Model of Facebook Net Worth

Last week, our group attempted to create an equation detailing the net worth of a single Facebook account. We imagined a situation in which Mark Zuckerberg decided to dis-invest all of Facebook’s net worth to each user, based on the net worth of their respective profiles. The question and scenario are both fairly straightforward, but the solution is not so much. There are many factors and variables to take into account, especially a social networking site as complex as Facebook. I wanted to create an index that would assign various “points” to a profile for having certain features. These features included having a large list of Friends, the fraction of the day spent on Facebook, number of comments, likes, and posts on the user’s profile, among others. I had originally wanted for each profile to have a certain numerical index (which theoretically should have no limit). To determine the “slice” of the Facebook pie earned, one would simply divide a profile’s index by the combined index values of all other profiles existing. For example, if my profile was determined to be worth 4 index points out of an aggregate index “pool” of 100, I would receive a 4% cut of the Facebook pie.

When I talked to Professor Sasha about the index idea, I do not think I was clear enough. He thought that each profile would have an index value, and the amount of money would be based off said index value. However, perhaps the most important aspect of my method involves knowing the index values of ALL OTHER Facebook profiles, as well. That way, a ratio of individual index value to the aggregate index value could be found and multiplied by the net worth of Facebook (which was $96 billion as of last year).

There are a multitude of other factors that must be taken to account that we had not gone over last week, as well. I brought up the possibility of Facebook groups and fan pages as well, which may be created by any Facebook user with an email address. That solution, however, is relatively simple; just calculate index values based on registered emails rather than just a Facebook profile. Another more difficult issue was ascertaining the value of a friend list. Clearly, a high school girl with 1200 “friends” would have a less valuable friend list than a prominent politician with 600 friends on his personal profile. But how would one calculate the value of a friend list? If we used the same index as illustrated before, we would be thrown into an infinite loop, in which an exact value of a friend list may never be known due to the extremely complex nature of Facebook networks. I am not a Course 6 major, so I am not trained in solving these types of loop issues. One rudimentary solution I have, though, is to calculate a “Preliminary index” based on all other variables EXCEPT the friend list. Thus, a “final index” can be calculated by determining the preliminary value of all the profiles in a Facebook friend list.

All in all, this was a very interesting activity, calculating the monetary value of a Facebook profile (or in my case, a registered Facebook user). There are numerous variables and challenges that our group had not gone over in the brief time we had. However, with additional learning in Civic Media (and perhaps some algorithm training), I am confident that this problem will be much easier to solve by the end of my MIT career.