Web "Hit Count" Skepticism

Although we've moved beyond the simple, and completely unreliable, basic Web hit counters of the mid-90s, there are still a lot of reasons to view hit counts with a solid dose of skepticism. If you're paying for hits, you should test your hit calculation methods, and compare multiple methods against each other. Even if you use a "traffic exchange" like EasyHits4U or FastFreeway and just use the free surfing credits, you're still paying with some attention and effort.

Your best bet is to collect stats in multiple ways, then compare the different numbers. (Trust me, they will be different!) Then make an educated guess, based on an averages, about the actual level of traffic you're seeing. This isn't a one-time effort, either. In addition to keeping your eye out for new ways to collect statistics, you should take time at regular intervals re-test the ones you're already using, to see how they're performing (and drop any that seem to be too far off). How often you do this will depend, of course, on how important accuracy is to you.

Photo credit: Ronnie Bergeron (ronnieb) / morgueFile.com

What better example to use for showing how you can't assume that hit statistics are correct, than this Lens? In the first week that this Lens existed, a hit-tracking link redirector reported that seven people came to this Lens by way of that system. Guess how many hits Squidoo reported coming from that system? Just one. Which number is correct? Probably neither -- the redirector probably counted hits that were either duplicate or where the visitor never made it to this Lens (perhaps due to network problems). And Squidoo may have reported a hit from the redirector as a "direct" hit, if the visitor's browser did not report the referring page.

So the main point here is that somewhere between 1 and 7 people reached this Lens via the redirector, but I'll never know exactly how many.