip address otherwise.

i like i
Friday, August 20, 2004

Webtrends does this sort of thing. Costs a bundle though.

If free is your goal, then look at setting up a virtual site, where every path on that server is mapped to the same thing. Give each user a different web bug (1 pixel gif) to download. Give everyone a session ID, use the page url as part of the path or something. Some spammers also do this to track who opens their emails, so they can send you more spam.

User A shows up, they get sessionID 555555, they are at /default.asp. Plop on the bottom of the page an image tag looking for /555555/root/default/webbug.gif
Then they go to the /fun/whatsnew.asp page, in that case, the image tag is looking for /555555/fun/whatsnew/webbug.gif.
And so on.

Then you write something that parses the web logs for the server.

You will spend a lot of time writing this. Are you sure your time is not worth the several thousand dollars that you would pay for webtrends smart source data collector?

Friday, August 20, 2004

Aspect Oriented Programming (AOP) does this for debugging, but it can easily be applied to other areas of logging...

Our apps log whenever someone logs in and wer're considering doing it in a few other places too.

Friday, August 20, 2004

Analog ( is free and decent, though somewhat challenging to configure.

AWStats is also free -

Friday, August 20, 2004

Thank you all for the help so far. I don't think I can find a tool which does exactly I want. I am looking for more than a log analyzer. I want to follow the clicking path of a given user around my web app. I might have to develop something on my own if I don't want to pay lot of money to WebTrends or something like it.

Friday, August 20, 2004

well, in php there is the free

maybe you could look at the code.

"phpOpenTracker is a framework solution for the analysis of website traffic and visitor analysis.

It features a logging engine that, either invoked as a web-bug by a HTML <img> tag or embedded with two lines of code into your PHP application, logs each request to a website into a database. One installation of phpOpenTracker can track an arbitrary number of web-sites.

Through the phpOpenTracker API module you can easily access the gathered data and perform complex operations on it, like for instance the analysis of your visitors' clickpaths.

The analysis of clickpaths is essential when you want to measure the usability of your website. phpOpenTracker's API functions 'shortest_paths' and 'top_paths' help to answer questions like: Is the navigation intuitive enough? Do visitors find the shortest, most direct way from A to B? How are ads percepted? On what document do they leave the website -- and where to? The answers to these questions help you to improve the user experience and the way your website is percepted by your customers. "

Friday, August 20, 2004

all I can think of is using some (free) basic logs analyzers such as:

But those will only give you the "Big Picture" stats...

maybe you can write a script that follows a client's referrer page based on access time?

Friday, August 20, 2004

The big question is:

Are your URL's human readable?

Are we talking:

or are we talking

Of course, that whole ref=xxx thing makes me think obidos does some hefty tracking. This is where webtrends comes in. It can recognize "0553294385" as being irobot.
Saturday, August 21, 2004

*  Recent Topics

*  Fog Creek Home