Where did Forrester get its Twitter data?

Peter Kim of Forrester writes on his blog “Our data shows that 6% of US online adults use Twitter regularly.”

I say bulls**t.

There is ABSOLUTELY NO WAY that many people are using Twitter.

My data shows that the regular users are between 50,000 and 300,000. A high percentage of which are outside the United States. That doesn’t come anywhere close to the numbers required for 6%.

Keep in mind that Hotmail has about 200 million users every month. Yahoo Mail says they have about 250 million worldwide users.

But, I’d love to be proved wrong. Where did this data come from? How was it collected? Does Forrester stand behind it? What’s in the report that Peter linked to (I am not a Forrester client, so don’t have access)? Does it contain other numbers that just don’t jibe with common experience?

UPDATE: Someone just Twittered me this: “Peter Kims’s source on the unique users (he says 447,000 in Aug07) is Nielsen//NetRatings.” I doubt that’s data for “regular” users, or even online adult users in the US. I could see total registered users being that high, but that’d be world-wide. Watch twittervision.com someday and you’ll see that there are lots of users outside America.

UPDATE 2: Peter Kim responded here, and says they didn’t get the data from Nielsen. I still think the survey is very flawed if it’s bringing back such numbers.

Comments

  1. “there are lots of users outside America.” Actually 65% of everything is outside America. And that number is rising.

    But if your view of the web is via Google and you live in the USA or UK, you’d be surprised because Google does a grand job of hiding that 65%. And if all you follow is the big analyst and media companies and the tech blogs, you’d be forgiven for thinking that there’s nothing much outside the Valley, let alone California. Does the rest of the world even exist?

  2. “there are lots of users outside America.” Actually 65% of everything is outside America. And that number is rising.

    But if your view of the web is via Google and you live in the USA or UK, you’d be surprised because Google does a grand job of hiding that 65%. And if all you follow is the big analyst and media companies and the tech blogs, you’d be forgiven for thinking that there’s nothing much outside the Valley, let alone California. Does the rest of the world even exist?

  3. As a Forrester Employee, I’m reading the data, and it comes from Forrester’s own Technographics Research, so yeah, we stand behind it, I’ll let Peter give more details on this report.

    Robert, how many bloggers are out there in North America? (I know it’s hard to track) Is it too much to suggest that a percentage of them are trying micro/mobile blogging? Not really.

    Ironically, a screenshot of twitter in the report, and your tweets are all over it.

    In the spirit of being responsive, just letting you all know we see these questions.

  4. As a Forrester Employee, I’m reading the data, and it comes from Forrester’s own Technographics Research, so yeah, we stand behind it, I’ll let Peter give more details on this report.

    Robert, how many bloggers are out there in North America? (I know it’s hard to track) Is it too much to suggest that a percentage of them are trying micro/mobile blogging? Not really.

    Ironically, a screenshot of twitter in the report, and your tweets are all over it.

    In the spirit of being responsive, just letting you all know we see these questions.

  5. Reading the quote carefully, neverness seems to be right. Peter says that 6% of the ‘online population’, not ‘resident population’ are using Twitter regularly.

  6. Reading the quote carefully, neverness seems to be right. Peter says that 6% of the ‘online population’, not ‘resident population’ are using Twitter regularly.

  7. I doubt 6% of online Americans have even heard of twitter, let alone use it regularly.

    I believe that 6% of Americans living in 415, 408, 650 and 510 area codes who have jobs with technology companies use twitter regularly. Twitter has essentially zero market penetration outside of the digerati.

    Of course I’m sure there’s a set of assumptions at the beginning of the report. I’m sure they have the caveat that the report’s accuracy is contingent on those assumptions. I’m also sure that some (or most) of the assumptions they use are laughable.

    Speaking of laughs. Go get one of those crazy expensive reports from 10 years ago and read it. The accuracy of these research companies is pathetic.

