Twitter: a sober look at statistics. Part II

In a previous post, we looked at the first doubts about the popularity of Twitter. Here we continue this (un) noble cause, with more specific numbers in hand. So.
A new blow to the statistics of the popular service was done by the study Participatory Marketing Network . According to their data, only 22% of the “Y” generation (18-24 years old) are using Twitter.. Despite the fact that 99% of young people are registered in traditional social services such as Facebook and MySpace! Moreover, 85% of users in this category noted that they “follow” their friends and only a third follow their company accounts. By the way, I myself recently asked an audience of 20 students, did they hear anything about Twitter. Only two raised their hands. Just to say that they learned about him from me. I believe that the 10% difference can be explained by the non-representativeness of the sample by Russia's overall lag in modern communications =).
Next came the study.Harvard Business School conducted on a random sample of 300,000 accounts. His main discovery can be considered the atypicality of Twitter in terms of gender separation. So, if in classic “social networks”, such as Facebook and MySpace , the center of attention and communication is a woman, then on Twitter this is a man. That is, despite the fact that 55% of users are women , men have 15% more followers . Moreover, the likelihood that an “average” woman will follow a man 25% more than a woman. For men, this ratio is 40%.
Another fundamental difference between Twitter and other social services is the Pareto distribution of the function of the number of messages from the number of users, which is more reminiscent of Wikipedia than Facebook and MySpace . In particular, on Twitter, 90% of posts are written by 10% of users . On Wikipedia, this ratio is 90/15, and for the “traditional” social service - 10/30. These distributions are clearly shown in the following graph:


Thus, we see that comparing Twitter with Blogger.com is not so strange.
As for the topic that interests us, here the statistics are also not very comforting. The average number of tweets for May is 1, i.e.half of users write updates less than once every 74 days . More emission-resistant statistics - the median - give a result of 4, which is also not very much.
Such results can upset anyone, and Twitter is no exception. This resulted in a halt in the explosive growth in the number of users (with a 7% increase in traffic). The annual increase in users of 3,712% and an increase in viewing time by 699% are in this case a weak compensation.
And so, the latest study “the state of twittersphere” ( pdf ) by Hubspot seemed to confirm alarming data. In a sample of 4.5 million accounts , the following data was received:

• 79.79% did not enter their site URL
• 75.86% did not indicate their biography in the profile
• 68.68% did not indicate their location
• 55.50% do not follow anyone
• 54.88% have no updates
• 52.71% do not have any followers

It would seem that it could be sadder? However, a couple of the findings of this study are still encouraging.
The first of these is to redefine the concept of “inactive users”. Traditionally, one who writes nothing is considered to be such. However, everyone knows the “lurkers" of traditional online communities, who, although they do not create content themselves, are also an integral part of the audience. Therefore, the authors of the study introduce a new definition of "inactivity". To be considered as such, the account must meet all three of the conditions presented and have:

• less than 10 followers;
• less than 10 friends;
• less than 10 updates.

And such, according to the data, there are only 10% .
The second encouraging finding of the Hubspot report is the increase in message length. According to the authors, it indicates that users are gradually getting used to the service and strive to fit their messages into the allotted 140 characters. This is graphically presented in the following diagram:

Thus, at the moment, we can safely say only one thing - Twitter is not a classic social service like Facebook and MySpace . Accordingly, it requires a fundamentally different approach to work and research.