Google Analytics Not Accurate?

[incrediblehelp]

Is anyone seeing the Google Analytics stats as being not accurate or in-complete? I am running Webalizer along with Google Analytics and the numbers are not even close.

[Serious]

Webanalyzer is quite optimistic. Could you give more info? Which particular numbers differ?

[greeneagle]

Google Analytics may be somewhat like Alexa, analyzing their own "channels" only... that makes perfect sense. What would make us think they had access to the other search channels for mensuration?

Ken

[incrediblehelp]

I am looking just at total SE hits and keywords referred. I don't see how it could be "optimistic" in these numbers.

Also Ken if the tracking code is on all your pages then it should have stats from all channels and search engines just like the Webalizer stats which is looking directly at your raw logs.

[dburdon]

I'm stickign with Statcounter. At least the people owning that business don't appear to want to take over the world.

If I want to take a look at Google Adwords I can go there. If I want to take a look at Overture I can go there too.

I just wouldn't want one company with all the info.

[brian.mark]

I am looking just at total SE hits and keywords referred. I don't see how it could be "optimistic" in these numbers.

Also Ken if the tracking code is on all your pages then it should have stats from all channels and search engines just like the Webalizer stats which is looking directly at your raw logs.

JavaScript turned off, spiders and other non-browser visitors, and a few others don't get counted by Google Analytics. Add in the optimistic part and you get more of the picture.

It also takes a bit of time viewing a page to register over at the big G. Depending on load, it could take quite a few seconds or error out. Logfiles don't have that delay. They don't even care if you ever get the page on the other end... you request it, it was viewed.

Brian.

[incrediblehelp]

yup, thanks Brian

[incrediblehelp]

I still would think a tool as powerful as Urchin would allow you to account for any and all traffic in your raw logs. I have never really used Urchin on my own websites just browsed it on client websites. I am sure the version that installs on your actual server is much more robust than the free Google version.

[brian.mark]

The analytics program we use includes a noscript tag with a clear 1x1 pixel image to track those without JavaScript, which this free Google version doesn't do. None of the tagging style accounts for non-browser visitors (spiders, bots, and other crawling programs.) Some count another pageview when someone hits "back", some do not.

Page tagging really has some inconsistencies in the way that they work, and most don't really say what they do. None of them account for spiders. Those can only be seen in logs.

Google's says the least of any I've ever seen. I'm sure it'll get better documentation as it progresses, but for now it's just an indication of where your visitors are coming from.

Brian.

[AccuraCast]

I use StatCounter and Google Analytics hand in hand. In fact, we'd been using Urchin long before it became GA, and noted some of the same problems back then too. The way I see it:

StatCounter (SC) provides excellent information down to an individual level
Google Analytics (GA) provides a much better overview for large volumes of data over time.

SC has live tracking
GA has some inexplicable delays which they dont explain the cause of, and cant seem to even pinpoint the exact times - i still cant figure out whether the hits i see are according to GMT or PMT or whatever!

GA integrates almost seamlessly with AdWords (albeit with the before-mentioned time delay)

GA provides the ability to track conversions, referals etc from other campaigns, by adding the tracking URL fields to your ordinary URLs.