
Facebook wants to become your favorite Internet TV by tracking what a television never could. Today it announced a News Feed algorithm change that will show you more videos similar to ones you expand to full-screen, un-mute or opt to watch in HD, even if you donât Like, share or comment. Those same signals will tell Facebook that a video is enjoyable so the News Feed shows it to more people.
The more Facebook understands which videos are great and which are boring to which people, the faster it will grow its view count, which has already reached 4 billion per day â astounding considering it was at 1 billion per day in September.
Just a few weeks ago, Facebookâs algorithm started factoring in how long you linger looking at posts and videos. Together with the new change, just pumping video content into Facebook wonât get you very far. Clips must be visually tantalizing from their first second, and vivid even when silent as thatâs how they play unless you click. To get more News Feed visibility and reach, theyâll now have to be so interesting that people watch them to the end in full-screen in HD.
The update comes just after Facebook overhauled its video analytics offering for Pages. Instead of having to export data to Excel, Page admins now have a dashboard to track views of all videos theyâve shared, measure auto-plays versus click-to-plays, compare time periods, and see their top videos by performance so they know what to share more of.
Out Tubing YouTube
Big data on viewership could give Facebook an edge on the leading Internet video player. YouTube is built as a host for videos pushed elsewhere around the web, and as a destination where hardcore fans visit to see the latest clips from stars they subscribe to. It thrives by being the only place to watch what you want, so you have to come back, not primarily on its recommendations. Sure, it has a personalized home page. But YouTube is not a feed.
Facebook is a feed, and one that hundreds of millions of people are straight-up addicted to. Many of its 1.44 billion users come daily, hourly even, to see the latest from their friends and Pages theyâve Liked. Itâs not about any one piece of video content, or even video at all necessarily.
And thatâs getting a lot easier thanks to big data. Facebook can crunch insanely specific info on viewership behavior to learn what you want to see. Did people watch for 3 seconds? 11? 96? How many seconds in did they turn on HD? Was it so visually stimulating they watched the whole thing full-screen but on mute, maybe because they were at work or didnât have headphones? Not to mention whether they Liked it, shared it, blurted out âsweet!â or commented an in-depth response to the content.
Multiply that by 4Â billion videos per day in an interconnected network where Facebook knows that a video popular with some people will probably be interesting to people similar to them based on all the biographical and other behavior data it has.
Suddenly it becomes quite clear why data-driven personalization could lead to a News Feed full of the exact videos you want to see, even if you didnât know you wanted to see them. Even if you didnât know you wanted to watch video at all. No Tube necessary.
–TechCrunch