Hi.. I dunno what to think right now about GA . I made a comparative test between my stats with GA and some others. One I installed using a SQL database, and 2 others using code like GA. The results are really impressive. In all case, I got 2x more Unique visitors with ALL the other stats than GA have. I mean if I have 450 Unique visitors with GA, with all the others i've arround 900 unique visitors. Something si goin' wrong, and i talk on forums about this and people tell me the same thing. Is someone can explain me what is the reason.. is exist an explication about that. I think that sounds really strange because damn.. thats a Google thing, thats not scam.. its suppose to be quality things. There are the stats i'm testting right now giving me good effective stats.: http://www.histats.com http://www.Piwik.org (this is the new version of PHPMyVisites) <- Really not bad http://www.statcounter.com/
They aren't all counting the same thing. Google Analytics counts pageviews. Not hits to the servers. So unlike some ways of tracking, it's not counting bots that don't call up the page. It also evaluates unique and returning vistors by dropping cookies, rather than by I.P - in my opinion this is much more accurate as it doesn't re-count dynamic IPs, such as generated by AOL. (Look in your statcounter logs for how it treats visitors that come via AOL - it's massively inflating the figures because every time the IP changes, it treats it like a different visitor, though if you look in the visitor journey data, it's clearly not - GA wouldn't do this) For my twopennorth - for these reasons I would trust G.A. figures. I know people always want to believe the larger figure, and suspect that Google is scamming them in some way (even though it's a FREE SERVICE) - but no - it's just filtering out more of the chaff.
Ok i undertsand, but when i look the logs for all stats services, I dont have a lot of AOL connections, not enough to make that big difference.
Yeah I know what you mean by AOL IPs, because they always change IPs. But there's a problem with cookies too.. some computers dont accept cookies. Anyways, i just want to find a reason why thats about double for the others. This is a good start, but its sure thats not the only reason So you sure that GA is more precise than others??