I've been wanting to calibrate my Bloglines-based analyses of RSS subscriptions [1, 2] with other sources of data. Feedburner came first to mind, and Richard McManus had the same thought:
Last week I wrote about my blog's subscription growth, inspired by Jon Udell's analysis of Bloglines Public Subscribers. In a subsequent email discussion with Jon Udell, we both wondered whether Feedburner data would show similar trends to the Bloglines data. Jon suggested a collaborative project could be launched in the blogosphere with the aim of calibrating the Bloglines data. Of course I thought this was a great idea, so I contacted Dick Costolo and Eric Lunt from Feedburner to ask if they'd be willing to help. Dick and Eric were very enthusiastic about the idea and Eric soon whipped up a couple of templates for us, accessible via RESTish URLs. [Read/Write Web: Collaborative Feedburner Stats Project]See Richard's whole item for the details.
We're not sure what we'll learn from this, but it seems worth a try. Apart from what the data may or may not tell us about subscription trends in the blogosphere, I'm curious about the collaboration itself. Typically a survey will collect your data and, maybe, show you aggregate results. The idea here is to do things in a decentralized and transparent way. Owners of data who are willing to release it1 post XML packets in an agreed-upon format, using an agreed-upon metadata tag. Nobody owns the aggregate data, anybody can collect and analyze it.
Could be interesting, we'll see how it pans out.
1 As Richard explains, if you'd rather anonymize your data you can still participate by sending it to him directly.
Former URL: http://weblog.infoworld.com/udell/2005/03/09.html#a1193