18 Comments

Summary:

Your hunches about the Twitter population are probably correct. A new study by an analytics firm of 36 million users confirms some assumptions but also present some intriguing facts about ordinary Twitter users.

Twitter’s endless funnel of data is producing an equally endless series of studies — including on Twitter itself. The latest study offers fun tidbits about who is using Twitter and why.

In “An Exhaustive Study of Twitter Users Around the World,” analytics firm Beevolve claims to have crunched data from 36 million Twitter profiles. What they found probably corresponds to what you suspected: most Twitter users are young iPhone users from English speaking countries.

Fond of gender stereotypes? The study has you covered: women tweet more than men; gals tweet about family and fashion while guys tweet about tech and sports; women like purple backgrounds while men prefer dark ones:

The most useful part of the study, however, is that it provides a good view of how ordinary people use Twitter. For instance, it reveals that 25 percent of Twitter users have never tweeted, the average number of followers is 208 and that 81 percent of users have fewer than 50 followers:

This is a good reminder that most people don’t use Twitter in the same way as those of us in the media-politics-tech-celebrity-sports bubble. It also appears to confirm BuzzFeed’s John Herman’s theory that “Your Twitter followers aren’t fake, they’re just shy.”

It will be interesting to see how Twitter reaches out to this passive population going forward. I’ve tried to persuade family and friends that Twitter is simply a great news service but they’re skeptical. They think, understandably, that Twitter is a club for loud mouths and ask me, “what would I tweet?”

As for the study, Beevolve CEO Goldee Udani said by email that the firm was “curious about the demographic make up of Twitter user base, topics that Twitter users found interesting at a macro level and how dense was the Twitter social graph compared to known statistics about Facebook.”

You can see the full study, which is presented in clear language and graphics, here.

Feature photo courtesy of Shutterstock user Andresr

  1. This article has a really misleading title, and is a great example of why one shouldn’t make user segmentation assumptions based on averages.

    If you look at the chart in the article, the vast majority of people (81.1%) have less than 50 followers, so we can’t really say a “typical” twitter user is a female with 208 followers. So the title, though enticing, is inaccurate.

    1. Its beehive’s fault — the report makes this exact claim AND sets it up for a reader to auto-tweet as a factoid, promoting its source, of course. Its tempting to make a rhetorical disparaging remark about that source.

    2. beevolve, not beehive. please excuse.

      1. Thanks Scott and David for weighing in. Yes, the numbers are taken from beevovle; yes, perhaps it would have been more accurate to have used the median (which was not provided) which would have put the number closer to 50 than 200.

  2. … and thinks I give a sh#t what she had for lunch

  3. QUOTE: “yes, perhaps it would have been more accurate to have used the median ”

    Having an editor look over article or fact checking or understanding statistics would have been more accurate. Do not make excuse for bad reporting. You will do nothing for your reputation.

    1. Jeff John Roberts Rick Thursday, October 11, 2012

      Rick, not sure how helpful your comment is.. It comes across as more officious and patronizing than instructive.. Note too that the story include a link to the original source where people can explore the methodology themselves

  4. Thanks heavens! This means a 47yo Gay man with over 1,000 followers is a-typical!

  5. 25% if the tweeter users never twetted because, they don’t know how to use, how to network with people

  6. does this mean I’m a stereotype? :(

  7. Tweeting is not a prerequisite to using Twitter. Some of us are simply followers.

  8. The distinction between the media-politics-tech-celebrity-sports bubble and the rest of us is interesting. I must admit that of the various social media, I’ve found Twitter the hardest to warm up to. Though last year I was a conference organizer and we invited a local newspaper columnist to join a panel on social networking. His key point was Facebook is dead and Twitter rules. As if to prove his point, he was tweeting non-stop the whole time he was on the panel. Then half way through he apologized and said “s*** happens, literally and you’ll read about it in tomorrow’s newspaper”. Then he jumped up and ran out the door. Turns out he was covering a story about a disaster at a sewage pumping station. Rather than being upset, the audience was absolutely fascinated.

  9. Bethan Jenkins Friday, October 12, 2012

    Has any such study been conducted about how those who use other languages use and access twitter? Its been suggested to me that those who use twitter via the medium of Welsh would not fit neatly in to these categories.

  10. I tweet, I have an iPhone, but that’s as far as the headline is accurate for me. Isn’t this just a load of rubbish like most statistics?

  11. Raymond Manders Saturday, October 13, 2012

    This is a good reminder that researches such as these are a complete waste of time and recourses.

  12. This “study” also (and more importantly imho) does not address the better stat of “Who is the typical twitter poster”.

    Analytics can get quite complex on this; Typical user with /n/ influence|retweets|views, longevity, acceleration of followers versus longevity – it’s rather hard to compartmentalize these data.

    (because I’d LOVE to publish a report, app, js, h5 or other tool) I’ve worked with statisticians around Chicago that are nothing but confused when trying to nail down who/what an average (or) typical (or) median (or) normative user would represent, statistically, and this is with collected data of over a billion ‘tweets’. (..not to mention metadata that users never see that developers would).

    I came here in “hopes” that someone could simplify/visualize some of these data in terms of interpreting data, but as with all studies so far on this topic, I’m left with dicke Eier haben, as it were.

    -j

  13. This is absolutely rubbish. these monitoring solutions are good enough for making a qualified guess and should not be making such misleading statements. How did they manage an article on gigaom?

  14. This has been observed in many online community platforms of varying types. Most people lurk and view the content without participating. Often they’ll register an account, but then chicken out when it comes to adding anything.

    In the OpenStreetMap community we’ve been puzzling over this for years. Why do so many people register on OpenStreetMap (the way to get access in order to edit the map), but then never edit the map? http://neis-one.org/2010/08/„nominal-members“-of-osm/

    As people involved in designing the system and trying to improve usability etc in order to grow the user base, this is frustrating, but it’s a well known phenomenon. Academics call it “participation inequality” http://www.useit.com/alertbox/participation_inequality.html

Comments have been disabled for this post