Server side tracking, client side tracking.

dif?

[mmmtweak]

I have been working on getting better use of my stats as of late. For my sites I use webtrends for log analysis. This helps determin where people are coming from and what words I am being found for. For my clients I have been installing Hitslink on some of their pages to let them track the effectiveness of their seo and marketing work.

The one question is this. Some of my competition keeps throwing around the terms "client side tracking" and "server side tracking" Now I need to know the difference ie. is client side = hitslink and server side = webtrends? or And why is one better than the other when trying to figure out real usage and details.?

[figment88]

webtrendslive is client side, other more common versions of webtrends are server side.

Definitions:
server stats: most web servers are configured to maintain an access log that notes every object served (html files, image files, javascript files, css files, etc.). Along with the fact that the object was served, usually indicated are a timestamp, ip address requesting, user agent requesting, and other data. Web server stat programs take this info and arrange it into easy to read reports.

client side stats: most client side stat programs require code insertion into every web page that triggers the recording of access data on some server. This code is usually a JavaScript include with a paramterized image tag placed in a noscript tag. Some of the monitoring services such as Comscore or netratings also do client side data collection but they use some type of software package (server proxy, bowser plugin).

intermediate stats: some stats programs also rely on data collected between the client and server. For example, webtrends enterprise can collect data at the outgoing port. This is useful for globally distributed server farms and other complex interactions.

advantages:
This is a lengthy subject but
server side advantages:
1) cheap
2) easy -- usually already installed
3) more standard for comparing across sites

client side advantages:
1) more accurate. people are able to browse cached web elements without downloading them from the server. This is a huge prblem for server stat packages.
2) can get additional types of data. Basically anything that can be read by javascript (cookie info, screen dimensions, javascript version, browser language) can be set back to the collection server

[mmmtweak]

Can you please explain?
"people are able to browse cached web elements without downloading them from the server"

also lets say my competitor tells clients to use client side tracking.

If I am loading Hitslink on their sites am I solving that issue?

[Receptional]

"people are able to browse cached web elements without downloading them from the server"

Many people can read your web page without even requesting data from your website. This is because (say) an AOL user requested a page 5 minutes ago, then another AOL user requests the same page. AOL says "hey, it I already have this in my memory, no need to go find it again"... so... no record in the log file that webtrends looks at.

lets say my competitor tells clients to use client side tracking.
If I am loading Hitslink on their sites am I solving that issue?

Yep. (and I agree with your competitor BTW - use a client side tracking system, which includes Hitslink, Hitbox or Wentrendslive) but this offers other problems... if the client (that's the web user) doesn't accept any cookies, data may be a bit scarce. But for most people it is much more accurate IMHO.

Really clever people combine both... Webtrends and Deepmetrix have a cookie plug in for IIS. But they both crashed our server when we tried it!

Dixon.