Deconstructing Facebook’s EdgeRank

The formula behind the Facebook news feed

During the latter part of April Facebook held its developer conference, f8. During this event, several Facebook executives and staffers shared insight about many of the developments rolled out during the conference. From the the new Like Button to the Facebook Open Graph, there was a great amount of news that came out during the conference.

One important discussion centered around EdgeRank, Facebook’s formula for determining what items show up in the Facebook news feed. To explain further, every Facebook profile has a live feed and a news feed. The feed is a stream of content and Facebook status updates coming from your friends, groups, causes and the pages you Like. For every unique profile (or Facebook account) there is a unique feed. Your feed is different then mine because we each follow and friend different people/organizations. The live feed is a real-time stream of posts that populate as they occur. The news feed is a little bit more complicated. Facebook determines what should go in your news feed based on your previous behaviors, and not everything makes the cut.

According to the Facebook consultants at BrandGlue, 1 in 500 updates make it to the news feed. How does it work? If you like or comment on updates from one particular friend quite often, you are likely to see that person’s update in your news feed on a regular basis.

To use a personal example, I tend to like or comment on updates from my friend Juan because he is hilarious, my sister who is expecting her first baby, and Arizona Highways’ page because they are always providing great content. Because of my tendency to do so, my personal news feed usually includes an update from one if not all of those profiles.

That’s the non-scientific method for determining what happens on the Facebook news feed. During the f8 conference, Facebook did what Google has never done; they provided the mathematic formula behind news feed optimization. Here it is:

Facebook EdgeRank Algorithm

Note: From this point forward I’ll use Facbook’s terminology regarding objects and edges. If you don’t speak Facebook, an object = a status update or post. An edge is a comment, like or interaction with that object.

Based on this formula, an object’s likelihood of appearing in a news feed is a function of three elements:

  1. Affinity
  2. Weight
  3. Recency

Let’s discuss each in plain English. I’ll also provide a brief explanation of what Facebook page administrators can do to take full advantage of this formula.

Affinity Score – This number defines the relationship between object creator and recipient. Those who comment and like your personal updates have a higher affinity than those that do not.

What this means to you: Create objects that appeal to different kinds of connections. If you see the same people commenting and liking a specific type of object like photos, create a different kinds of objects like questions or polls to elicit a response from other connections. The more variety you have with the profiles that provide feedback and like your objects, the greater overall reach your page will have in the news feed.

Weight Score – Each object is assigned a score based on the number of comments and likes it earns. An object with 15 comments and 20 likes has a much greater weight than an object with 0 comments and likes.

What this means to you: Develop a strategy for creating objects that will generate feedback from your connections. Entertain, educate/enlighten, make invitations, say “thank you”, ask questions, etc. View every post opportunity as a method for garnering important insights and learning more about your page connections.

Time / Recency Score – This number is a reflection of how long ago the edge was created. Each edge will lose its mojo very quickly. For example, the oldest edge in my news feed right now was created 42 hours ago. Most were created in the past 18 hours.

What this means to you: Allow each object you create to gain some exposure. Creating back to back to back objects in a single afternoon will stifle each object’s ability to garner comments and likes. Meanwhile, adopt a regular posting schedule (like once or twice a day) so that you always have an object and its related edges placed in the news feed of your most loyal connections.

Helpful? Feel free to comment if more clarification is needed.

Comments

  1. Interesting read. Thanks Chris!

  2. Very useful information, thanks. I hadn’t really understood the difference between the News Feed and the Live Feed before this either.

  3. Chris Sietsema says:

    Thanks, Victor. Happy to help.

  4. Fascinating. Thanks for taking the time to share.

  5. Great explanation, thanks!

  6. This terminology will be of great use to my team. Thanks for sharing your information.
    Thanks,
    Brian

  7. A very interesting read, this formula should also be applied to content generation so that it can be driven towards what is most likable and popular.

  8. Chris Sietsema says:

    Thanks, Simon. I definitely agree. It seems Facebook is on to something here with how they measure what is seen and not seen. It’s similar to the Google algo in a way in which links = likes. We could all use some more Like builders.

  9. Interesting read. Finlay those Social Heads that say “SEO-is-dead” have a sock in their mouths.

    SEO is changing & it’s breaking up into verticals.

  10. Great content and explanation. When you really think about it – edge rank as you’ve explained in this post is exactly what makes you a hit at a party – or not. I wonder if facebook has a ‘black-hat nfo’ team? No doubt they do. Cheers and I look forward to more updates.

  11. Some of these facts were clearly noticable even before facebook came out with the official announcement, and now that the official announcement has been made, it would certainly help admins garner more fans.

