"I just don't get it." [i.e. Twitter]

When I talk about Twitter to corporate audiences, a very common reaction is "I just don't get it."  Hence, the post title.

If you want to learn more about microblogging check out this report (client access only).

Our data shows that 6% of US online adults use Twitter regularly.

If you want to reach an affluent, well educated, and early adopter audience, there might not be a better communication channel out there.

BTW each line in this message is < 140 characters.

@forrester is on Twitter, as are many individual analysts like @peterkim, @jowyang, @charleneli, @birdahonk, @sureshvittal, @carriejohnson, @SFOSkyGod, and @jbernoff.

UPDATE: Many people - namely Robert Scoble - believe that 6% is way too high.  Here's related information from Cynthia. For those interested in the methodology, you can read more about it here.  The specific question that survey respondents saw: "There are also more general activities you can do on the Internet. How frequently do you do each of the following activities? (Select one for each row)"  and "Use Twitter" was one of 11 options, with the frequency choices of "at least daily," "weekly," "monthly," "less than monthly," and "never."

A researcher on my team tells me the number is corroborated from data from another survey as well.  It isn't drawn from Nielsen//NetRatings, which has its own methodology.  While researching, I did d biz and email pownce, but neither responded.

It's fun to talk about whether 6% is BS or not, just like it's fun to watch traffic spike when one gets Scobelized.  To move that part of the discussion forward, I'd like to hear from people who have facts and data to refute the figure, not just feelings.  [But I do care about your feelings - deeply. :) ]

IMO the number isn't critical to the report's premise; more importantly, I explain why brands should pay attention to a small percent of online activity and how microblogging fits into social computing strategy.

Comments

[Robert Scoble]

There is NO WAY that 6% of US residents are on Twitter. Absolutely no way. My friends inside Twitter tell me they only have about half a million people world-wide on the service. A high percentage of which are outside the United States.

[Mickeleh]

The big gotcha in using Twitter as a medium to reach "an affluent, well educated, and early adopter audience" is that most Twitterers won't see anything you post. Twitter lets all users create a channel of folks that they follow. If you're not followed, you're invisible.

Until, of course, Scoble notices you.

[Joe Szabo]

The key to successful Twittering is A. Get followed and B. Optimize those 140 characters to the best of your ability so that you don't get de-followed.

Gibberish like "I'm eating a Big Mac" is what gives microblogging a bad name.

And I'd have to agree with Scobilizer; 6% seems a little inflated to me.

[Scott Monty]

I'd be interested to see the data teased out a little more. How did you arrive at 6%? Twitdir lists only 518,000 Twitterers, and I wonder how many of them are active (I have a couple of dormant accounts myself).

If Twitdir is to be believed, then that means there are only about 8.5 MM online adults. I really question that.

[Robert Scoble]

Scott: and of those they would ALL have to be within the US, which, if you watch twittervision.com for more than one minute, you can prove to yourself very quickly that's provably false. And like you said they'd all need to be online and regular users (what's a "regular" user anyway?)

[Daniel Riveong]

Peter,

While I'm a twitter fan, I've got to say that right off the bat that I'm pretty skeptical of the data.

Could you please expound more on this?

Does "6% of US online adults use Twitter regularly" mean people who check other people's twitters? Are they lurkers?

Even including lurkers (and not active twittering users) that's a substantially high amount (1 in 20).

Please give us a peak at your data!

Thanks!

[Daniel Riveong]

Peter,

Thanks for the extra information. I noticed that the page states that the methodology is “not a random sample”, which is fine but as a Marketer, I beg to ask:

1. Who are the Panelist who took the Survey?
2. What are Comparative Data Points?
3. What is a “regular user” on Twitter?

The Panelist
Basically, how tech and media savvy are the survey takers? Are there substantial representatives from the mid-west? High-Income? How many are from major cities from the west & east coasts?

Comparative Data Points
So we know 6% say they are regular twitter users. How many are iPhone users? How many do RSS feeds regularly?

If the general US/Canadian population say that 5% of them own iPhones and only 10% are RSS reader users, but the survey takers are 20% iPhone owners and 75% RSS reader users – that says a lot about the type of people who took the survey.

Using Twitter
The survey asks if the person is a “frequent” twitter user, but as I mentioned what does that mean? Plenty of Americans are frequent Wikipedia users, but that doesn't mean they contribute to Wikipedia. Ditto for Twitter.

