I will shortly be publishing the new list of top analyst blogs. This list has been compiled from Tekrati’s excellent blog directory. This league table will take into account new analyst hires and firms as well as a tweaked methodology.
I have taken the feedback I have received from the previous research and modified the methodology.
Scores are now calculated as follows:
Google PageRank – Google PageRank is a link analysis algorithm that interprets web links and assigns a numerical weighting (0 to 10) to each site. High-quality sites receive a higher PageRank. The ranking uses the actual PageRank as part of its algorithm.
Yahoo Inbound Links [date unlimited] – Yahoo counts the total number of inbound links that go directly to a blog. Each number was assigned to a range which was then used as part of the algorithm.
Google Inbound Links [3 months date limited] – Google allows people to search the number of inbound links to a specific blog but limit this to a predefined date period. Similar to how Technorati only looks at six months of data, this method was used in combination with the Yahoo Inbound Link count to assess which blogs were considered to be important due to the number of links that came to them, but also currently relevant as measured by the limitations on the timescale. Each number was assigned to a range which was then used as part of the algorithm.
Google Reader Subscribers – Google reader lists the total number of subscribers to a blog. I believe this is a more realistic number to that which Bloglines provides. Mihai Parparita confirms that “these numbers include subscribers across all Google services”. To account for people using other readers (e.g. Newsgator) it has been suggested that this number is multiplied by 3. Subscriber ranges were determined (i.e. more than 20, more than 30, etc.) and each range was assigned a number that was used as part of the algorithm.
Frequency of Posts – Updating relevant and interesting content frequently onto a blog will naturally cause more people to find this blog important. This score is established via Google Reader to understand the precise number of posts per week that the blogger makes. Frequency numbers were determined and assigned to a range that was used as part of the algorithm.
Date Last Blog Post Published – Working in combination with ‘Frequency of Posts’, this score mitigates against blogs that were once popular but haven’t been updated for a long time. The number of days since the last blog post was calculated and assigned to a range which was used as part of the algorithm.
Comments – A simple way to judge how valuable a blog is to other people is through the number of comments (where this is enabled) that visitors make. In a similar way to linking and subscribing this user requested service shows a significant value. The number of comments made over the last five posts were calculated and assigned to a range which was used as part of the algorithm.
Twitter Inbound Links – There are various online tools available to count the number of links inbound to a blog from Twitter. Backtype was used to count the number of these occurrences over the past five blog posts. The number of times this happened was calculated and assigned a range which was used as part of the algorithm.
Weighting – Each specific variable listed above was given a standard score out of 10. Using a weighting scale I varied the importance of the each metric to establish a blogs total score.