  8. I doubt 6% of online Americans have even heard of twitter, let alone use it regularly.

    I believe that 6% of Americans living in 415, 408, 650 and 510 area codes who have jobs with technology companies use twitter regularly. Twitter has essentially zero market penetration outside of the digerati.

    Of course I’m sure there’s a set of assumptions at the beginning of the report. I’m sure they have the caveat that the report’s accuracy is contingent on those assumptions. I’m also sure that some (or most) of the assumptions they use are laughable.

    Speaking of laughs. Go get one of those crazy expensive reports from 10 years ago and read it. The accuracy of these research companies is pathetic.

  9. Hi – just an acknowledgement, will respond later. I don’t stay up all night like you and Jeremiah, so I just woke up. :) Need to make coffee and catch a train to work, will post a reply about methodology so your readers can stop speculating.

  10. Hi – just an acknowledgement, will respond later. I don’t stay up all night like you and Jeremiah, so I just woke up. :) Need to make coffee and catch a train to work, will post a reply about methodology so your readers can stop speculating.

  11. I reckon 0.25% if you’re generous and assume 500,000 U.S. Twitter users out of 200 million online adults.

  12. I reckon 0.25% if you’re generous and assume 500,000 U.S. Twitter users out of 200 million online adults.

  13. Are Hotmail and Yahoo mail stats really users or are they accounts? What are the stats on users who have multiple accounts? Is there good data on how many accounts are inactive? abandoned?

  14. Are Hotmail and Yahoo mail stats really users or are they accounts? What are the stats on users who have multiple accounts? Is there good data on how many accounts are inactive? abandoned?

  15. As with the Jericho campaign, caused by Nielsen and their antiquated system of ratings and research gathering, here we see them blow it again.
    I am heavily involved in Jericho, and the campaign to save and now keep the show…… Nielsen was a major reason for our whole problem. They never took into account the users for DVR, and based TV ratings off the Box system, now proven to be a huge issue for the Networks.
    Now we get to see more of their faulty services, their lack luster performance to do the job.

  16. As with the Jericho campaign, caused by Nielsen and their antiquated system of ratings and research gathering, here we see them blow it again.
    I am heavily involved in Jericho, and the campaign to save and now keep the show…… Nielsen was a major reason for our whole problem. They never took into account the users for DVR, and based TV ratings off the Box system, now proven to be a huge issue for the Networks.
    Now we get to see more of their faulty services, their lack luster performance to do the job.

  17. So we all agree that 6% is not possible. Dare we ask what they list as the number of US adults online?

    From a scientist standpoint it’d be nice to get some explanation of their metrics to determine the source of [possible] error.

  18. So we all agree that 6% is not possible. Dare we ask what they list as the number of US adults online?

    From a scientist standpoint it’d be nice to get some explanation of their metrics to determine the source of [possible] error.

  19. Funny, I left you a comment around 3:21 am your time that seems to have disappeared. Finally into the office now (which you might have noticed if you’re following me on Twitter) and will post soon. RE: the Nielsen stat on your update – that’s not where I drew the 6% figure from.

  20. Funny, I left you a comment around 3:21 am your time that seems to have disappeared. Finally into the office now (which you might have noticed if you’re following me on Twitter) and will post soon. RE: the Nielsen stat on your update – that’s not where I drew the 6% figure from.

  21. Agreed–more like .06%. It is hard enough to get people outside of the tech or marketing industry to understand what a blog is, let alone micro-blogging like Twitter.

  22. Agreed–more like .06%. It is hard enough to get people outside of the tech or marketing industry to understand what a blog is, let alone micro-blogging like Twitter.

  23. Erm, yeah, sure. 6% of online Americans (137 million according to Pew Research Center, October 2007) would be 8.22 million. There’s no way Twitter has remotely near 8 million regular users.

    This isn’t the first time Forrester has published some outrageous stats. Recalling a few while back, they had published another preposterous statistic only to be forced to backtrack with numerous changes and footnotes after being called out to explain the methodology.