Closing Thoughts
The "6% of US online adults use Twitter regularly" is a substantial sum (at least for me) no matter how you cut it - unless you're at a tech party in San Francisco.

So, again thanks for sharing, but we need more data!

Daniel

[Cynthia Pflaum]

For all those who challenge the data, I think a little context will help. I'm a researcher who works with Pete on survey design and data analysis at Forrester.

The 6% that we wrote in the document comes from one of our online surveys of US adults -- not of US residents. We define online adults as those who self-report that they go online at least once a month or more. To delve in deeper, our sample of respondents comes from a volunteer online panel, which is a a concentrated sub-sample of the US online population.

Also, consider your own definition of Twitter users. If you self-report/define yourself as a Twitter user, you probably consider usage to be at least daily or even weekly to be a Twitter-er. Forrester does not take that approach. For online activities, we define usage by at least monthly or more. If you look at the breakout of our data, you'll see that only 1% (and yes, that is rounded) use Twitter at least daily; 2% at least weekly; 3% at least monthly; and, 6% have used it all together. The goal of this report was not to focus on the Twitter-er -- or the daily user -- but rather to demonstrate the technology in its early adoption phase, i.e., of those who have interacted with it at all.

Defined, the statement -- use Twitter -- can mean have the application, viewed it on someone else's blog and so forth. This is the statement we used in our survey. The many applications where people can find Twitter, and interact with its content, in a respondent's mind can be considered "usage."

If you try and interpret this data as those who only use Twitter.com or are the Twitter-er authoring the tweets, of course, this percentage seems high. However, if you ask people if they plainly use it or not -- at least some portion of it, in some way -- this begins to make a little more sense.

Think about the traffic to your blog, your readership and the fact that you may have Twitter streaming on your Web site. Then, Twitter usage has a more effective reach than a respondent who jush pushes the content, or sends out the message. A 2nd hand effect is Twitter reaches more people by a sheer pass-along effect. Use can be indirect, not just direct, i.e., going to Twitter.com.

I hope this helps.

[Glenn Gow]

Peter, I have several issues with Twitter and with the claims around this 6%.

If Twitter themselves claim 500,000 users, then you’re telling me that there are just over 8 million online U.S. adults? The numbers don’t add up. What is your assumption of the number of U.S. online adults? If you assume 100 million (there are approximately 200 million U.S. adults), then you’re saying that there are 6 million regular Twitter users. Again, no matter which way I look at it, it doesn’t add up.

Furthermore, I don’t get it either when it comes to Twitter. I expound on this in The Top 10 Reasons Why Twitter Scares Me (http://www.achievemarketleadership.com/?p=74).

[Cori]

Ah, the curtains begin to part:

"6% have used it all together" is vastly different than "6% of US online adults use Twitter regularly"

Cynthia this is very helpful. Given how you've defined Twitter user, and given the self-selection of the respondants which skews the results somewhat, this is much closer to believable.

Thanks for the clarification!

[Steve Spalding]

Hey Peter, you wanted some numbers to refute the Forrester Statistics. Well, here are a few that I pulled together.

http://howtosplitanatom.com/news/twitter-this-is-your-life/

I really, really have a hard time believing that 6% thing.

[Steve Rhodes]

Even with your narrow definition of use, 6% seems way high. It might perhaps be half that for San Francisco, nowhere near that for the US as a whole.

I'd be surprised if even 6% have heard of twitter let alone used it.

Now it may just be that your respondents wanted to seem up on the latest thing, but it could also mean your sample is flawed.

Now I use twitter and like it, but numbers used for online are all too often overhyped. And figures like this lead people to question all the claims about numbers online.

And while you certainly can reach a small "affluent, well educated, and early adopter audience" (even the largest commercial or media twitter accounts only have followers numbering in the thousands. If you wanted
to read a larger number, you'd be better off advertising in Wired or on Engadget.

And with twitter, you can get some actual numbers. And twitter has even more numbers, but they probably won't be releasing them now. Still, they probably would be surprised to learn a survey says so many people use twitter.

