Faststats versus Google Analytics

Which is more reliable: Log Files or JavaScript Tags?

[Ben_Neake]

We've been comparing the results from Faststats (or Mach 5 Analyzer, as it's now known) and Google Analytics, and they are vastly different.

In short Faststats is showing 500-600 visitors a day for the past month, whilst Google Analytics is showing 250-300 visitors - roughly half the number of visitors - an alarming difference. Faststats claims that these are unique visitors, not hits.

Faststats is based solely on the server log files, so my question is which one should I believe?

[Receptional]

Well - I don't use Google Analytics unless pushed, but I've been a firm believer in the TAGGING based approach for years (see here for the debate, now a bit out of date) [webmasterworld.com].

[Altstatten]

As a rule, logfiles are best for tracking technical, server-performance information and tagging is the currently preferred, "best practices" method of collecting visitor clickstream data.

[mike73]

I just started using Analytics a few days ago. It seems to report about 1/10 of the traffic recorded by my own log files. It even shows less page views than Adsense reports, so I'm not really impressed.

It does have a nice GUI though, and I like the map feature. I'll probably keep using it just to keep an eye on trends, as they say.