  24. Erm, yeah, sure. 6% of online Americans (137 million according to Pew Research Center, October 2007) would be 8.22 million. There’s no way Twitter has remotely near 8 million regular users.

    This isn’t the first time Forrester has published some outrageous stats. Recalling a few while back, they had published another preposterous statistic only to be forced to backtrack with numerous changes and footnotes after being called out to explain the methodology.

  25. The funny thing is 6% of online adult Americans is usually the low-end estimate of the number who blog. It ranges between 6-8%. Right statistic, wrong label?

  26. The funny thing is 6% of online adult Americans is usually the low-end estimate of the number who blog. It ranges between 6-8%. Right statistic, wrong label?

  27. Peter Kim: sorry, for new commenters my moderation system grabs it. I’d love to see the thinking behind this statistic. It doesn’t match what Twitter employees are telling me.

  28. Peter Kim: sorry, for new commenters my moderation system grabs it. I’d love to see the thinking behind this statistic. It doesn’t match what Twitter employees are telling me.

  29. Guys, please notice what they are claiming.
    “Our data shows that 6% of US online adults use Twitter regularly.”

    They are not saying that 6% of Americans or 6% of American adults are on Twitter.

    Question I have for Forrester:

    1. Define “online adult.” We need an age demographic and what makes them “online”

    Then we can comment as to how full of shit they are when they make that statement.

  30. Guys, please notice what they are claiming.
    “Our data shows that 6% of US online adults use Twitter regularly.”

    They are not saying that 6% of Americans or 6% of American adults are on Twitter.

    Question I have for Forrester:

    1. Define “online adult.” We need an age demographic and what makes them “online”

    Then we can comment as to how full of shit they are when they make that statement.

  31. Study was performed online: 5,223 individuals.

    The sample was drawn from members of MarketTools’ online panel, and respondents were motivated by a sweepstakes drawing. The sample provided by MarketTools is not a random sample.

    ———-

    So all in all, not very useful since you don’t know where MarketTools gets their user base from.

    You could equally conclude that 100% of U.S. adults online are interested in winning an online sweepstakes, based on the same results.

  32. Study was performed online: 5,223 individuals.

    The sample was drawn from members of MarketTools’ online panel, and respondents were motivated by a sweepstakes drawing. The sample provided by MarketTools is not a random sample.

    ———-

    So all in all, not very useful since you don’t know where MarketTools gets their user base from.

    You could equally conclude that 100% of U.S. adults online are interested in winning an online sweepstakes, based on the same results.

  33. Typical analyst make-up numbers, and double-whammy as here you have to start with defining “usage”.

    “We define usage by at least monthly or more” with “at least some portion of it, in some way” in a “sheer pass-along effect” that’s wholly “indirect”.

    Gawd, what a mouthful of pure nonsense.

    Soooo if within the month you have maybe glanced at a Twitter, “in some way” or maybe even just heard someone talking about it, indirectly, in a “sheer pass-along” way, from our handpicked unscientific sweepstakes list, then you are a “Twitter user” or maybe just “use Twitter regularly”.

    5 million ways to bend the data to your liking…

    PS – You go, kystorms, you go. Nuts. Hope we won’t be disappointed with just the half-hearted half-seasonisms. ;)

  34. Typical analyst make-up numbers, and double-whammy as here you have to start with defining “usage”.

    “We define usage by at least monthly or more” with “at least some portion of it, in some way” in a “sheer pass-along effect” that’s wholly “indirect”.

    Gawd, what a mouthful of pure nonsense.

    Soooo if within the month you have maybe glanced at a Twitter, “in some way” or maybe even just heard someone talking about it, indirectly, in a “sheer pass-along” way, from our handpicked unscientific sweepstakes list, then you are a “Twitter user” or maybe just “use Twitter regularly”.

    5 million ways to bend the data to your liking…

    PS – You go, kystorms, you go. Nuts. Hope we won’t be disappointed with just the half-hearted half-seasonisms. ;)

  35. Here’s an “explanation of [the] survey methodology” but “it’s not likely to satisfy”…

    So why even bother to explain things then? Here’s an idea, how about an explanation that, in fact, WILL satisfy those that “feel” that 6% is wrong.

