Unless Twitter has released how many users they have and I'm not aware of it, then how many users they have is really up for speculation. And there is a lot of speculation going on.
I've read that Twitterholic has guestimated about 12 million users while Twitdir has just over a million in their database. If you do as I did and accidentally click the Google search this site button from the Google Firefox toolbar while on your Twitter home page, you'll get back 3,770,000 users.
I did some simple math based on when I joined and when a friend joined. He joined just about six months later than me. I joined in early May of 2007 and he joined in early December of 2007. According to Twitterholic I was user 5,870,022 while my friend was user 11,197,712. So in the six months between us there about 5 1/2 million new user accounts. That was in December of 2007, and we're now just about May of 2008. So how many news users have they added in the last 5 months? Well you would have to figure at least another 5 million bringing the total to around 16 million. But remember this is based on Twitterholic data. How accurate is it really? I don't know. And how does this mesh with the 3.7 million accounts Google has?
Right now my search company Hyperix is crawling Twitter. in fact we just switched yesterday from using our discovery web crawler to using our big daddy web crawler as we've got enough user accounts now to start the big crawl. At some point in the next couple of months we'll have an application ready to go and we'll release our numbers. In the meantime, does anyone want to speculate further?
Update @ 3:34 PM PST
What's the coincidence that Techcrunch would publish a story on the same topic today. According to their numbers from an insider at the company Twitter has about 200,000 active users per week and 1+ million users as of March.
That's very interesting in that would mean that Twitdir has nearly all of the user accounts in their dabatase. And it also means that our initial discovery crawl did very well.
Update @ 8:09 AM PST, April 30, 2008
One user brought up an interesting point. Every time a new account is created they get an xml feed with an ID number in it. Mine is number 5,870,022 which matches the number that Twitterholic defines as the nth person to sign up. Last week I created a new account which as the ID 14,461,298. This suggests that if Twitter is using these id's in sequentially that there have been nearly 14.5 million accounts created. However, what % of those accounts are actually active? That's still an open question.
Well, I did a little homework myself, and found the following:
11,200,000 is the number Google gives me when I change the search preferences in Google for just Spanish and English pages (I don't know if this is a bug with Google, but in the past it has given me both larger and smaller numbers). Changing to other languages has no effect on this search number.
Asking Yahoo gives me 6,573,494
So, I'm thinking the twitterholic is probably right.
Also just checked twitdir and searched myself (motorbikematt) and found two accounts
The primary one I use (@motorbikematt) and one that I had inadvertently setup via a test of the IM function, (@motorbikematt1). I do consider myself a fairly capable webuser and this mistake was an easy one to make. Given that I cannot logon to this account since it was added via IM, I also cannot delete this account. Therefore I think it's a safe assumption that while there may be a maximum of 12 million accounts, there is most certainly a lower number that are relevantly active.
What do you consider to be a user? Do you mean all accounts or ones runs by 'people?' I mean, are @NASA or @NASAWatch accounts considered users also, though they're just RSS feeds converted by Twitterfeed into Twitter?
I'm curious about how many accounts are owned by regular (say, once per week) individual users.
I'm referring to all accounts including those RSS feeds. However I would also like to know how many users are "regular" users. With our data I think we'll be able to post that number at some point.
Looks like TechCrunch was thinking the same thing as you, Mark: http://www.techcrunch.com/2008/04/29/end-of-speculation-the-real-twitter-usage-numbers/