SocialSphere, Inc.

You are here:
Home Blogs John's Blog Twitter, Come Clean Please
Twitter, Come Clean Please
Written by John Della Volpe   
Sunday, 15 February 2009

I will buy dinner (or maybe the entire restaurant) at a restaurant of your choice for anyone who can prove to me that what I read from multiple sources, including Wikipedia and SlideShare, last week is true:

"Twitter has more than 55 million daily users who post short updates on anything from their latest date to their views on Congressional hearings."  

Impossible -- and not only is this number WILDLY off, it's probably off by a factor of between 10x and 50x.  I am not sure who's to blame for this huge mistake -- whether it's Twitter, their VCs, the Bloomberg reporter who wrote the story, or some optimistic analyst -- it just doesn't make sense -- and I think most people know this. 

Hopefully Twitter, like Facebook Stats, will help us out to give us the real story, but I am not holding my breath.  I was disappointed to search the Twitter blog under "stats" to find the most recent update was literally about a year ago. 

So therefore, here are three back of the napkin calculations that get us close to the real story -- which is still phenomenal by the way:

1)  Check out our friends at Compete.com -- for U.S. based inquiries like this, I think they're the most reliable.  According to their data, Twitter's monthly unique visitors have increased a whopping 800% in the last year, but they still chart monthly users at 5.97 million. I assume that this is mostly, if not all, US-data -- and if you assume Twitter still gets 40% of its traffic domestically that would extrapolate into about 15 million monthly unique users worldwide.  (NB w/ thanks to Jonathan and Robert:  Safe to add a few more to this total (not several million) to account for those who use sites like tweetdeck, twihrl, tweetfeed, facebook and other platforms to post their feeds)


2)  Look at Pew's recent survey -- and you will find that 11% of America's online adult population say that they have EVER used Twitter, and 4% reported doing so in the last day. Let's say that the online adult population in the U.S. is about 180 million -- 4% of 180 million is 7.2 million.  Keeping in mind that the Pew survey, which is excellent as always, has a margin of error of 2.8% -- even the most optimistic calculation looks much closer to Compete.com's number than the 55 million quoted in the Bloomberg story.

3)  Conduct the family and friends test. According to Facebook, they have about 175 million active users, of which 70% are outside the US. That means that there are about 52.5 million active Facebook users in the US, according to Facebook (NB.  I don't think the Facebook Factsheet has been updated in a few months at least).  Looking at the Compete.com graph, again, which is mostly US data, this is in the ballpark with their estimates. 

 


Ask yourself, do you have more friends on Facebook or Twitter? I know the answer -- and I am willing to be that the average person in America knows at least 10x as many people who are active on Facebook than tweet daily.

This post is not intended to be anti-Twitter.  While I think they should come clean with their real user base stats (isn't this the era of authenticity and transparency?) -- this message is intended for others who consume, analyze and are otherwise involved in the Web 2.0 and social media space today.  There's plenty of great stuff happening every day -- I cannot imagine us ever turning away -- it's best for everyone though if we maintain composure, understand that this space is in its infancy and will continue to grow and expand in very meaningful ways.  

And oh yeah, Bloomberg -- check the facts please.

Trackback(0)
Comments (6)add
...
written by Todd Van Hoosear , February 17, 2009
One final update: Bloomberg issued a correction after several emails with the reporter, who was VERY professional and responsive throughly the process: http://tinyurl.com/bk3bgf

I've subsequently updated the Twitter Wikipedia page to reflect the most recent accurate data from Compete. It now reads, in part,

"In November 2008, Jeremiah Owyang of Forrester Research estimated that Twitter had 4-5 million users. A February 2009 Compete.com blog entry ranks Twitter as the third largest social network (behind Facebook and MySpace), and puts the number of users at roughly 6 million and the number of monthly visitors at 55 million."

Full Wikipedia entry on Twitter: http://en.wikipedia.org/wiki/Twitter
report abuse
vote down
vote up
Votes: +0
...
written by Todd Van Hoosear , February 17, 2009
Now that I look into the numbers more, and look at a recent blog post from Compete.com (http://blog.compete.com/2009/02/09/facebook-myspace-twitter-social-network/), I think the reporter has the right number, but the wrong scale: it's 55 million MONTHLY visitors, NOT daily. Total number of users: 6 million.
report abuse
vote down
vote up
Votes: +0
...
written by Marcus , February 16, 2009
And yet the Bloomberg reporter stands by his numbers.
report abuse
vote down
vote up
Votes: +0
...
written by Todd Van Hoosear , February 16, 2009
Well, as I reported on the Wikipedia Twitter entry (http://en.wikipedia.org/wiki/Twitter), I exchanged emails with the Bloomberg reporter. His information was obtained from a source inside Institutional Venture Partners, one of the VCs behind the recent funding round. He stands by the 10x growth number that 55 million implies, saying that according to Institutional Venture, the daily Twitter audience is "roughly twice the size of the American Idol audience and 10X the size of CNN nightly viewers." American Idol averages over 30 million viewers according to several sources (e.g., http://www.realityblurred.com/...ut_ratings). So Bloomberg is standing by the number. As I mentioned to our own Jonathan, the numbers are internally valid, but I want them confirmed by a third-party, as the Compete numbers are WAY off...

Here's the link to the Wikipedia discussion page: http://tinyurl.com/cq5gwh
report abuse
vote down
vote up
Votes: +0
...
written by John , February 16, 2009
Good morning Robert,

Didn't mean to bust your chops on this -- your Tweetdeck, Facebook, Twihrl, etc. comment is definitely an excellent point. Would def boost the numbers. My gut is that this is a good case study of a reporter mishearing, or misremembering :) stats, using it in a story, not being fact checked because the real data is not available -- and then ending up on Wikipedia.
report abuse
vote down
vote up
Votes: +0
...
written by Robert Davis , February 16, 2009
Hey John - ya busted me on the slideshare deck, I was putting it together VERY early on Saturday morning and should have double-checked that number myself. There actually was a minute when I was going to go to Compete to do so, but I remember being concerned that Compete wouldn't track usage through the API. (I use TweetDeck for all posts, and almost never go to the .com site.)

Duly chastened at my haste, I will repost an edited deck at http://www.slideshare.net/thinkinc/whos-on-twitter

r.
report abuse
vote down
vote up
Votes: +1
Write comment
smaller | bigger

security image
Write the displayed characters


busy