"60,000 out of 73,000 file fownloads were from bots"
Web analytics based on web server log file data using Urchin 6 web analytics software provides some very useful points of information that Google Analytics and any other tag-based web analytics solution simply can't deliver. However, one of the biggest challenges with logfile based analytics is pollution of the data by "bots" - search engines robots, crawlers, spiders, scapers, etc... I've found that the data generated by non-human activity can easily account for 60% of the hits in your logfiles, and if you don't exclude it, your resulting reports built on your server logs can be off by that margin. Typical installations of popular logfile analysis tools like Webtrends, AW Stats, Webalizer, and even most Urchin installations won't exclude robotically-generated data by default.
