Time for another data deep dive. This week we're taking a look at how video duration impacts high level engagements. For this study we analysed videos posted on Facebook or YouTube in the last 30 days with more than 10k engagements - 24,000 videos in total. With that kind of data, we show you how to capture the best engagement through optimal video length, and what platform will work best for your content type.

Optimal Video Length: YouTube vs. Facebook

This data was just a WOW moment for us. We expected YouTubers to be making longer content, but according to the data from the past 30 days, videos on YouTube are nearly 10X longer than Facebook.

Using data from Tubular, we can confirm that creators on YouTube are making not just longer content, but WAY longer content. Some of this is certainly a function of YouTube rewarding longer retention, but it really points to the type of content each platform currently values most. Facebook tends to value short, flash in the pan type content to keep you moving and tied in to the feed, while YouTube is more focused on serving you longer, higher quality content to keep you on the site longer.

Although we keep looking at YouTube and Facebook as competitors, they may actually be on opposite ends on the video spectrum not just in duration, but in what metrics indicate a video’s success as well.

PlatformAverage of Duration (seconds)

Facebook Vs. YouTube & Views Vs. Engagement

Our previous deep dive which was specific to Buzzfeed, showed that video posted natively to Facebook tends to generate more engagement. But in this study, we see that YouTube has a lower views to engagement ratio than Facebook, which means it takes less views to generate the same engagement on YouTube.

Our gut reaction is this as an indicator for view quality. YouTube fans continue to tout the value of a YouTube view compared to Facebook and this data would tend to support that claim. It may also suggest that the type of engagements occurring on Facebook the most, namely likes, are more freely given and less indicative of meaningful engagement.

PlatformAverage of Views:Engagements

But not so fast. Analyzed another way this data would suggest that an engagement on Facebook can generate more views for a given video. It all depends on how you view video and what comes first, the view or the engagement.

For YouTube, a combination of views and engagements can help it rise in search, so there is a distinct partnership in these numbers. On Facebook, however, views are somewhat less indicative of the performance of a video - the almighty engagement actually causes it to propagate across multiple feeds, granting the video more views.

What this bit of data may really suggest is that Facebook videos may be better judged on the number of engagements they get, while YouTube seems to have focused more on the quality of their view counter.
Let’s revisit the first chart real quick, but add in the average view count as well. While Facebook is beating YouTube in views more than 2:1 on these videos, YouTube is only losing 3:4 in the engagement department. To me, this suggests that Facebook has a lot of what I'll call "empty" views. But that also suggests that Facebook is clearly getting the traffic and may even be beating YouTube in raw traffic.

PlatformAverage of Views:EngagementsAverage of Views

Looking at this data broken down by duration categories, YouTube has clearly trained viewers to not only stay around, but interact a bit before they leave. On the other hand, Facebook seems to trend towards getting the video in front of somebody, getting a like/share and moving on.

As for optimal video length, as duration increases, the likelihood of high level engagement goes down on Facebook, where YouTube is just the opposite. YouTube engagement seems to perform the best early in the video, like Facebook, but remarkably they get more high engagement videos the longer they go. There is a definite plateau around the 3 minute mark for YouTube, which has often been used as a marker for the general attention span of a YouTube viewer. As viewers watch past the 5 minute park on YouTube, it seems that the interaction drops quite a bit. This could point to a larger issue of actual viewer retention. How mentally engaged is a viewer after a certain point?

PlatformAverage of Views:EngagementsCount of Video_Title
Grand Total28.4922583

Looking at this chart would seem to suggest the best place for a call to action is right around the 30 second mark for a video, or wait until it is just about over. Personally I’d advocate doing both if it makes sense for your content.

If you want high engagement, the sky is the limit on YouTube. The videos analyzed averaged nearly 15 minutes in length for YouTube while Facebook was just shy of a minute and a half. But if you are looking for short form content, the fleeting nature of the Facebook feed may provide you with a better place to get massive exposure. Either way, we may have been measuring Facebook video all wrong this whole time. The secret sauce may lie in their engagements and not in their views. The engagements ultimately lead to the exposure that those in video desire.

Optimal Video Length Takeaways:

  • Videos posted natively to Facebook generate more engagement - but it takes less views for YouTube videos to generate the same engagement rate
  • Engagement rate for Facebook video is a more reliable metric than view count
  • The most engaged videos on YouTube are nearly 10x longer than the most engaged videos uploaded to Facebook.
  • Short-form video content tends to do better in terms of engagement on Facebook.
  • The longer the video on YouTube, the more viewers will engage with it.
  • Best place for a CTA on a YouTube video is around the 30 second mark.