Jed Cohen

An Archive

Do Numbers Equal Value?


Every so often, someone new starts to follow me on Twitter.  Most of the time I look at the e-mail, click through to their profile, and decide whether I want to follow them back.  But every so often I get a follower notification that I just delete immediately.  More often than not it’s spam.  But not this time.

Here’s what Twitter told me about a certain “person” who just started following me (names omitted to protect the not-so-innocent):

Does this seem kind of ridiculous to anyone else?  What sort of earth shattering revelations must be in this person’s 99 tweets that over 40,000 people are following him?

Let’s take a look…

I decided to take a slightly scientific approach to this.  Here’s a few vitals:

  • According to When Did You Join Twitter?, this person joined Twitter on March 25, 2009.
  • I looked the handle up on Twitter Grader, where the account scored a 98.
  • I then used Twitterholic to get historical data on friends, followers, and updates.  Then I graphed the results:

To give myself a sense of perspective, I ran a number of other users through Twitter Grader as well.  Guy Kawasaki’s, Justine Ezarik’s, Drew Olanoff’s, Evan Williams’, Biz Stone’s, and Jack Dorsey’s accounts all scored a 98.2.  I plugged in another 15 or so users (not all celebrities), and only four of them had scores lower than 98 (sampling biases include following them and finding them interesting).  Only one had a score lower than a 90.

It’s absolutely obvious that this user is gaming the system.  They’re using any one of the hundreds of ways out there to increase their follower numbers, probably by agreeing to follow everyone back.


We’ve come to use following/follower numbers as a measurement of influence or value.  But it’s simply not true.  I don’t care what rationale you have, 99 tweets in 270 days can’t enrich the lives of 40,000 people.  It’s just not possible.

Numbers have become a heuristic for the social web.  We have a mental shortcut that says that higher numbers are better than lower ones, and we use it to judge authenticity or likability or popularity or any other number of traits.  But we can’t keep on using numbers as an indicator.  This user that I’ve featured here shows us that it’s much too easy to fake them.  And if you don’t believe me, consider that there are 15,740 experts on Twitter according to Mashable (Pete Cashmore scores a 100 on Twitter Grader by the way).

What else is there?

There’s a lot written on how to sell social media to your boss, and almost everything will at least mention return on investment.  And ROI is always going to be measured based on numbers – whether it’s sales, cost per thousand (CPM), click-throughs, subscriptions, or customer surveys.  It’s a numbers game, sadly.

But maybe, together, we can come up with some better numbers.  What do you think?

Written by Jed

December 27th, 2009 at 11:12 pm

  • Zoey

    interesting article. People definitely use number as a heuristic, although i'm not sure how…Case and point, you looked at the numbers associated with this user suspiciously while others might have taken these numbers to be a sign of authenticity or importance. I've actually been thinking about the twitter “phenomenon,” and the popularity of certain users (e.g. the one you quoted) boggles my mind. Ideas? What's the intention behind the user and people who agree to follow him/her? Perhaps some people take twitter to be like facebook, and thus they “collect” followers like people collect friends on facebook. Or perhaps they see Twitter as a genuine information exchange (which I find unlikely for 99.99% of the users out there). And my personal favorite explanation-put eloquently by one of my professors-“mental masturbation”.

  • Jed Cohen

    I'd say that people use these numbers as a heuristic for popularity. In the physical world, we can say with much more certainty that the more people you know the more popular you are. But in the digital word, it's not necessarily the same. This user probably mass follows 50 people at a time, then unfollows those who don't follow him back (he's no longer following me for example). It's a simple way to game the system and get around Twitter's requirements to have a certain followers/following ratio.

    Many of the users I quoted are internet celebrities, and like movie stars the average individual wants to be close to them, something that can be accomplished through the follow button. There's also the chance to see inside an organization you don't have access too, which is part of why I follow Twitter's founders. Or the opportunity to learn from people you respect. Or to see what content they can find on the web (Guy Kawasaki).

    Twitter, like Facebook, can have its serious applications for both professional marketing and personal relationship building. And it certainly has its uses for link sharing and real time trending internet nonsense. Yes, I would agree that a good portion of the Twitter community is marketers, and chances are they're marketing their products to other marketers. But there are also some serious conversations going on. And that's worth looking into.

  • Zoey

    I agree with the idea that people use the follower number as a heuristic for popularity-a restaurant with people waiting out the door will attract more people-however, I do think there are other interpretations of such numbers . My whole point above is that while number convey popularity to many, they may also be cues of insincerity, self-importance etc.

    And you are right, despite commercial “misuses” (i.e. companies that are on twitter just to be on twitter…and they make no valuable content contribution), there are definitely interesting conversations going on, mostly between consumers . The challenge right now is how do we track and aggregate such information so that we can describe consumer behavior in terms more specific than “WOM is happening” or “people are talking about xyz”

    i still think you should look into my recommendation regarding atlanta

  • Jed Cohen

    I see what you are saying about using numbers as a proxy for a user's feelings of self-importance. I think there is an argument to be made for that from an analysis standpoint. But we run the risk of misclassifying celebrities or brand representatives if we don't factor in more information. Perhaps we need to expand the data set to include @replies both from and to the user? This should supply more information that we can use to determine who is truly popular and who is gaming the system.

    But at that point I'm not sure we're not looking at a heuristic anymore – we're looking at building a system designed to measure the authenticity, popularity, sincerity, or actual importance of an individual in a social network. Which is probably a big part of describing consumer behavior on social media (think of all the Gladwell-esque articles to be written).