Facebook algorithm: a virtual creation of Gladwell’s Tipping Point Theory

By Lauren Wessel

Aware of all the beauty the world has to offer, I often find myself sitting behind my laptop, scrolling through Facebook. Sometimes, my entire timeline seems to be in the grip of something. This can range from topics like the Paris attacks, to people hyping over this white and gold (or is it black and blue?) dress that kept the community busy for a solid 24hs. In his book “The Tipping Point”, Malcolm Gladwell explains that certain messages can be spread as epidemics, which are determined by “The Three Rules of Epidemics": The Law of the Few, The Stickiness Factor, and The Power of Context. It seems as if Facebook has exactly included these characteristics in its algorithm to ensure that a post goes viral.

Facebook, instead of randomly picking posts from your (probably way too extensive) friend list, has designed an algorithm that determines what and who is visible on your timeline. This is a very complex algorithm, but it comes down to these variables: the interest of the user in the creator, the performance of the post among other users, how popular the creator has been amongst other users, the type of post the user prefers, and the recency of the post.

The success a post is heavily dependent on the involvement of people with a particular and rare set of social gifts, which Malcolm describes as Law of the Few. Furthermore, the content of a post weighs heavily, which resembles the Stickiness Factor. If a post is usually interesting to the user, it will appear sooner. Furthermore, the Power of Context is important. It could be that a post has all it takes, but is overshadowed by other things that are posted at exact the same moment.

Facebook’s algorithm is complicated, and cannot be explained in one blog post. However, it should be clear that like Malcolm’s Tipping Point, popular posts do not go viral on accident: it is a combination of all these variables that allows Facebook’s algorithm to spread these posts like epidemics.