Forum Moderators: DixonJones
I'm trying to explain to my colleagues why we only have referer information for about half of our "human" visitors and I realize that I don't actually know the answer!
I screen out user agents that are most likely to be spiders/robots, etc, but I know there are surely some UA fakers that sneak through. Nevertheless, I still have about 50% of unique visitors who leave no referer on their first hit on my site and just a "-" mark is recorded. Interestingly, that number has declined steadily from almost 80% 12 months ago.
I am guessing that some folks use browsers or browser add-ons that hide this info from me, which I don't mind. But what are some of the other reasons why UAs don't leave any referer? Bookmarks? Email links? Type-ins?
I'm also wondering if others among you have similar rates of no referers. Is 50% high? Low? Is it a sign I'm not screening robots well enough?
Thanks in advance to any insight into this phenomenon.
But the firewall scrubbing makes a lot of sense. We have a large portion of our audience coming from government, and I suspect they are more likely to be behind firewalls that would leave these columns in my log blank. I'll have to do a cursory check on some of their IPs to see if there's a pattern.
I'm definitely not sure I'm screening out all robots, but I try to catch the most significant fakers by screening out any UAs that have a lot of hits in a short time, i.e. spider behavior. FYI, for anyone who's interested, my latest list of robot user agent strings (804 total) is in this text file [johnmccrory.com].