I recently read Time Magazine’s Special Double Issue from May 10, 2010 (Vol. 175, No. 18), “The 100 Most Influential People in the World”. The centerfold of that issue included “The Influence Index”, an attempt to quantify each of the honorees’ influence by measuring the breadth of their brand through their social networking sites. They charted this influence index against each person’s age for an interesting graphical interpretation of influence.
To measure this personal brand breadth and influence, the statistical analyst who wrote the article used the following formula: Divide by two the sum of the number of Facebook followers (F) and double the number of Twitter followers (T). In graphical representation, that would be [F + (T*2)] / 2. Taken from the Time website, the formula for networking index = (Twitter followers ) x 2 + (Facebook Connections) divided by 2. (http://www.time.com/time/specials/packages/article/0,28804,1984685_1984713,00.html )
According to the article, the rationale for doubling the Twitter followers and taking the average is that the average person has twice as many Facebook followers as they do Twitter followers. So in their own words, they doubled the Twitter count to adjust for this. Of course, if the honoree only has one account, the influence score is simply the number of followers on that specific account.
The equation works fine in those cases where the person does indeed have nearly twice the number of Facebook followers as Twitter followers. In fact, it pretty much equates the influence index as the number of Facebook followers. For example, for someone who has 2 million Facebook followers and 1 million Twitter followers, the influence index yields 2 million. The problem with their equation is that it gives undue weight to Twitter over Facebook at the extremes.
For example, say you have 1 million Facebook followers and 100 Twitter followers. Your influence index is now 500,100, or just about half of your Facebook followers. On the opposite extreme, if you have 1 million Twitter followers and 100 Facebook followers, your influence score is 1,000,050. Needless to say, the arbitrary doubling of Twitter followers to adjust for the lower follow rates of Twitter is not the most effective way to to measure social media influence using these two services as a proxy. Instead, if you assume a certain number of duplicate followers across the platforms (say 50%), you could do the following: Take 50% of the larger network. If the smaller network exceeds this number, take the difference and add it to the number of followers in the larger network. If not, the influence score is simply the size of the larger network.
Using this new method, the three examples above would yield, respectively, 2 million, 1 million, and 1 million, a much better representation of influence and much more in line with the attempt made by the analysts creating this chart for the Time 100 Influence Index.
I have one final quip. The picture they decided to use for Lady Gaga, one of Time’s 100 Most Influential People, has 4-5 times the volume of any other person on the graph due to the massive headdress/wig she is wearing. This image distorts her ranking and makes it appear as if she is the most influential person of the list, when in fact, Barack Obama received a higher Influence score (by more than 13%) and ranks at the top.