Cookieless tracking

Question

I'm working on university final year's project which includes website visitor tracking (not conversion tracking so server-to-server postback tracking isn't of my interest). Obviously it's dead simple to implement using cookies, however as a fallback I'd like to implement cookieless tracking. I know it's not reliable, but the articles I've read were quite outdated so maybe things have changed and there's some clever way to do it.

I understand that the only possible footprint that could be used is a combination of UA(I could store browser update information because the UA header won't be matched if it's been updated so I could somehow figure out that the same visitor just updated his browser) and IP (and timestamps maybe? if so how? what patterns I should be looking at?), but I'm starting to look into client-side data caching which could be sent to the server using AJAX or some other alternative.

What I'm looking for are ideas and tips that could help me built this sort of tracking system.

I'll be storing each visitors profile in a database so that I could analyse each visitor individually. So on page view, I'm going to lookup that visitor and if he already exists I'll continue to store data linked to him.

Thanks for any ideas.

Answer 1

I implemented a basic tracking based on IP address, but as it collects all the other information it can (browser, OS, screen resolution, colour depth, flash version, etc) so it's definitely possible to define a user by a collection of these data points.

I tend to store the raw data and then use queries after to perform the analytics to make it faster, but it depends whether or not you need to store other information that you actually use throughout that users visit.