Coming from a background that includes print publications (magazines), the ability to gather data about a Web site’s use is so refreshing.

Back in the print days, an effort was often made to find out just what the readers really wanted, as well as give a balance of what the readers wanted vs. what they needed.

However, this was usually an anecdotal method – talking with readers at trade shows or on the phone (for whatever reason) and trying to get some sense of what’s working and what isn’t. Gives one an idea, but not exactly science.

Even the more empirical efforts – readers’ surveys – suffered from not one but two Achilles’ Heels:

• Small sample size, and
• People lie on any survey – intentionally or not. Get over it.

So, at best, we were shooting in the semi-dark over what readers wanted/needed. Which is frustrating.

With the Web, however, the reality is often the inverse: Via user comments, e-mails and – especially – server logs, there is a mass of data that clearly states what the user actually is seeking out. The trick, of course, is to cut through this mass of data to identify the nuggets.

This comes to mind because of the Top 10 lists I’ve added to the site. The referers are particularity fascinating:

• First off, the Top 10 area gets more referrer traffic on my site than any area besides the blog (after the blog proper, the Gallery section gets the most overall traffic).
• Big surprise: The top referrer – by a wide margin – is Google. Duh.
• The most frequently searched-for item that users hit my Top 10 list for is the author Robert Coover, whose Pricksongs and Descants is listed in the Short Story list. It’s most frequently a search for Coover and/or his story The Babysitter. I don’t know if it’s because not many people know Coover – so there are only a handful of places to find him compared to, say, John Updike, or if there is a greater interest in Coover than I’m aware of. Interesting.

Data diving. It’s not just for breakfast anymore.