We’ve gotten quite a bit of feedback from Healthcare 100 bloggers after adding some new metrics and features to our algorithm this month. Based on this feedback, we’ve now made adjustments to Technorati, Bloglines and Alexa in the algorithm, decreasing the influence of Technorati and Alexa and increasing Bloglines (although it is still a smaller factor than it was in the original Healthcare 100 algo.)
We made these adjustments based on remarkably consistent feedback we were getting from people across all positions in the Healthcare 100 rankings. Essentially, we became concerned that our inclusion of Alexa, while a valuable indicator of Web traffic, was overweighting the list in favor big domains (wsj.com, scienceblogs.com, etc.)
Meanwhile, our downgrading of Bloglines seemed to disproportionately hurt independent blogs — some of which had large and loyal subscriber followings. As flawed as Bloglines is, it provides the only publicly available API for measuring blog subscribers.
We also agreed with the criticism that we had overweighted Technorati in the new algorithm. What we love about Technorati is that it SHOULD be the go-to source for measuring and comparing blog influence. It’s precisely what they’re set up to do.
Unfortunately, Technorati has had a number of issues and failures over the past year or two that have really pecked away at its authority in the eyes of many bloggers. So we decided we needed to reflect this by better balancing Technorati’s influence relative to Google, Yahoo, Alexa and Bloglines.
Finally, a word or two about proposed changes we didn’t make. We didn’t increase the influence of Google PageRank in the algorithm — a big reason being that Google PageRank, like Alexa, is biased toward big domains. We also aren’t crazy about Google’s black-box methodology; we like to know what’s being measured by one of our metrics, and how it’s being measured.
We also were asked to do a lot of things that we simply can’t — like include Feedburner stats, or StatCounter or SiteMeter metrics, etc. If the data isn’t available via API, and if the metric can’t be applied to all 900+ blogs in our system, we can’t incorporate it in the Healthcare 100. (However, if you know of an interesting metric with an API that is not included in the Healthcare 100, please let us know.)
So there you have it. This stuff is hard work! But we hope we’ve created a list that serves as a valuable directory of health and medical blogs based on their relative influence in the blogosphere. Thanks to everyone for your input.141