    Spin me round round, like a record…

  36. Here’s an “explanation of [the] survey methodology” but “it’s not likely to satisfy”…

    So why even bother to explain things then? Here’s an idea, how about an explanation that, in fact, WILL satisfy those that “feel” that 6% is wrong.

    Spin me round round, like a record…

  37. Peter – your blog post was bad enough, your near-insulting comments here are worse. Implying that WE are all stupid because we “FEEL” 6% is crazy? Tell you what – you go do some simple, direct, hands-on surveying OUTSIDE the bay area and see how many folks out there USE Twitter in any way.

    My suggestion to you: How about you stop making life hard for your sales team whose job is to actually help make you money off these kinds of ridiculous assertions?

  38. Peter – your blog post was bad enough, your near-insulting comments here are worse. Implying that WE are all stupid because we “FEEL” 6% is crazy? Tell you what – you go do some simple, direct, hands-on surveying OUTSIDE the bay area and see how many folks out there USE Twitter in any way.

    My suggestion to you: How about you stop making life hard for your sales team whose job is to actually help make you money off these kinds of ridiculous assertions?

  39. There’s ZERO way 3% of online adults use Twitter ‘regularly’ and there’s less than zero chance even 2% of online adults will be using Twitter regularly 6 months from now, there traffic is stagnant the past 3 months and will soon start decreasing because it’s going to be hard convincing anyone outside of Silicon Valley that it’s actually cool it just makes boring men having a midlife crisis nostalgic for the first time they heard about the internet or messed around with Q Basic…

  40. There’s ZERO way 3% of online adults use Twitter ‘regularly’ and there’s less than zero chance even 2% of online adults will be using Twitter regularly 6 months from now, there traffic is stagnant the past 3 months and will soon start decreasing because it’s going to be hard convincing anyone outside of Silicon Valley that it’s actually cool it just makes boring men having a midlife crisis nostalgic for the first time they heard about the internet or messed around with Q Basic…

  41. Well, whataya know, a Blue Moon, Scoble got this one right.

    When even the hype-king himself says your hype numbers are off, you are in deep trouble. And that’s my sheer pass-along “feeling”, at least some portion of it, in some way, more or less, indirectly, that is.

  42. Well, whataya know, a Blue Moon, Scoble got this one right.

    When even the hype-king himself says your hype numbers are off, you are in deep trouble. And that’s my sheer pass-along “feeling”, at least some portion of it, in some way, more or less, indirectly, that is.

  43. Twitdir says it’s tracking 521,933 Twitter users. I believe that is every user — not regular. If you look through Twitter accounts you can find a lot of orphaned users – haven’t posted in ages. Of course, that doesn’t mean they aren’t lurking.

  44. Twitdir says it’s tracking 521,933 Twitter users. I believe that is every user — not regular. If you look through Twitter accounts you can find a lot of orphaned users – haven’t posted in ages. Of course, that doesn’t mean they aren’t lurking.

  45. From Forresters site:

    “The sample was drawn from members of MarketTools’ online panel, and respondents were motivated by a sweepstakes drawing. The sample provided by MarketTools is not a random sample. While individuals have been randomly sampled from MarketTools’ panel for this particular survey, they have previously chosen to take part in the MarketTools online panel.”

    http://www.forrester.com/ER/Research/Survey/Excerpt/0,5449,589,00.html

    There’s a long way between 6% of people who opt in to take online surveys about internet habits for sweepstakes chances and 6% of online adults.

    It’s amazing these research firms stay in business, their data is utterly worthless.

  46. From Forresters site:

    “The sample was drawn from members of MarketTools’ online panel, and respondents were motivated by a sweepstakes drawing. The sample provided by MarketTools is not a random sample. While individuals have been randomly sampled from MarketTools’ panel for this particular survey, they have previously chosen to take part in the MarketTools online panel.”

    http://www.forrester.com/ER/Research/Survey/Excerpt/0,5449,589,00.html

    There’s a long way between 6% of people who opt in to take online surveys about internet habits for sweepstakes chances and 6% of online adults.

