Twitter is not Statistically Sensible
If you’re in the business of marketing, or if you work with marketers, or if you read blogs done by marketers you may have noticed a new game that’s gotten wildly popular: Statistic Says. It’s an exciting new game that takes Twitter to a whole new level by finding statistics relating to the demographics and use of Twitter, then making up stories about it’s future boom or bust. While it’s a lot of fun, it’s about as useful as looking to Hollywood for accurate depictions of history.
The unfortunate reality of the situation is that most of the competitors in Statistic Says don’t actually know how to read, or use, statistics. Tossing fuel on the fire, Harvard Business released a set of statistics that examine who is using Twitter and who they follow. The real trick to going through this report is interpreting the data.
Watching this hit, I noticed a general reaction to a particular statistic: Among Twitter users, the median number of lifetime tweets per user is one.
It seems that there are two camps on this.
1) Nobody actually uses Twitter! If most people only post once, it shows that it’s pretty useless!
2) Twitter isn’t about posting, or Tweeting. It’s about other people Tweeting and gleaning information.
What Harvard Business did, instead of trying to interpret the data alone, was look at trending from other “social media” sites. They compared contribution rates from users of Twitter, Wikipedia, and Facebook and noticed that Twitter was far more like Wikipedia than Facebook. As a user of all three, I can’t say I find the trend surprising.
During a day of using Twitter, I find that there are several kinds of people, including those that never tweet. The catch, however, is that they still use Twitter. Compare me to Guy Kawasaki for instance. You’ll notice a small (read: enormous) disparity in our posting rates. Me: 826 updates, 66 followed, 107 following. Guy: 25,411 updates, 122,767 follwed, 127,137 following. We may not use Twitter equally. I have 3.2% of his total posts and only follow .05% of the total number of users he does.
However, this follows the exact trend that Harvard Business points out: the top 10% of active users account for 90% of the content. It’s just like Wikipedia. Most of the people who use Wikipedia (and Twitter) don’t actually go there to push content. They don’t have a whole lot to say. Maybe nothing. Instead, they find value in watching it and listening. A few examples of my own use include following the Indianapolis Star’s shopping feed on Black Friday to find out where the lines and sales were. Or following RedBox to get a free rental every Monday. Maybe it’s following Mike Massimino to find out how the Hubble repair mission is going. It could be following the Yankees to get scores on my phone without having to pay for a subscription.
Whatever the case may be, Twitter users aren’t always “users” in the sense that they contribute content. To see the tweeting rate and take it to mean the usage rate is to ignore some very common sense facts about statistics. With that, I leave you with one more thing to think about: lurkers. If you participate on message boards, you likely have noticed the large number of people who never post as well as the very few people who post a lot. The same is true of a very public, very real time message board.