Just as I said a few month ago, after being rather skeptical in the beginning, I turned out to be a Twitter-fan: many informations that would have passed my by before are now showing up in my Twitter-stream. I do agree that there are a lot of useless informations, too, though that is not really a problem thats uniqe to Twitter but more of one that shows up once you open your browser. Those of you who are following this blog regularly will have noticed, that I like to know the reasons for developments and backgroun informations. So it seemed just natural to take a look at data on Twitter.
Seeing how Twitter was sadly not able to send me a database-dumb on a USB-stick, I had to find another way to get a large enough data basis. Luckily, Twitter has a rather vast API, which sometimes even works, when the Sun and Moon align just right. When you sign up with Twitter, you receive an ascending numer (ID). At the moment this number is arround 85 million. I went ahead and wrote a little script that would ask for data from Twitter-accounts through the API and save them. I felt that 10% of all accounts should be sufficient to come to a convincing conclusion. This is just to give some background information on the data basis that was used for the following diagrams and evaluations, where the numbers were projected (if those values were absolute) for all accounts. If you want to use these diagrams, feel free to do so, though I do expect a note of origin.
Amount of Twitter-Accounts
Lets start out with the absolute number of Twitter-Accounts. Twitter started in March of 2006, the first account seems to be number 13, which belongs to the Twitter-founder Bizstone. After that, not much happened for a long time. Half way through 2008, Twitter had their breakthrough and gained a lot of traction:
Currently (meaning the end of October 2009), they just arrived at the 66 million-account-mark. Those of you who are curious why there are only 66 million accounts and not the 85 million that were mentioned earlier should consider that: theoretically, the User-ID will grow with ever user but there are large ‘gaps’ in this numbering. I could imagine that technical changes, which are associated with the exploding growth, are to blame for this.
Besides the absolute amount of accounts it is also interesting to see how many users are gained each month. The following diagram shows the number of new accounts per month:
Here we can see that Twitter is still growing rapidly but that this growth has stagnated over the last few month. It even looks like the opening of new accounts is slightly on the decline. Here it will be interesting do run the same analysis again in six month and see how the development goes.
Signing up is one thing, using the account is another. That this is especially true for Twitter shows this diagram of the amount of accounts that are actually actively used:
I marked those accounts as active, that have at least one Tweet and how are following at least one other account. If we were to, for example, move this threshold to incorporate only accounts with a minimum of 10 Tweets and more than five followers, then the accured conclusions would be even bleaker for Twitter.
And seeing how I can’t stand the sight of Exel anymore, we will finish this tomorrow 🙂