Twitter – Influence and Popularity Measurements

We live in the Google, Facebook and Twitter world as they dominant our online lives. For Twitter, its convenience and simplicity have made it a critical tool for product launch and brand management. It provides the pulse of the nation, seconds and seconds, and has attained a status that the Library of Congress is archiving its contents. From celebrities to politicians, iconic Wall Street firms to charities, a Twitter strategy is no more an afterthought.

Traditionally, people measure the popularity based on the numbers of followers an account has. That is certainly a good metric in the Internet age where more always win. As I wrote recently, a blog post that is read by hundred casual people is more valuable in the world of online marketing than one only ten people read but was cited by a professor. The former could bring more ad money since it was read by more people though the latter might have been better in quality.

So to estimate popularity, we usually check the number of followers for an account. The problem with that is that it gives an inaccurate measure of effective popularity or influence. In this case, I try to differentiate between popularity and influence. Someone could be popular in Twitter and yet not influential. By influence I mean when a post does generate a buzz or attention in the network.

To examine this, I used the very simple metrics on every account: “following” and “followers”. In the Twitter world, the former is the number of people an account is following while the latter gives the number of people following that account. With the followers, we can directly obtain Twitter Popularity Index (TPI) and ascertain how popular someone or brand is.

To move beyond popularity and ascertain influence, I divide the numbers of “followers” by the “following”. If you have a number greater than unity, you are indeed influential since more are following you than you are following others. To me, this is a better metric of ascertaining the effective popularity for each account and that eliminates the inflation people could create on their profiles. I call this Twitter Influential Index (TII). I initially called it Twitter Effective Popularity Index (TEPI). Either goes.

If you notice in Twitter, celebrities will not only have high followers, but also high TEPI. Britney Spears has a TEPI of about 13.9 (Sept 5, 2010) with 5,851,463 and 419,016. That is very interesting when one considers what it takes Ms Spears to add 419,016 people. How many of those friends are effective followers? Possibly, some added her just to appease her for adding them in her network. And they may not be actually interested in her posts! They are followers, but may not be influenced by her.

Twitter has this robust expansion tool that puts peer-pressure on members. By sending that email to you when someone adds you, it basically informs you that you can return the favor and add that person in your own network. (I estimate additional 5% hits on Twitter because of that email; they want you to get back to Twitter website). If not, why must Twitter do that assuming your account is already public? It is unlike Facebook where you have set your profile private and need to guard the contents.

With that system, we have bloated and fake popularity indices where someone can decide to add 1000 people in a day and will hope to get a favor back from say 80 people. Those 80 people may not even care for this person, but they just do.

For marketing executives, this is very important. A comparison between a person with 1000 following and 2000 followers with TEPI of 2, and one with 5000 following and 2500 followers and TEPI of 0.5 could give an impression than the latter could move the product easily since it has more followers. That is wrong. Most of the followers in the latter may be casual that reciprocated an aggressive “marketing” of the account even though they are not interested in the account. In the former, there is a sense the person is influential since more people are interested in the account than the account is to others. Personally, I will use the former with TEPI of 2 to launch my product in Twitter.

In the last few weeks, I tested this principle in Twitter by carefully adding people and noticing how my followers increased. The more friends I added, the more followers I got, though I did not retain most after few days. However, most of the accounts that added me without my covet marketing are easy to retain. And they are the ones that followed my posts more. Many of the marketed ones rarely bothered to comment on them.

So in summary, it is not the number of followers that matter, rather, it is the TEPI of an account that shows its vibrancy and influence over its networks. Use TEPI in planning your Twitter strategy and you will be better off. The followers could give you popularity, but TEPI provides you with level of influence. I consider George Stephanopoulos with TEPI of 2823 as more influential that Ms Spears though he has only 1,654,492 followers, millions less than Ms Spears’. My algorithm shows that he generates more comments and secondary patterns on his posts and it is possible that a brand pushed through him will penetrate more.

Finally, by combining TEPI, TPI, network preferences (music, politics, news, etc), location, post update rate, etc, I came up with Twitter Brand Composite Index (TBCI). TBCI provides a near accurate assessment between a brand and each account in Twitter. This is what the marketing executives must look for.