Twitter’s real number is closer to 60 million

comscore-twitter-mar09

Twitter’s worldwide traffic growth just keeps going. TechCrunch reports the latest stats, comScore’s traffic data at the end of March:

[…] Worldwide visitors to Twitter.com increased 95 percent in the month of March from 9.8 million to 19.1 million, according to [comScore’s] estimates. This compares to 9.3 million visitors in the U.S. alone.

These numbers only count visitors to Twitter’s Website, which is not the same as active users and also does not include people who interact with Twitter via desktop or mobile clients (a large portion of users). But the comScore numbers provide a good proxy for Twitter’s overall growth.

The bold emphasis is mine.

I reckon it’s three times that number, so closer to 60 million visits if you take into account some assumptions about people who use third-party applications on their computers (TweetDeck, Twhirl, Seesmic Desktop and others) and mobile devices (Tweetie on the iPhone, Twibble and Gravity on Nokia and other S60 devices, etc) to interact with Twitter, all of which use API calls to twitter.com to retrieve content as well as send your outbound tweets, @s and DMs.

Traffic to the Twitter website, in other words. (And remember: this isn’t the number of users we’re talking about, but the aggregate number of visits to the Twitter website.)

What’s the basis for my number? Nothing scientific, just some back-of-the-envelope thinking, a feeling from looking around at all the activity combined with something Hitwise’s Robin Goad said a few weeks ago in a comment on an earlier post:

This question of how much Twitter traffic comes via 3rd party apps is tricky one, but I did hear somewhere that it could be up to twice as much as comes through the web. If that is the case, crudely multiplying Twitter’s market share by 3 to compensate for this would make it around the 20th most visited website in the UK – neck and neck with iPlayer but still lightly behind MySpace.

So projecting the traffic number the way I’ve done doesn’t seem too outlandish.

Like John Gruber writing yesterday about how Twitter clients are a user interface design playground, one of the most interesting aspects to me in the amount of attention Twitter’s getting and it growth explosion, in users as well as traffic, is a similar explosion of Twitter API client software.

How does the overall picture look to you?

Related posts:

Neville Hobson

Social Strategist, Communicator, Writer, and Podcaster with a curiosity for tech and how people use it. Believer in an Internet for everyone. Early adopter (and leaver) and experimenter with social media. Occasional test pilot of shiny new objects. Avid tea drinker.

  7. James Cridland

    Except.

    To something measuring traffic (a faux proxy, as Comscore or Neilsen use, or a complete net over ISP traffic as Hitwise use), there’s little difference between:
    http://www.twitter.com/jamescridland (my Twitter page) and
    http://twitter.com/users/show/jamescridland.xml (a typical API call to my data via XML).

    Whether it’s Firefox, Internet Explorer or Tweetdeck doing it, it’s still a standard HTTP request (aka “web page”). So, it’s hard to see why these figures wouldn’t include apps like this.

    (What is clear is that all three companies – Comscore, Neilsen and Hitwise – are poor at measuring at-work usage, which – to me at least – appears to show that they are almost entirely pointless as a decent gauge of traffic.)

  8. neville

    […] it’s still a standard HTTP request (aka “web page”). So, it’s hard to see why these figures wouldn’t include apps like this.

    That’s my thought too, James. How hard can it really be? Clearly it must be hard or surely such metrics would be in the stats that have been published so far.

Comments are closed.
Close