How to Be Noticed By Your Friends: Introducing EdgeRank
We’ve probably all wondered how Facebook decides to sort the various forms of activity on our Newsfeed. How does it determine what appears in our feed, and which posts should appear higher than others? It turns out that Facebook uses an algorithm called EdgeRank to decide for us what we want and don’t want to see. And while Facebook is notorious for keeping the workings of their ever-so-mysterious algorithms a secret, its developers have been kind enough to divulge the elements that make up their EdgeRank algorithm, at least on a very basic level. Perhaps the most noteworthy advantage of understanding EdgeRank inside and out would be for a company to get an edge (no pun intended) in its Facebook marketing campaign, as making sure your announcements reach your followers’ Newsfeeds can make the difference between a business-changing success and an unnoticed failure. However, for the purposes of this blog post I will focus mainly on social networking as it applies to the average college or high school student.
There are plenty of examples of Facebook posts that typically achieve high EdgeRank scores and stay on the top of one’s Newsfeed for quite some time. The classic example of such a post would be a relationship status update. The newsflash that two people are now publicly in a relationship (assuming it is genuine) catches the eye and garners the appreciation of many; in fact, nowadays Facebook is how a large portion of people first find out about newly established couples. Not surprisingly, this type of post would generate many Likes and comments, boosting its EdgeRank so it makes it to the Newsfeed of other friends who may also potentially contribute to its popularity, boosting its EdgeRank even further until eventually the post goes “viral” and appears on the top of everyone’s Newsfeed. Other common examples of such posts include things such as college/grad school acceptances, amusing status hacks, humorous/controversial photo uploads…the list goes on.
So what exactly are the components that make up EdgeRank? First of all, EdgeRank is an obvious analogue to the PageRank algorithm we discussed in class, in the context of search engine optimization. We learned about how the PageRank algorithm determines the importance of search results by calculating the authority score of each page, which depends on the hub scores of the pages that link to it. Now I will proceed to make an attempt in analyzing the mechanics of EdgeRank by direct analogy to PageRank and web search jargon. For the purposes of EdgeRank, the Newsfeed is analogous to a web search in that things deemed the most relevant will show up more often than things deemed less relevant. A post or status update is equivalent to a page/node, and activity on other people’s posts/walls is equivalent to a link. A post has a high authority score if it has many inbound links (i.e. comments and likes). While it is less straightforward to come up with an analogy hub score, it seems to make sense (at least to me) that the hub score of a link/user who interacts with a post is roughly correlated with the amount of mutual friends the user has with you. While this definition of a hub score does not portray the “legitimacy” of a link per se, it does somewhat accurately capture the probability that a post with which this user interacts is something in which you would be interested. At the very least, it seems to explain why a friend’s status update with 20 comments from your friends shows up more often in your feed than, say, a status update from the same friend but with 20 comments from people you don’t know.
This is not all there is to the EdgeRank algorithm, however. According to Facebook, the algorithm actually consists of three, simple-looking but deceptively complex factors: affinity, edge weight, and time decay.
– Affinity: This is a score that is basically determined by how interested you are in someone and what they are doing. Commenting on someone’s photos, liking their status updates, even simply stalking their profile…these will increase the likelihood of them appearing in your feed in the future. It’s as if Facebook just “knows” and fills you in on the lives of the people you think are important. Conveniently (or maybe not so conveniently, depending on who looks at it) enough, affinity is strictly one-way, so you can stalk someone all you want and they won’t be seeing any more of you than usual on their feed. Affinity is different from authority score (at least the way I defined it here) in that while authority score determines the popularity/relevancy of a single post, affinity serves to increase the relevancy of a person’s entire profile and his/her activity as a whole.
– Edge weight: Different types of content are more likely to appear in people’s feed than others. Facebook has decided that photos, videos, and links (as in hyperlinks) are more important to people than other types of posts, and thus weight them more heavily. Edge weight also works in the favor of each individual user’s interests; for example, someone who has a clear preference for browsing photos is more likely to have posts including photos in their feed than someone who rarely does so.
– Time decay: This seems relatively straightforward; a post becomes less relevant the older it is. There might be more to this but I’m not sure.
Despite all the similarities, there are some clear differences between EdgeRank and PageRank. Perhaps the most glaring is the fact that while the EdgeRank values of posts are constantly being updated due to the time decay factor, PageRank values of search results are more or less temporally static. Consequently, Facebook posts cannot achieve hierarchical dominance in the way that websites can, only fleeting moments of glory at best. On the other hand, while PageRank only affects those who actively search with a particular query, EdgeRank exerts its influence on anyone who happens to be logged in and on the home page. Before what you just posted gets swamped by whatever else is going on in people’s lives, it will, for those first few moments, hold the spotlight, whether your friends care to see it or not.
So to all those who crave attention or desperately need to have their post seen for whatever reason, here are some general suggestions for optimizing a Facebook post to increase its potential EdgeRank score:
– Be active on Facebook (i.e. create objects and edges) at times when your intended audience is most likely to be logged in. This is already obvious to many people and can easily be approximated by noting the number listed on your Chat sidebar (though not to an exact degree because some prefer to sign on as invisible). Explained in terms of EdgeRank, your post will have a higher probability of making it to people’s Newsfeeds with a lower time decay.
– Devise strategies to increase the affinity between you and your audience, including in your posts things that will provoke interaction and responses. For example, studies have shown that incorporating question words such as “what”, “why”, and “where” in your status updates tends to generate more comments and/or likes from your audience, which will in turn increase the chances of your posts reaching them in the future. Humor is simple but also works quite nicely in this regard.
– Post things like photos and links that have a higher edge weight than simple plaintext status updates.
For the curious Facebook user, there is actually an unofficial EdgeRank checker available on the Internet that can be used to check the EdgeRank score for each of your posts. It can be found at www.edgerankchecker.com!
Source: http://econsultancy.com/us/blog/7885-edgerank-the-most-important-algorithm-you-ve-never-heard-of