Sat 1 Apr 2006
How many Williams students, or even people on the Williams campus, read EphBlog? I am not sure. Being transparent, we provide links to two different hit counters (Extreme Tracking and Site Meter) along the right hand side bar. I suspect that there are ways to use these tools to figure out what percentage of our visitors come from Williams, but I am not smart enough to do so.
Fortunately, the natural experiment of Spring Break allows us to compare our hits between periods when students are on campus and periods when they are not. It seems that, in the last two weeks, our readership is down around 100-150 per day. Since the posts are of approximately the same quality (?), I infer that about 100 students who read EphBlog while they are on campus do not read it over Spring Break.
The College, by the way, has similar tools but refuses to make the data public. When I pursued the matter — after all, finding out which pages on the williams.edu domain are most visited would be interesting — I was told that the College was afraid that the information would be “too confusing.” Whatever.
At some point, the College will realize that the more open and transparent it is as an organization, the more successful it will be in the long run. Allowing the rest of us to know which (public!) web pages are most visited would be a good place to start.
April 1st, 2006 at 12:42 pm
Site Meter shows the domains people came from (http://www.sitemeter.com/?a=stats&s=s18ephblog&r=8), so with a spreadsheet you could figure out the percentage of visitors from Williams.
While knowing which pages people visit most might be interesting from a factoid point-of-view, sometimes it’s embarrassing — for example, visits can go down after a Web site makeover, if the designers go for glitz and not for easy navigation.
I believe the college doesn’t really track its Web site traffic in any comprehensive fashion, or if it does, the reports are read by several analysts in IT and that’s it. So I think “confusing” is a code word for, “Geez, I wonder who runs those reports, and if I can get them, what they’ll show? Better to obfuscate than admit we don’t really have a handle on it.”
That said, I agree with the college that there’s not a lot to be gained by releasing raw numbers. At this point in the discipline, that’s a simplistic, late 90s thing to do. The real reason for doing Web analytics is to track where people go after they get to a page, how long they stay, where they get confused. You’re gathering the numbers so over time you can figure out how to improve the site.
As a destination site, the college isn’t doing too bad: Alexa ranks it at 38,273, slightly ahead of Amherst. (Details here: http://www.alexa.com/data/details/?url=www.williams.edu)
April 1st, 2006 at 6:36 pm
The domain tracking shows that 7% of the past 100 visitors (i.e., seven visitors) were from williams.edu. If you look at this every so often, you can find a general idea of how many of the visitors tend to be from williams.edu.