Forum Moderators: DixonJones
I know, I know.. log files are best for traffic analysis. While, I agree with this, sometimes I just want a quick estimate of how many visitors one of my websites receives on a particular day.
For the past year or two, I have been using Awstats to answer this question. I always choose Awstats over Webalizer because it seems to give a more accurate reflection of actual traffic. The question is, I don't understand why.
It seems like Webalizer always reports about HALF as many visitors as Awstats. I don't understand why the discrepancy is so large. For example, assume I download my log files, and view my stats w/ webtrends or a comparable log analyzer. If this reports 2,000 visitors for a particular day, then the exact same day Awstats will ALSO report approximately 2,000 visitors, BUT Webalizer will only report 800 to 1100 visitors. Can anyone tell me why this is?
There is an even bigger disparity between the keyphrase referrals reported by Awstats and Webalizer. For example, Awstats may report 1,000 instances of a particular keyphrase per month, whereas Webalizer will only report 100 to 150.
Can anyone shed any light on this?
I have wreslted(sp) with the same problems too.
How do the two define visitors? And are you talking about unqiue or returning visitors?
Also, it maybe that they both parse log files differently.
How do the two define visitors? And are you talking about unqiue or returning visitors?
I am referring to total visitors, not "unique visitors". However, this still does not explain the difference, because my unique's are not less than half of my total visitors. Not even close.
Also, it maybe that they both parse log files differently.
I don't understand why this would make any difference. The same raw data should be parsed out of the log file. Man, I don't understand this problem.
Use any of the paid systems and you will again get different answers.
I don't know these two in particular, but I'll put my money on one of them having a huge number of users from Virginia (AOL users). That's the one that you should dump if that's the case.
