YouTube’s New Suggested Video Algorithm: New and Slightly Improved?

YouTube has recently unveiled a new change-up to its algorithm of suggested videos; namely, the video-sharing website has adjusted the algorithm to give precedence to “engaging content”. This decision is part of efforts to improve the YouTube experience: prior to this, the video algorithm was a source of frustration to as it presented misleading thumbnails that when clicked, did not deliver the anticipated content. The change-up to the algorithm is meant to address this. Let’s take a closer look at this decision of YouTube’s

Engaged Content:

The main question this change asks is: what exactly determines “engaged content”? A YouTube help article provides this clue: “the viewer’s total watch time is what matters for the algorithm.” In other words, videos that are longer, for example, 3 minutes in length, would be considered more ‘engaging content’ than shorter videos, say those that are 30 seconds long; YouTube’s comment indicates that longer videos would therefore receive greater emphasis within the new algorithm.

Potential Problems:

This move is certainly a step in the right direction, offering a much welcome alternative to the previous practice that provided related videos according to the number of views – problematic as a ‘view’ can be 10-15 seconds. However, the move is problematic if this total watch-time is the only determining factor for the algorithm. What should also be considered in the process is the percentage of time viewed.

When creating content, it is understood that the video should engage the audience for the majority of the video; even if users do not watch until the very end, the video would (read: before these new practices) be considered a success if the audience was engaged for about 90% of the time. A 40-second video that users watch 36-seconds of should be considered ‘engaged content’, no? However, if this algorithm is indeed based only on watch-time, a longer video where the majority of users only watch 50% – say 2 minutes of a three minute video – would be emphasized more than our 90% example, and thus have a higher likelihood of coming up in the users’ suggested videos.

Other Factors

However, even if YouTube’s new algorithm is based solely on overall time viewed, there are other social factors built into YouTube that will also affect the ‘engaged’ factor, and have the ability to elevate videos in the recommended video section. These social factors include video likes, video comments, and, of course, social-sharing across networks such as Facebook and Twitter.

Conclusion:

YouTube is clearly making efforts to improve its service, and this change to its suggested video algorithm is more progressive than problematic. As it stands, the only major flaw is that the percentage of time viewed is not considered alongside overall time viewed; should this be tweaked, then the algorithm would prove to be a definite improvement.