What’s the Ideal Length for Online Video?

The difference 68 minutes and 56 seconds makes

Blaise Zerega
4 min readMay 29, 2014

It’s a question I hear often from conference and event organizers.

“What’s the ideal length for an online video?” they ask.

When I pause before replying, they whisper furtively, “You know, to get the most views?”

I give them a simple answer, the fix they sought. Then I add, “But it’s more complicated than that.”

Consider the video of Sir Ken Robinson at TED: How Schools Kill Creativity. Posted in 2006, it has been viewed more than 26,581,643 times*, making it the most viewed TED Talk to date.

http://www.ted.com/talks/ken_robinson_says_schools_kill_creativity

Now consider a video of Robinson at the Los Angeles Public Library: A New View of Human Capacity. Posted in 2009, this video program has been viewed more than 84,243 times*.

http://fora.tv/2009/01/29/Sir_Ken_Robinson_A_New_View_of_Human_Capacity

The Robinson TED video is 19 minutes and 24 seconds long; the Robinson LAPL video is one hour, 28 minutes and 20 seconds long. That’s a difference of 68 minutes and 56 seconds.

Do the math. The video that is about 4.5 times as long, received just a tiny fraction —.003 to be precise—as many views. Even if we accounted for the three year head start given the TED video, the conclusion would be the same: Shorter videos always win, right? Not so fast.

Consider yet another Robinson video: This one an animated, condensed version of a talk he gave at the RSA: Changing Education Paradigms. Posted in 2010 on FORA.tv and on TED.com, this video has received more than 1,405,068 combined views* from both sites.

http://fora.tv/2010/10/14/Ken_Robinson_Changing_Education_Paradigms

This video is 11 minutes and 40 seconds long. The shortest of the three. If we accounted for it being posted four years after Robinson’s TED talk, and duration were the sole determinant of views, then this program, the shortest of the three, should have the most views over time, right? But even without a calculator we know it doesn’t.

A direct, causal connection between video duration and number of views simply does not exist. I would guess you suspected this from the outset. My apologies for presenting you this necessary red herring**. It’s in service of a broader inquiry.

However, a correlation between video duration and number of views does exist, no matter how difficult to prove out.

Look to the world of sports. It’s the brief highlights showing a slam dunk by LeBron James, a great pass by Colin Kaepernick or a homerun by Buster Posey that garner the most views, not the full video transcripts of the games themselves. Or better yet, something a bit more highbrow than professional sports: Shakespeare’s Hamlet. As a video excerpt, the soliloquy “To be or not to be” from Act 3, Scene 1, would have more views than a video of the full play.

Further, with video consumption habits changing rapidly, one person’s 117 minute delight on a flatscreen TV is another’s punishingly long Gunga Din on a poorly-charged mobile phone. Or take House of Cards and the advent of binge watching, where the “ideal length” is determined by the viewer. One hour per episode if that’s their desire, or perhaps eight hours if they’re going to watch an entire season in a single viewing.

“Ideal lengths” for feature films — two hours, TV sitcoms — 22 minutes, and TED videos — 18 minutes, were developed largely to suit the dominant viewing modes in use: movie theaters, broadcast television, personal computers, respectively. New “ideal lengths” for emerging content forms will be driven by both the consumption devices and the consumption desires of the viewers. Is Vine establishing an “ideal length” for mobile video? Are connected TVs best for lean-back viewing or are they better for lean-to viewing? What’s your “ideal length” for an online video?

A simple rule: If you want your video to get its maximum number of views, then make it as short as possible.

Shorter videos will generate more views than longer videos. Why is that? Shorter videos are shared socially more often; they’re more mobile-friendly and they’re well-suited for short attention-spans.

While implicitly obvious, such factors as promotion, distribution, timing and so on, can have greater influence on a video’s popularity than its length. The massive views accorded Robinson’s TED video are due in part to the marketing juggernaut that is TED.com, certainly as compared to the modest online audiences for the other two videos, drawn from the Los Angeles Public Library and the RSA and FORA.tv.

This simple rule, however, is disingenuous. As disingenuous as the very question about the ideal length for an online video. The ideal length should be governed by the narrative arc of the content, the complexity of the subject matter, the audience’s appetite for engagement and even the target mode of viewing — web, mobile app, or connected TV.

In the case of Robinson’s TED video, 19 minutes and 24 seconds is the ideal length and the view count proves it. If a three minute excerpt of the video — call it a highlight — were to be published, this shorter video invariably would attract a higher number of views. But to publish such a short video would be less than ideal. It would dishonor Robinson’s presentation and his message. It would also reduce TED’s motto to “Snippets of Ideas Worth Spreading.”

For conference and event organizers seeking a simple answer, I reply with certainty, “19 minutes and 24 seconds.”

Then I sit them down and explain, “But it’s more complicated than that.”

And that’s the answer they’re really seeking.

--

--

Blaise Zerega

Editor in Chief at All Turtles| VentureBeat, FORA.tv, Conde Nast Portfolio, WIRED, Red Herring, Forbes I Michener Fellow at Texas Center for Writers| @BeeZee