How does Facebook's news feed AI decide which posts to show you and when? Facebook has released a full explanation and, as you may assume, the system is very complex.
If you're interested in hearing the Facebook engineering team's perspective, you'll get to learn more about the way the multi-layer algorithm code is built.
"Designing a personalized ranking system for more than 2 billion people (all with different interests) and a plethora of content to select from presents significant, complex challenges. This is something we tackle every day with News Feed ranking."
Heads up, be prepared for a lot of this: Vijt = wijt1Yijt1 + wijt2Yijt2 + … + wijtkYijtk.
Facebook's tech team and newsroom do a nice job of explaining the algorithm methodology and news feed ranking system from a higher level.
You'll see in the video how the algorithm analyzes our behavior and activity, constantly taking in new information and recalculating, to determine what it thinks we want to see next. Everything is individualized and team building this algorithm is conducting interviews and doing lots of research to better understand what our actions really mean.
All of this is interesting to think about, but what does it mean from a marketing perspective? Learning how the algorithm works will help you strategize better.
There aren't any tricks you'll learn to cheat the system here, but the algorithm is designed to deliver value. Facebook determines this value by learning what people like and dislike through their interactions with content.
That means that there's only one real way to get your content in front of your audience on Facebook (and Instagram).
- Know your audience and what they value.
This takes time, listening, research and a careful analysis of your audience.
- Deliver that value.
Once you truly understand what your audience wants, it's time to deliver them high-quality, engaging content.
To learn more about how Facebook's research team is thinking about AI, check out updates from them here.