A few days ago a Technorati watchlist for Quova turned up an article by The Register about their November 2005 circulation numbers. While The Register announced that they crossed the 4m unique threshold for the month (as audited by ABCe), I was more excited to see them quote their distribution of readers by country according to Quova.
While this is a very simplistic use Quova’s technology, it made me think back to the mid-1990’s when I was on the netGenesis and I/Pro boards and I first pondered the analog analog of magazine circulation as applied to the web. At the time, this was really hard data to pull together with any accuracy, especially at the geography level, and I remember talking about different approaches beyond simply parsing HTTP_REFERED to get the country code in the URL.
At the time, we weren’t even thinking about content localization or fraud detection. I was just excited by the pretty pie charts that showed distribution of visitors by country. We’ve come a long way in a decade.