Forum Moderators: phranque
what is the bad side effects of doing this really?
the good one must obviously be that you get trustable stats.
ill just use webalizer then.
One problem I've seen with sites that use free stats tools comes when they depend on calling out to a third party server for a script or an image (and they often do.) The page can just hang and hang and never finish loading when the thrid party server is overloaded.
And then there's the obvious fact that you are sharing your stats. Hmmm...not always what you want to be doing.
What are "trustable" stats is a major question in its own right. Unless you know the methodology behind crunching the stats, you only ever have a rough idea and not all that trustable. Take the same set of server logs, run them through two stats packages (even very high-end expensive analysis) and the end reports will not be the same.
For instance, how do these services work with bots, especially "rogue" bots? I've seen server logs where 70% of the traffic is not a human being - and some where it's more like 15%. I doubt that any free stats service works very hard to get close to a number for "real people".
this thread is interesting
Which Stats Package is right for you? [webmasterworld.com]
no surprise that I agree with Ted, I am not a supporter of sharing my logs with anyone for any reason. I also suggest to most people when they want to understand logging and site traffic to dig into and read their raw logs. It is difficult to understand various interpretations of data when you don't truly understand the data it is interpretting.
