Liz Gannes

Recent Posts by Liz Gannes

Twitter Numbers: Cool, But How Many Users Does It Actually Have?

Twitter today published a variety of impressive stats about its growth on the occasion of its fifth anniversary; for instance, it now has 140 million tweets per day, up from 50 million a year ago, and 460,000 accounts created per day.

But what the company didn’t give is a very standard stat for any Web service: How many users it has.

Twitter has for years avoided sharing its number of active users (often defined as people who visit at least once a month). At times it has said how many total registered accounts it has; I believe the last number was “about 200 million,” from an email sent to all users in early February.

It’s true that Twitter is more of a publishing platform than other social Web services, and that many members of its audience are readers and celebrity followers rather than tweeters. Many of its visitors don’t have registered accounts. And much of the messaging service’s distribution is out of its control, on cable news broadcasts and the like.

But the company can indeed measure its usage, and better than third-party stats providers, who don’t have access to all its platforms. By giving out its own specialized stats–like the very cute tweets per second (TPS reports!)–it’s hard to understand Twitter in the context of the rest of the Web.

But! There are now a billion tweets sent per week!

Image via Flickr user Bill Moser.


comments so far. Add yours.

  • http://twitter.com/McProulx Mike Proulx

    Liz, it’s a fair question but at the same time, is the total number of users the “holy grail” metric or is it around how engaged the userbase is (regardless of its size). The people within my social circle on Twitter are heavily engaged and, for me, as an individual that’s what matters.

  • http://twitter.com/davzimak Dav Zimak

    To monetize off of ads, twitter will be concerned with how many times a pair of eyeballs ends up looking at a promoted tweet. So, I imagine the relevant number should be something like pageviews, but include also syndicated list views to third party sites where then can insert a promoted tweet, as well as client views where they can hit the users with an ad. Twitter clients probably account for many of these views, so I can see why they don’t want to rev share on that in the future.

    Along that line of thinking, if they tell the number of pageviews in that way, they would pretty much give away their revenue ceiling (pv/1000*cpm).

  • Anonymous

    That really does make a lot of sense dude.

    http://www.anon-tools.es.tc

  • http://networkeffect.allthingsd.com Liz Gannes

    Ultimately engagement is more important but it’s even more difficult
    to measure. And engagement is a subset of activity–you have to start
    somewhere.

  • http://networkeffect.allthingsd.com Liz Gannes

    You’re right–pvs are increasingly important as twitter becomes more
    of a media company.

  • http://twitter.com/tweetysez TweetySez

    Wouldn’t they also have to count bots??;)

  • http://networkeffect.allthingsd.com Liz Gannes

    Bots are people too!

    But yes, they are not exactly like every other social network or publisher. Still, they are not from an alien planet.

Latest Video

View all videos »

Search »

I break down a product the same way I break down a character I’m going to play. I try to get inside the mind of that person — the user, the consumer — and figure out why they’re doing something and what they want from it.

— Ashton Kutcher’s investing philosophy