THE TIMES THEY ARE A-CHANGING

posted on Oct 30 2012

Changed Edge Rank I: Why “Talking About” is not “Engagement” and why this ultimately leads to a deteriorated user experience

We have wanted to publish information on our Facebook insights for quite some time now. But the changes made to the Facebook Edge Rank algorithm in the past two weeks have also altered everything we wanted to talk about. So in this note and in the consecutive ones we will only concentrate on the affected changes and what page owners can do to anticipate them.

The Edge Rank Algorithm determines which story makes it into your personal news feed. You may recall that your feed is not sorted chronologically by default but according to relevancy. When this was introduced a couple of years ago there was an uproar in the community, but pretty quickly everybody seemed to appreciate that Facebook was fishing out those cat pictures from old high school classmates you hadn’t talked to in years and instead highlighting stories from people closer to your daily life.

The Edge Rank Algorithm determines which story makes it into your personal news feed. You may recall that your feed is not sorted chronologically by default but according to relevancy. When this was introduced a couple of years ago there was an uproar in the community, but pretty quickly everybody seemed to appreciate that Facebook was fishing out those cat pictures from old high school classmates you hadn’t talked to in years and instead highlighting stories from people closer to your daily life.

The Edgerank algorithm is what determines this relevancy. And it takes into account a variety of factors that are accessible to all page owners via Facebook Insights. Facebook has done a really great job of offering you all data relevant to your pages’ performance. You can even download this data as a CVS file for any given period of time. (Obviously, you don’t get the whole recipe though, neither how important which KPI is for you overall relevance nor 100% of all ingredients used in the Edgerank agorithm. But it’s still much more than Coca Cola would ever do.)

The latest update in the Edgerank algorithm has again stirred quiet a lot of negative reactions especially among Brand Page owners who have reported a significant loss in Organic Reach and Viral Distribution of their page posts. You can find a good analysis of the algorithm update by Edgerank Checker here.

When speaking about the new factors that influence the updated Edgerank algorithm Facebook claims that two factors have increased their impact in the update: The first is the “Talking About” value also referred to as PTAT (People Talking About This). The PTAT rises when people like or comment on your story thus reproducing a story “about you” in their stream. (The second aspect is “negative feedback,” which we will address in our next posts.) Facebook has also informed us that the algorithm is designed to promote content with high “engagement” from the user perspective, but what they are actually referring to is the PTAT.

In order to understand why both terms are not the same, one has to know that “Engagement” is also a term from the Facebook Insights section. But in comparison to the PTAT value it takes into account a much wider set of actions a user can take while interacting with content. For example, Engagement also includes clicks to open a picture to its full size and clicks on referral links provided in your page posts.

Referral links affect publishers, like Moviepilot, especially in comparison to pure advertisement brand pages, who are more concerned with exposure than with website traffic. A publisher’s goal is to write intriguing page posts that catch the readers’ attention and make him want to know more about the content. The brand page manager, on the other hand, wants to get across his commercial message. His goal for the post is not to leave the Facebook canvas.

Let’s look at an example how different page post compositions reflect these opposite goals: An entertainment editorial piece could be based around an A-List star’s birthday (Clooney, Pitt, Depp). The publishers interest is to generate readers for the article written in order to celebrate the star. So after the jump the reader might find interesting and fun facts about the past career or future plans of the star.

From an Edgerank perspective this is not optimal. Alternatively, the page manager could simply wish the star a great birthday without an external link and invite fans to do the same by liking the story. (Everybody remembers Steve Krug, right?)

Since this alternative does not offer a jump to more content about the star, the call to action is streamlined and likes and comments rise. The updated algorithm enhances this composition, despite the fact that the first composition might offer more engagement because the clicks on the external link are counted as well. Comparing both compositions from a user perspective, we argue that the page post that offers more in depth information and content about the actor offers the better user experience.

So the added value of PTAT ultimately encourages posts to be composed of rather simplistic content with easy like and commenting call to actions. This is mostly done with sweet and/or funny pictures and video content. A true optimization from the user perspective should have enhanced the “Engagement” value rather than the PTAT, thus also promoting more challenging, in depth content and higher quality editorials just as much as extremely witty viral posts. To sum up, one can say that Facebook is currently putting more effort to keep you within the Facebook canvas than to let you go somewhere else. A fact that does not correlate too well with the mission statement of “connecting the world” and a general growing understanding that Facebook is more than just the blue canvas, but the entry door to thousands of awesome apps. The apps lose. The canvas wins.

Please leave your comments and remarks.

Next time we will publish an overview how directly PTAT affects the Organic Reach of your Page.

comments powered by Disqus