accuracy of Google Urchin Traffic Stats vs 1&1

[killerbeame]

Hi,

Does anyone know the accuracy of the Google Urchin traffic stats?

On my personal project site - I installed the Urchin code on my website hosted by 1&1. I've had sites on 1&1 for several years. Based on conversions, I have always suspected their traffic statistics were fairly accurate. When traffic was up --so were sales.

Then I installed Urchin on my brand new site (22 days old)-- and the traffic stats are soooooo way off. They aren't even close. I know this is a new site.. but my deep dark fear is that these Urchin stats are right..

SEO has not been done on this particular site just yet.. so no laughing at my numbers. :) I've been relying heavily on traditional forms of marketing until I figure out how to use textpattern.

Here are a sample of the numbers I have been looking at

Wednesday
1&1 Stats: 84
Google : 11

Thursday
1&1 Stats: 93
Google : 14


What the heck? They are soo far off right? Does anyone have any ideas on this?

[optimalwebsite]

Are you sure your stats are measuring the same thing? It might be one's counting visitors, the other hits, or Urchin might only be couting hits to the homepage, while one and one include all hits.

People rank Urchin very highly, I can't imagine it being wrong.

[incrediblehelp]

Some software's look at the raw logs, some use JavaScript to track visitors. I have seen wide variance when using different software's. The key to true numbers to read your raw logs. I think Urchin/Google Analytics uses JavaScript tracking.

[killerbeame]

I double checked on my site and the Urchin tracking code is installed on almost all of my pages except a few minor like contact.

So I guess I will go by 1and1 stats.

[Conficio]

Hi there,
I'd guess the difference is in reading raw logs (1&1) or checking via Java Script (Google Analytics/Urchin).

The raw logs analysis counts all accesses to pages if they come from a robot or a human browser.

The JavaScript does by nature only count human browsers. Because most robots/crawlers do not interpret and execute Java Script (some do a slight of hand scan if they can find some links in it or some other obvious tricks).

Look at the raw logs and count the number of robot/crawler related requests. That should account for the difference.

I'd stay with Google Analytics. It does a much better job at analyzing the flow of visitors through your website and you can see if people reached your target pages (conversion). Most other packages don't offer that. However, have a look at your raw logs, for requests that fail. This is something that Google Analytics misses.

K<o>

[killerbeame]

thanks!!