Website stats vs Index pg counter stats

Divergent stats

[insyte1]

The site stats on referrals provided by my website host differ in a major way from stats provided by counters on my index page.
Besides search engine referrals, I was particularly keen to assess the amount of referrals from my own theme sites to my product-sales site.

I put advertisements on my theme sites all pointing to the index page of my product-sales site, and then I put a counter (2 actually) on that index page to measure the incoming referrals.

The stats provided by my website host show one theme site referring massively, but the counters register virtually nothing coming from that site even though the advertisement links point only to the index page where the counters are.

And conversely, the counters show two of my theme sites as major referrers, but the site stats show zip coming from these theme sites.

I contacted the website host who basically said "Our stats are accurate - talk to the counter people". So not much help there.

I need to know what theme sites are working for me and which ones aren't, but these conflicting stats leave me confused.

Insyte1

[Chris_R]

Your website stats are PROBABLY more accurate - as they probably come from your logs.

The counters work off an image tag which takes longer and can be inaccurate depnding oon how it is done.

There is one solution - use AXS.

I can't tell you how good of a program this is. It is EASY to install (as easy as any perl script out there) and gives you oodles of info.

It is shareware.

[xav.com...]

try out their demo.

[xav.com...]

You don't have to install anything to use their demo. Your host - should install it for you relatively cheaply if you feel like you can't.

It takes about 15 minutes if you know perl - 5 minutes if you have installed it before.

[Mikkel Svendsen]

You will always see large differences in numbers tracked with different methods. Trackers only track page views - and only if they are viewed (not if they fail). Log files only track what the webserver "see" - so e.g. not cached pages. On network level you get something even different.

So to say "Our stats are ccurate" is a funny statement. Accurate according to what standard? Accurate according to what filtering and referrers tracking?

You bascially have at least 3 completely different methods of data collection, an even higher number of filtering and data mining options and endless ways of analysing and reporting the numbers. With so many variables it's pure luck if you get the same stats from using two different methods :)

[insyte1]

I'm not so much concerned about exact numbers as with the fact that the website log files show such a large stream of referrals from one of my other sites (more than Google), but the index page counters show virtually nil - even tho the links from that other site all point to the index page where the counters are.

And these same counters ARE picking up referrals from several other sites of mine which use the exact same links.

I'm really baffled.

Insyte1

[MHes]

%B<
?

I suspect your pages are getting cached by proxy servers. Change the link to your main site on your theme sites to www.mysite.com/?id=12345

check the link works!

Then monitor your raw logs on your main site.... this may stop the isp's serving a cached version of your main site, and instead draw on your own server... hence the referral will show on your logs.

Counters are very dodgy, I would believe the raw logs with this 'fake' dynamic url request.

good luck

[insyte1]

OK MHES, I've changed the links to www.mysite.com/?id=12345. They DO connect properly, so we'll give it a day and see if it makes a difference.

Insyte1