    It’s amazing these research firms stay in business, their data is utterly worthless.

  47. It’s amazing these research firms stay in business, their data is utterly worthless.

    Ahh see, worthless data, sadly, has it’s own value — used when needed to buffer something low, lower something high, justify some corporate policy, provide survey data in justification of some press release, drum up consulting and extra services monies (say it’s broke, then rush in to fix) or crystal-ball justify long-term corporate strategy changes.

    Heck, they STAY in business on ACCOUNT OF bad data. Actually good data they wouldn’t know what to do with, and is not an easy sell, everyone needs some “inside track”, not commodity reality data. Bad data is the backbone, trick is packaging bad into some sort of believable framework, however in this case the methodology was so far out as to be space-cadetish, with even the insiders saying it’s outlandish. And with Jeremiah morphing apologist, defending the indefinsible, yet one month on the job…heh. ;)

  48. It’s amazing these research firms stay in business, their data is utterly worthless.

    Ahh see, worthless data, sadly, has it’s own value — used when needed to buffer something low, lower something high, justify some corporate policy, provide survey data in justification of some press release, drum up consulting and extra services monies (say it’s broke, then rush in to fix) or crystal-ball justify long-term corporate strategy changes.

    Heck, they STAY in business on ACCOUNT OF bad data. Actually good data they wouldn’t know what to do with, and is not an easy sell, everyone needs some “inside track”, not commodity reality data. Bad data is the backbone, trick is packaging bad into some sort of believable framework, however in this case the methodology was so far out as to be space-cadetish, with even the insiders saying it’s outlandish. And with Jeremiah morphing apologist, defending the indefinsible, yet one month on the job…heh. ;)

  49. Interesting data. However, I think a number of people signed up for Twitter, me included, and use it initially and then stop after getting bored. After all we have mobiles/cells for texting rubbish to each other don’t we?

    Jim

  50. Interesting data. However, I think a number of people signed up for Twitter, me included, and use it initially and then stop after getting bored. After all we have mobiles/cells for texting rubbish to each other don’t we?

    Jim

  51. Thanks for very interesting article. I really enjoyed reading all of your posts.
    It’s interesting to read ideas, and observations from someone else’s
    point of view… makes you think more. So please keep up the great work.
    Greetings.

  52. Thanks for very interesting article. I really enjoyed reading all of your posts.
    It’s interesting to read ideas, and observations from someone else’s
    point of view… makes you think more. So please keep up the great work.
    Greetings.

  53. [...] Scobleizer wonders where Forrester got its Twitter data? A report claiming that 6% of American adults are using Twitter, leaves a bad taste in Robert’s mouth. However, the numbers are broken down and explained here, stating that the numbers come from an online survey, and consider the nature of a volunteer, sample panel and the results from that. [...]

  54. An excerpt from Forrester’s “Microblogging for Marketers” report…

    “Twitter users are, on average, 78% male and 31 years old, and they draw an annual income of $78,000. In comparison, members of the online US population are, on average, 49% male and 43 years old, and they draw an annual income of $68,000.”

  55. An excerpt from Forrester’s “Microblogging for Marketers” report…

    “Twitter users are, on average, 78% male and 31 years old, and they draw an annual income of $78,000. In comparison, members of the online US population are, on average, 49% male and 43 years old, and they draw an annual income of $68,000.”

  56. I landed here via a google search.. I’m doing research on twitter and I’ve spidered the entire twitter social graph, and the number of users with at least one follower is 280,000 so I’d say the number of active users is slightly less than that.

  57. I landed here via a google search.. I’m doing research on twitter and I’ve spidered the entire twitter social graph, and the number of users with at least one follower is 280,000 so I’d say the number of active users is slightly less than that.