There was a time not so long ago when the world looked upon web-based video streaming as a curiosity more than a legitimate way to view content.
If there was ever a need for proof that that world is long gone, YouTube has delivered. A recent article in The Wall Street Journal reports that YouTube viewers are now watching more than one billion hours of video daily, a ten-fold increase since 2012.
Personalization is Key to YouTube Success
Marketers who rely on video advertising for their messages would be smart to give YouTube a try. Nielsen reports that Americans watched an average of 1.25 billion hours of live and recorded television last year, but that figure is dropping rapidly.
The next closest competition for YouTube video are platforms Netflix, which reported 116 million hours watched daily, and Facebook, with 110 million hours. YouTube is only gaining traction, largely due to parent company Alphabet’s dedicated effort to increase viewer attention.
Although anonymous sources are reporting that the $4 billion of revenue for 2014 helped YouTube to just break even, YouTube CEO Susan Wojcicki stressed that growth was the company’s first objective. It would seem that much of that money has been invested in improving the machine intelligence that drives YouTube’s suggestion engine.
YouTube Getting Smarter
There have been some rocky roads in the past, but YouTube’s AI is getting smarter and better at solving problems. For example, when YouTube first started suggesting videos that visitors might enjoy, content with misleading titles or preview images were often included in the results.
Now, Alphabet’s video streaming unit utilizes deep neural networks developed by Google Brain to make predictions that keep watchers watching. Instead of just feeding more videos like the one a user just finished watching, the neural network can suggest areas of related interests to try to better read each person’s preferences.
Google Brain is still helping YouTube understand its users, but each small tweak is a step in the right direction. The fear of many is that YouTube’s learning algorithm will end up creating personalized content experiences that are so unique to users as to eliminate the shared experience television once was, but the real likelihood of that is slim considering the tendency of users to share videos they really enjoy with friends and family.