How Does the Facebook Algorithm Work?
If you're still using Facebook (FB) to promote your business then you might have already noticed a few changes to the organic reach of your posts since FB updated the algorithm a couple of months ago.
The changes were made by FB in order to 'prioritise “meaningful” person-to-person interactions among friends and family over posts from Facebook pages'. This means fewer video and public posts from pages in the news feed.
BUT,
how does the algorithm decide what to put in your newsfeed or not?
Here's an overview of how it works:
1) Ranking your content:
The algorithm uses a 4-step process to rank your post on the news feed:
Inventory - looks at all the stories posted by your friends and pages you follow.
Signals - HUGE range of influencing factors that are used to make an educated guess about what posts you might be interested in. These include time of day, how you're accessing FB, what you've looked at before.
Predictions - what are you likely to comment on, spend time reading, share, etc.
Score - the post is given a relevance score based on it's calculations of the predictions and probabilities.
FB have admitted that they really have no idea what posts you'll be interested in seeing and that this is an educated guess! However, this ranking process occurs EVERY time you access FB.
2) More about Signals:
“The most critical input into what you see in your news feed is who you decide to friend and what [pages] you decide to follow in the first place.”
Of course it is, because you're the one who decides to friend someone and follow a page!
However, there are a few additional considerations:
Who posted a story:
Frequency of posts from that person or publisher
Previous negative feedback on an author
Engagement:
Average time spent on content
Overall engagement a post already has
When the story was posted:
Friend tags
A recent comment from a friend
Story type
Completeness of page profile
Posted from a friend or page
How informative the post is
3) The Predictions - The algorithm uses all of the above to predict the likelihood that:
you will click on the post.
you will spend time spend time reading the post.
you will like, comment, and share the post.
you’ll find the post informative.
the post is click bait.
the post contains nudity.
Because the algorithm looks at the information pertaining to your previous use of FB, each relevance score is specific to you.
Watch out for my next blog which will talk about Facebook Zero and how it will affect you as a marketer!