Twitter: a sober look at the statistics

While some were surprised at the astronomical growth of the Twitter audience (2,565%, yes, two thousand five hundred and sixty-five percent of unique users per year), others decided to approach the data critically.
Matthew Hurst, a Microsoft MSN researcher and creator of the BlogPulse service, conducted the following improvised research on his blog . Using the Twitter search for requests, he looked at how many people who were looking for something a month ago are “active” at the moment. By “activity” he meant updating the status in the last 5 days. And what data he got: out of the 15 first accounts on request 6 (40%) were not updated for a fortnight; 9 (60%) have not been updated in the last 24 hours.
Then he did another experiment - he looked at the profiles of the first and last 20 “followers” ​​of Obama, assuming that the lists of followers are sorted in reverse chronological order. Among the “recruits” 50% have never been updated. “Old Guard” in this regard did not differ much from the first and second twenty followers in the last 5 days five (20%) were not updated, and from the third ten all the same 50%.
The statistics for Ashton Kutcher (aplusk), which recently set a record of 2 million followers, calculated according to the same principle was as follows: 13 out of 20 followers from a fifty thousand page page were inactive (65%). In Britney Spears (britneyspears), this figure reaches 75%. Thus, for fans to see soap bubbles in all fashion trends, the first arguments appeared. But that is not all.
The main evangelists of Twitter on the open spaces of the English-speaking Internet - Mashable super- blog - a week after the post, Matthew also began to doubt their shrine. Based on data from Nielsen Online , they reported that 60% of new tweeter converts stop all activity in a month. However, as it turned out later, users of clients such as Seesmic Desktop and Twhirl were not included in the Nielsen report. It is also worth noting that even Facebook and MySpace had twice the return percentage in the same “takeoff” period. In this regard, the authors of Mashalbe were forced to assume that Twitter is, by its nature, closer to Blogger than MySpace or Facebook.
And just now the first substantial studies of the audience of this service began to appear. I will tell about them in the next post.