Do you hear a ringing sound? As I suggested earlier, the iPhone launch is going to be a case study vehicle for social media anlaysts. Hey presto - BuzzMetrics put out a press release and will follow up with a web cast. In addition, they will be offering a syndicated report.
Interestingly, while the press release includes a BlogPulse trend graph, it fails to support the bullet point factoids with any data. For example, it describes the visual voicemail and the music library as being hotly anticipated features. Here's my graph:
My prediction is that chatter in the blogosphere about the iPhone will hit 1% tomorrow.
Update: Relevant Noise is also blogging about this from an analytics point of view, Max (BuzzMetrics), Pete (BuzzMetrics).
Is there anything to differentiate between discussions of these individual features and their inclusion in a list of all iphone features?
Posted by: FD | June 28, 2007 at 10:46 PM
I am surprised to notice that BlogPulse graphs are so often used in posts commenting trends in the Blogosphere. I have to admit that I am also guilty ...
On the other hand I have noticed that the choices BlogPulse makes in choosing which blogs to index are rather abitrary (ignoring a lot of other blogs). I have doubts concerning the completeness and the quality of the BlogPulse graphs. The graphs can show a trend, but I would not attach much importance to the percentages shown.
Unfortunatly there is no better public source of reliable information for graphically representing trends in the blogosphere, which explains why a lot of bloggers continue to refer to BlogPulse graphs.
Posted by: Bruno | June 29, 2007 at 05:23 AM
Bruno,
While I agree that one needs to be careful with the interpretation of results of this sort, I'd like to know more about your concerns with BlogPulse. Given that the results are percentages, and that BlogPulse covers a reasonably high percentage of the blog posts out there (and that its quality is better for those posts than Technorati), what would you like to see changed to make the results more reliable? What percentage recall would you require to make the percentage based graph useful?
It is one thing to grumble about coverage (and quite a different thing to grumble when one has been in the trenches so to speak). But you are complaining about the reliability of a metric which is designed to help remove the user from the coverage issue.
If you are correct, then you must be claiming that there is - in this case - a large portion of the blogosphere out there that is significantly contributing to the mass of non-iPhone posts. By your estimates, how big is that mass? Alternatively, your complaints could be interpreted as suggesting that there is a large mass of spam posts in the index about the iPhone which we need to ignore - is that the case?
Posted by: Matthew Hurst | June 29, 2007 at 03:30 PM
Matt,
Agreed...some of the analysis in our press release did not have back-up data in the actual press release. The assumptions, however, were backed up, but the press release was a story angle and a tease to let people know where to go for much more in-depth supporting data. That's what our Webcast, press interviews and client briefings and reports were for! :-)
Bruno,
Matt is one of the brainchilds of BlogPulse, as well as one of the smartest people in social-media analytics. While Matt no longer works at Nielsen BuzzMetrics, I'd be happy to forward your concerns or questions to the team that manages BlogPulse now.
Regards,
Max
Posted by: Max Kalehoff | June 30, 2007 at 09:15 AM
My remarks are based on my own experiences. I have the impression that Blogpulse is not doing so good in indexing non-English blogs.
I have submitted my blog (http://bvlg.blogspot.com) in Dutch several times, but it still does not get indexed by Blogpulse - although my blog is considered by Metatale as one of the most influential blogs in Belgium (information unfortunately only in Dutch at http://metatale.eu/top-100).
Recently there was quite some buzz in Belgium caused by someone who started the website "medium4you" where he copied without asking blog posts from a lot of blogs. Technorati is able to find 42 blog posts for medium4you, Google Blogsearch does a little bit better with 53 blog posts, Blogpulse only finds 14 posts.
Another example. One of the most popular marketing blogs in the Netherlands is called marketingfacts. Technorati finds 1848 blog posts containing the word marketingfacts, Blogpulse finds 668 occurences.
I do not have any guarantee at all that the blogs indexed by BlogPulse are representative.
Posted by: Bruno (BVLG) | June 30, 2007 at 06:17 PM