Web Metrics: Measuring Blog Popularity
Web site usage metrics is far more art than science, most numbers come with a healthy dose of bias, I think mostly due to demographics. Alexa will not give accurate numbers for sites which have a larger chunk of Firefox users since their data collecting toolbar only works for IE[1]. So you will get poorer numbers for sites that focus on technology, since they have a more savvy audience who tend to use Firefox. For example TechCrunch reported in April 2006 that 60% of their users use Firefox[2], when the most generous overall number I've seen is probably something like 25%.
Even looking at one single site for data you can get vastly different numbers based on which tool or algorithm is used to calculate see Kerry Watson's Browser Statistics.
RSS
RSS adds another tricky layer on top of web metrics. How many subscribers to an RSS feed are there? This sounds easy enough when dealing with individual clients, but when using a site which fetches and displays RSS such as Bloglines, My Yahoo, Google Personalized, etc... can you get an accurate count? Some of the services do report how many subscribers they have, others don't.
Another issue with RSS metrics is measuring who is reading not just subscribing. Bloglines record I subscribe to a blog and makes the content available to me on their site. What it does not report is if/when I may read those articles. Do you want to boost your reader numbers? just create a bunch of dummy bloglines accounts and subscribe to your feed. No change in pageviews but subscribers would look better.
See the following comparison which shows even for the very top blogs measuring popularity is tricky and the further down and more niche demographic/market the harder it is to measure. The list is of the top blogs listed by Hitwise and those top blogs compared to each other in Alexa and Technorati.
| Blog | Hitwise | Alexa | Technorati |
|---|---|---|---|
| huffingtonpost.com | 01 | 03 | 02 |
| dailykos.com | 02 | 04 | 03 |
| thesuperficial.com | 03 | 02 | 08 |
| engadget.com | 04 | 01 | 01 |
| postsecret.com | 05 | 10 | 04 |
| michellemalkin.com | 06 | 08 | 06 |
| crooksandliars.com | 07 | 07 | 05 |
| trent.blogspot.com | 08 | 09 | 10 |
| defamer.com | 09 | 06 | 09 |
| gawker.com | 10 | 05 | 07 |
The Technorati numbers are only when compared to the other blog, it is not their Technorati rank. For example, defamer.com and trent.blogspot.com are ranked #53 and #95 in Technorati.
A bit unrelated, I came across this cool site which shows how the top blogs link/relate to one another, check out Linkology
Sources
Technorati: http://technorati.com/pop/blogs/
Weblogs Hitwise: http://weblogs.hitwise.com/le...nd_gadge.html
Alexa: http://www.alexholic.com/ used to chart and compare
[1] http://www.alexa.com/site/company/technology
[2] http://www.techcrunch.com/2006/04/28/microsoft-blows-live-shopping-launch-no-firefox/