  12. klaudia says:

    great help. Big thanks.

Trackbacks

  1. [...] Deconstructing Facebook’s EdgeRank Facebook's EdgeRank algorithm, from Teach to Fish Digital [...]

  2. [...] an article over at Teach to Fish Digital that I was reading a while back and I wanted to share some of with [...]

  3. [...] http://teachtofishdigital.com/facebook-news-feed-optimization/ This entry was posted in Uncategorized. Bookmark the permalink. ← New system for keeping track of our information [...]

  4. [...] Teach to Digital Fish has done a superb post addressing just that topic. [...]

  5. [...] only one in 318 tweets are being retweeted. And, thanks to Facebook’s EdgeRank algorithm, only one update in 500 is actually seen by a company’s fans. One real reason social networks see more engagement is because their content is often more … [...]

  6. [...] Baer, vers l’augmentation progressive de ce .5%. Comment? Quelques pistes ont été ouvertes ici et là, avec à chaque fois l’importance de créer du contenu intéressant, d’inviter [...]

  7. [...] Facebook News Feed Optimization: EdgeRank | Teach to Fish Digital [...]

  8. [...] Now I feel stupid. Why didn’t I figure this out before? Amplify’d from teachtofishdigital.com [...]

  9. [...] to the EdgeRank algorthim that decides what appears on your wall and what doesn’t, it is estimated that only one in 500 or [...]

  10. [...] is not all members of your crowd's News feed are treated equally. This blog post elaborates on how the feeds work: To explain further, every Facebook profile has a live feed and a news feed. The feed is a stream [...]

  11. [...] chronological.  That’s right, just like your personal News Feed (which uses Facebook’s EdgeRank to determine which posts are most relevant to you), Fan Pages are now subject to a Facebook [...]

  12. [...] avec autant d’entités. Pour limiter le flux de données apparaissent sur leur mur personnel, Facebook utilise un algorithme de filtrage qui sélectionne lesquelles des mises à jours de ces pages aimées (et amis) y seront publiées. [...]

  13. [...] They are not (even) a measure of impressions! Now, measuring impressions has been described as a very old fashion metric but counting fans isn’t even that. Hypebot recently released a study showing that one in a hundred fans at most LIKEd brand updates. Previous studies of the EdgeRank algorithm that filters updates appearing on fan walls showed that 1 in 500 brand updates reached their targets! [...]

  14. [...] recently released a study showing that one in a hundred fans “liked” brand updates. Worst yet, previous studies showed that 1 in 500 brand updates reached their targets. Jovisst. Antal Facebook fans innebär [...]

  15. [...] the Send button is that a use of it also adds to the pages “Like tally”, which means it impacts Edge Rank [...]

  16. [...] sich mit dem Themen NFO oder Newsfeed Optimization bzw. EdgeRank, sozusagen das SEO von Facebook, noch nicht beschäftigt hat, wird die Folien Nr. 27 bis 30 [...]

  17. [...] to Chris Sietsema 1 in 500 updates on Facebook make it to the news [...]

  18. [...] by Facebook at their developers conference f8, in April 2010. Digital marketer Chris Sietsema blogs here about these three important pieces that together will determine the visibility of your content on [...]

  19. [...] Uno de los grandes atractivos (aunque casi nunca se mencione) es el grado de interconexión entre los usuarios de las redes sociales. El grado de separación (el número de intermediarios necesarios para comunicar un mensaje de una persona a otra) es más bajo en las redes sociales que en el resto del mundo: 5.7 en Facebook, 4.7 en Twitter, contra 6.6 para el email. No obstante, este fenómeno se reduce bastante debido a la baja difusión de los mensajes en estas redes: sólo un tweet de cada 318 es re-twitteado.  Y, gracias al algoritmo EdgeRank de Facebook, solamente una de cada 500 actualizaciones es vista por los fans de una compañía. [...]

  20. [...] chronological.  That’s right, just like your personal News Feed (which uses Facebook’s EdgeRank to determine which posts are most relevant to you), Fan Pages are now subject to a Facebook [...]

  21. [...] Slides 39-42: Tools and concepts to maximize efficiency and exposure with social media. There are references here to Facebook EdgeRank. [...]

  22. [...] few descriptions of the EdgeRank algorithm from TeachToFish, Travel Quest and [...]

  23. […] provide a measure of ongoing temperature control for your passions on Facebook. That’s how the Facebook EdgeRank formula works, governing what you see in your Top News stream versus what’s relegated to the less popular […]

Speak Your Mind

*