Looking at some of the current ones featured on the timeline page

http://twitter.com/public_timeline

http://twitter.com/vma 1580 (it probably has dropped a bit since the awards aired)

http://twitter.com/nytimes 1240

http://twitter.com/WIREDScience 634

http://twitter.com/TMZonTV 78

[O'Grouchy]

I'll bet $100 that Habbo has at least 10x the real, active user base of Twitter. Twitter has yet to demonstrate widespread appeal or any kind of staying power. As you point out yourself, it has yet even to pass the "What the hell is this thing?" test.

[Kaleem]

Thanks for the explanation of your methodology, Cynthia.

Peter's statement that "6% of US online adults use Twitter regularly" can clearly be seen as a misstatement or misrepresentation of the data when, as you explain:

"6% have used it all together" and "the statement -- use Twitter -- can mean have the application, viewed it on someone else's blog and so forth."

When people read "6% of US online adults use Twitter regularly" neither Forrester's redefinition of the word "use" nor "2nd hand effect" is what comes to most people's minds.

Please tell us where the corroborating data that Peter cited comes from.

Peter, in your view "the number isn't critical to the report's premise" of "why brands should pay attention to a small percent of online activity and how microblogging fits into social computing strategy".

Being able to believe the data that you are using is accurate -- data that you are using to bolster your opinions -- directly affects the weight with which your words are considered.

Thanks again for the clarification, Cynthia.

[Paul Soldera]

I think the survey results are probably right, and as the above poster points out, there is no way you could claim that 6% use Twitter regularly from the data. But I also agree that the actual size of the market is somewhat irrelevant to the argument Peter makes.

If you ask questions like this is a survey of a new technology like Twitter, you get the whole curve responding. What I mean by whole curve is people who are currently using it regularly, people who just found out about it, people who tried it and forgot, etc. The 6% is the size of the current Twitter curve.

Cynthia points out the only 1% (rounded) are using it daily and 2% weekly. Considering Twitter is designed to be used pretty frequently, I would say 1% is the best estimate you have of people who are currently 'hooked'. The Pew Internet Research study puts the US adult online population at about 147 million (18+). That makes Twitter about a 1.4m user base. But it's a survey folks! With a margin of error at that level of approx 0.5% @99% confidence. So the real Twitter audience could be anywhere from 700,000 to 2.1m. From other data, it looks like the 700,000 might be closer to the mark.

Of course, once you get to arguing the numbers of real users from a survey sample and you're nit-picking over a few hundred thousand, time to go shoot yourself in the head. This is all back-of-the envelope stuff and you have to accept the error involved.

So while I don't agree the Peter's 'regular user' tag, the point is still valid. It's pretty obvious this is a technology that is gaining traction and a lot of influential people are using it. You don't need a survey to tell you that. And if you believe the survey data, a 6% curve is not bad.

[Jeremy Toeman]

Gosh where to begin. There's just so much BS floating in the post AND the comments.

re: 6% of US online adults. how about no way. if you stand behind the numbers I dare you to bet your personal salary on it. how about 6% of Silicon Valley area online adults? I'd accept that number.

Aren't you supposed to provide USABLE data to companies for a living, not hype that barely even belongs on a blog?

[David Berkowitz]

This post would be more credible if you at least retract that 6% use it regularly. Clearly that should be halved based on Cynthia's info. What's at stake here isn't metrics on Twitter usage; it's about credibility of Forrester data.

[Chris Brogan...]

While everyone fought over the actual stat, I used Twitter to help send a woman to college:

http://chrisbrogan.com/help-send-a-woman-to-college/

Twitter is. Stats or not. It is.

[thom singer]

"scobleized"? That is a verb. Sounds like household cleaning product. How does one get "scobleized? ha ha

[Ron Shevlin]

Sorry to be so late to the party...but I can only conclude that you've missed a decimal point to the left of the six. In July, Twitter Facts (http://twitterfacts.blogspot.com/2007/07/number-of-twitter-users.html)
blogged that TwitDir reported 340k twitter accounts w/ 2000 being created daily. Generously, that puts the number today, at the end of 2007, at about 800k. According to Forrester's research, a little over 70% of the 105m of so of the HH in the US are "online". 6% of them makes about 4.5 million HH using Twitter. Move the decimal place one to the left and your #s are a whole lot more in line with Twitter's own numbers.

p.s. I don't understand why Cynthia would raise the point that respondents came from a panel. Doesn't Forrester weight the data to represent the general population? And even if there was some sampling bias because this was a paid panel, the survey results are so far off from what Twitter reports, that sample bias doesn't seem like a good explanation.