Gaming Facebook’s EdgeRank Algorithm

Facebook thinks of everything. They’ve figured out how to connect hundreds of millions of people and how to expedite and accelerate sharing among them.  They also figured out how to parse and moderate all that inbound information from your friends by embedding an algorithm that automatically decides which posts you see and which you don’t.

This bit of mathematical legerdemain is called the EdgeRank algorithm. Like the formula for Coke or Google’s page ranking system, its top secret. If you buy the Facebook spin, this not-so-well-known tool allows you to focus on the most important people, brands and organizations in your life; the ones you care about most. By editing your newsfeed, Facebook spares you from a deluge of blah-blah posts, baby pictures from people you used to work with, infrequent posters, long lost relatives or people you friended out of courtesy.

From what we can discern, the algorithm looks at frequency of posts, relevance (number of friends clicking, liking and commenting) plus your history of clicking and messaging within Facebook. Facebook is sorting content based on affinity, weight and relevancy.

Each item in your news feed is considered an object. Any action taken relative to an object (like, share, comment, etc.) is called an edge. Each edge has a different mathematical value as does each object.  Edge compared to objects equals a score that includes or excludes a specific post. Facebook does the math and presto! Your news feed and top news sections are populated.

While this might be good news for individuals, it’s a serious challenge to brands; especially those brands who were counting on a multiplier effect to get free news feed space among the top stories of their million plus friends. Think about the math.

If the average Facebook user has 130 friends, getting one brand posting to an individual yields 131 impressions. If your brand has 1 million friends that’s 131 million potential impressions for each post you land in the news feed. Yet the EdgeRank degrades that significantly. Baked into the formula is the fact that most people like a brand quickly (50% within 80 minutes of visiting the brand page) and without a depth of emotion.  Most “likers” rarely return to the brand page and are less likely to comment or interact with brand postings than they are with friends and acquaintances.

As a result, according to the available research, no more than 20 percent of brand posts get through the algorithm and into the news feeds. That means, at best, you get 26 million rather than 131million potential impressions, eighty percent less than you were counting on. This presents a SEO-like challenge for marketers, which has spawned a new sub specialty dubbed “newsfeed marketing.”

The idea, like SEO, is to game the system to get more brand content into news feeds. Working to achieve this subversive objective forces brands to better understand the Facebook gestalt and abandon their primitive thinking that Facebook is just another new, slightly nuanced advertising channel. To finesse the EdgeRank algorithm, you have to create content that will be scored positively on the basis of affinity, weight and relevance.

The way to do this is to build and sustain frequent and topical engagement that your friends find interesting, worthy of comment and stimulating enough to spend time with. The building blocks for prompting more intense engagement are …

  • Asking open-ended questions
  • Inviting opinion, comment and conversation
  • Connecting organically to offline news, cultural or sports events
  • Limiting brand and sales language and catalog-like copy
  • Creating games, quizzes or surveys
  • Soliciting user-generated content
  • Posting photos and videos
  • Including links in posts
  • Developing limited time offers

Ironically, in our efforts to beat the system, we’ll better understand and utilize the Facebook platform as its being developed and extended. Let’s hope they don’t change the rules before we get a few wins on the scoreboard.


Comments are closed.