[Updated: 10/04/12] // We’ve received a lot of feedback regarding this data. A Facebook Ad Rep had emailed the following:
- We’re continually optimizing newsfeed to ensure the most relevant experience for our users
- One of the key factors in our optimization is engagement: the amount of clicks, likes, comments, shares etc. generated by a piece of content
- While overall engagement should remain relatively consistent as a result of our most recent optimization, your organic reach may be impacted
- The more engaging your content, the lower the impact this optimization should have on your reach going forward
- Feed is optimized to show users the posts they are most likely to engage with, where engagement is defined as clicking, liking, commenting, or sharing the post – or in the case of offers, claiming the offer.
- Posts that are more likely to be engaging tend to appear higher in feed. Some of the strongest factors that influence this are how engaging an individual post has been for other users who have seen it, and how engaged a user has historically been with other posts they’ve seen from that page. Feed also takes negative feedback into account, which is the number of people who have hidden a post or reported it as spam.
- Finally, if a page has a piece of content that it feels will be very engaging e.g. A good offer, a great photo, an announcement, etc. then using paid media to “boost” that post to fans in newsfeed can be an effective tool to increase engagement with fans.
Matt, from another agency, also dropped a great insight into the issue along with their Facebook Ad Rep’s statement. //
Since Ogilvy reported that Facebook announced a drastic change to the EdgeRank algorithm on September 20th, we’ve heard widespread complaints that Page Admins are experiencing a decrease in Reach. We decided to look into the data ourselves to see if this is, in fact, a widespread issue.
Over time we’ve seen Reach slowly decrease as more Pages, and more users, create content. The more content that is posted to the news feed, the less likely your Page’s content will reach your fans. Facebook has also been rumored to provide 80% organic content, and “20% paid content in the form of sponsored stories” for Pages. So, tweaks in EdgeRank can cause fluctuations in metrics for brands on Facebook.
What Did We Find?
The typical Facebook Page in our data set was experiencing 26% Organic Reach the week before the 20th. The week after the 20th, these same Pages were experiencing 19.5%. These Pages lost approximately 6.5% of their Reach after the 20th.
Other Page Admins suggested that they have received detrimental losses to their Viral Reach. When examining Viral Reach Per Post Per Fan, we found that before the 20th Pages were experiencing 0.69%. After the 20th, Pages suffered a drop to 0.38%. This resulted in a nearly -45% change for Pages’ Viral Reach. Over the long term, this could have serious negative impact on Pages’ ability to enter Fan’s Friends’ news feeds.
We looked at a few other metrics to provide a bit of perspective. These same pages were experiencing 0.59% Engagement/Fan before the 20th. After the 20th, these Pages’ Engagement/Fan decreased to 0.49%. However, it’s difficult to determine if this loss in Engagement was related to the decrease in Reach.
We also looked at Virality over these same time periods. Before the 20th, pages were seeing 1.91% Virality, and after the 20th, they were seeing 2.04% Virality. As Facebook defines Virality, it implies how your content is able to engage your audience. This change in Virality, which resulted in a slight improvement, seems to imply the content was similar in terms of engaging the people who saw it.
What Does This Mean?
Overall we found that Organic Reach (-25% change), Viral Reach (-45%), and Engagement/Fan decreased (-17%), while Virality marginally increased (+7%). With both Organic and Viral Reach dropping fairly drastically, it seems to indicate that visibility was possibly reduced within the news feed. This decrease in Reach may have also caused the decrease in Engagement/Fan averages. Virality was relatively maintained, which seems to suggest the quality of the content was roughly the same (or was able to engage people who saw it).
It should be noted, that this is a fairly small sample size and should be monitored further. We’ll be monitoring these changes moving forward.
How Did We Analyze The Data?
We analyzed approximately 3,000 Facebook Pages that posted content in both the week prior to September 20th, and the week after. The Pages varied in size and posting frequently.