Facebook asks How Many Friends Can You Have? (but doesn’t answer the question) and then shares its new concept of the “active network.” The ideas build off of The Economist’s Primates on Facebook piece and the data provided by the Facebook Data Team:
Your active network consists of all the people with whom you stay up to date. What makes your active network different from other networks is the way you communicate with the people in it.
This stream communication, rather than reciprocal and direct communication, forms your active network.
By this definition, Facebook states that active networks indicate a user tracks two to four times more people through the stream than through direct or reciprocal communication.
Several thoughts immediately come to my mind:
- Why is the definition of “active” linked only to the stream? This definition seems to try to sell the stream over being accurately defined. Active should include stream plus reciprocal communication. After all, reciprocal communication shows that a relationship is healthy and engaged.
Further, it could be easily argued that the reciprocal communication should be the sole metric to define the active network. A commonality of all direct and some stream communication is one-way. In the direct scenario, I send a communication with no response. In the stream scenario, I click on a stream activity and similarly have the *possibility* that this intention is not reciprocated (i.e., the recipient of my stream activity does not click on anything from *my *stream).
- Direct and reciprocal communication are mutually exclusive and stream communication should be considered the same way. That is, Facebook should break out stream communication only when it is the sole means that a user interacts with another person.
For example, I message Billy and Billy writes me back. I also click on a photo of Billy’s dog in the stream. It appears Facebook considers this later interaction a stream communication and thus considers Billy both a friend in the reciprocal and stream category. This accounting biases the stream category. It essentially gets to count people that might also have had direct or reciprocal communications. The same does not occur for the other categories. A person cannot exist in the direct and reciprocal category. It is one or the other. By removing people that are a part of the reciprocal or direct communication category, the proper importance of the stream will be revealed. Consider the Facebook definition versus thinking of stream as *stream-only*. Notice how the stream *could* reduce in size once the intersections of the reciprocation and direct were dropped: <img style="display: block; float: none; margin-left: auto; margin-right: auto" title="FB Active Network " src="https://kenyarmosh.com/wp-content/uploads/2009/04/image2.png" border="0" alt="FB Active Network " width="358" height="237" /><img style="display: block; float: none; margin-left: auto; margin-right: auto" title="Active Network " src="https://kenyarmosh.com/wp-content/uploads/2009/04/image3.png" border="0" alt="Active Network " width="342" height="227" />
- Would the stream-only communication still show that a user tracks two to four times the number of people than other communication options? If a user is a fan or “friend” of a large number of profiles, it might. But it might also show that for a person with a smaller number of actual friends, the stream simply acts as a complementary interaction. This strengthens the argument the active network is likely not accurately defined in its current form.
Facebook is doing some very interesting research but I think the definition of active network and the accounting of the stream needs to be reconsidered. Perhaps the above actually is part of their current methodology but it doesn’t seem so from their descriptions.