|Huge diferences with Google Analytics|
Any idea on how to solve this?
Hi webmasters we are on a difficult stats conflict. Our local stats are fine but Google Analytics is showing almost 20% of the visits.
The company I work for runs several websites, I'm in charge of the design and updates of a few (but just that, the administration of the server is in other hands). The sites rank just fine and we have several Ks of visits per day.
The marketing department insist on using GA as a traffic and reporting tool of choice to show the graphs to the clients, but GA shows only a fraction of the traffic. I once used GA on MY SITES and the difference was very little, not this huge. I researched about this in the past and found some reports of little differences, kinda acceptable, and also found forum posts stating the same as me: huge difference but no solution.
Any idea on why? how to solve this? any explanation? consider that it's a BIG company (burocracy), the admin part is not in my hands. I moved the GA code up the page to give it priority on execution but so far the same.
Our server logs are fine, no errors of access. The sad part is being a BIG company the server is behind whatever infrastructure and firewalls, not always *fine*. We even had config problems in the past.
I remember there was a local Google Analytics option, I guess it is Urchin, but it costs heavy money.
Gosh... I know that after explaining the complications and limitations there is little to do, but perhaps somebody had this experience before.
Any comment will be welcome.
I feel your pain. I deal with clients who look at Google Analytics in a similar way. Only when you deal with Analytics programs from multiple companies you find that no one Analytics platform gives you a complete picture of what's actually going on.
Most tracking pixels are designed to benefit the creator of that pixel.
I'm gathering this is a chronic problem?
I can't help you, but I've noticed today that Analytics stats are way behind on my articles posted on a third party site. As in, they're at 50% of my AdSense numbers. There are sometimes delays, but this is a pretty big one for me.
@CPC: true, I agree 100% with that. Too sad clients like fancy graphs and want to stay with that.
@Lapizuli: the problem derivates from being a big company where too many hands are involved. In the past the servers wrongly configurated and visiting any of the child sites (domains) take per example site B, C, D, E, without the www took visitors to the main site... being A, terrible. But that's one of the problems of big companies (even if you work there and know how to fix it).
That was finally fixed, but the stats problem continues. I guess that's the only clue I can give: the domains are now resolving well with and without www's but the stats are still wrong. Sounds familiar?
The data quality is obviously only as good as the setup.
Also, when talking about discrepancies between 2 different styles of data analysis you have to make sure you are comparing apples to apples.
You mentioned that the sites experience several thousand visits a day.
What constitutes a visit in this statistic? How do you define a genuine visitor vs. a bot?
My advise, really narrow down your comparison. Something like day-by-day breakdown of 7 days worth of unique visitors to a single page, from a specific browser.
Personally, I have had no issues with using GA in enterprise environments.