In SocialBakers engagement study, they found that both options performed similarly to each other when it came to Likes and Shares. From January 22-27 they studied (with YouTube links being the obvious choice among admins) 3,684 YouTube links against 458 natively-uploaded Facebook videos. There were double the comments with Facebook videos compared to YouTube links, as illustrated by the graph below:
But where there was a 40% engagement difference between Facebook videos and YouTube links. Now, the reason SocialBakers gives for this is entirely reasonable: the videos show up not only in the posts but in the Fan Page video gallery, and people can watch the videos right there on Facebook without it having to jump over to YouTube. This is something I’ve actually encountered with Facebook when it comes to the mobile site and YouTube links. It can often take forever or not connect with YouTube at all. The ability to watch a video completely on-site keeps viewers on the page and they don’t have to jump around to interact.
Further, in a study conducted from December 4th to March 3rd, SocialBakers found 10 times the viral reach with Facebook video versus YouTube links (YouTube still led in actual use, 4,371 to 554):
SocialBakers avoided paid media here, as they wanted to measure the reach of a video that is more “honest” about how they reach consumers.
I think there’s no doubt that what SocialBakers found here had a lot to do with the ease-of-use for Facebook users to watch the video completely on-site as opposed to being re-directed to YouTube. Obviously, it’s easier for those who share video to simply upload it to YouTube, which is what a brand/marketer is going to do anyway, and then provide the link which is very easy itself. However, getting people to jump around, even though it’s not much, makes a pretty sizable difference.
Thanks to SocialBakers for their study!