Seeing more video in the news feed? You're not the only one. Facebook said on Wednesday that video activity on its platform grew dramatically in 2014, bolstering its ambition to become a video platform contender to rival YouTube.
In a blog post Facebook said the number of video posts per individual user grew 94% on average in the US and 75% globally. And exposure has increased even faster. People saw 3.6 times more video in the news feed in 2014 than they did in 2013. And they now generate 1 billion video views collectively every day, 63% of it on mobile devices. (YouTube viewers meanwhile watch 4 billion videos a day, according to its latest public numbers from 2012. Presumably the number is much higher today. )
All that video media is like gold bullion for Facebook, which is eager to get users comfortable seeing video ahead of an upcoming increase (launched officially in December) in the number of auto-play video ads that it serves. The increase also lets Facebook prove to brands and agencies what it can do from a "sight, sound, and motion" standpoint.
But Facebook says this was not a case of warming the soup kettle slowly to keep the frog inside. "News Feed is unique to each person," a spokesperson tells AdExchanger. "So if you watch a lot of videos, we’ll show you more since it’s a signal that you like videos," the representative said.
But that signal may not be as clear as Facebook would like. Since videos uploaded to Facebook play automatically, it's fair to ask how the company can be sure its users really do love seeing videos and want to see more. Could it be that many users, scrolling idly downward, are arrested by a video that plays automatically – thereby creating a false signal that tells Facebook's algorithms to serve more video? The human brain is wired for motion after all.
Ultimately it's an academic question. Whatever the agent of the video increase, Facebook or its users, video has increased dramatically and will likely continue to do as the platform increases as a sharing nexus for moving pictures.