A question that comes up a lot among learning professionals making digital content is, “What’s the right training video length?” Is it two minutes? Five minutes? Or even 10?

It’s not an easy question to answer because at first thought, something short like two minutes sounds about right. But as you consider it, a declared ideal duration like two minutes doesn’t seem so clear.

Ten years ago, I ran a workshop for newspaper editors in the south of France on digital storytelling. At lunch, over a chilled glass of rose, two editors from France and Spain asked the very same question.

I had my own views—I was comfortable with two minutes. But I decided to check in with a major European newspaper I had worked with about their analytics.

Back then, video was viewed with media players like the Real Player. This newspaper’s analytics showed people closed their media players at about one minute and fifty-one seconds.

Of course, the data only told them when the media player was closed. Not whether the user had watched all of it. But I was happy to conclude that keeping videos under two minutes was a good idea.

But is long video bad?

A few weeks later, at a conference in London, my new rule was thrown on its head. The conference was about 360-degree storytelling. That’s the term we used in the UK at the time for content across platforms and modalities—the concept is like transmedia storytelling.

During a panel discussion, the “how long should your video be” topic came up again. Folks from independent production companies behind some of today’s biggest TV shows attended this conference, as well as some folks from broadcasters like Channel Four and ITV.

During a panel discussion, one prominent producer said he didn’t think it mattered. I disagreed. But then he showed stats of what was then a top rating TV show getting hundreds of thousands of online views, lasting longer than forty minutes.

These contradictory pieces of data made it tough for anyone wanting a concrete rule. Until I realized the obvious. The newspaper analytics were for factual content. The TV statistics were for entertainment. They were different. I could live with these variables.

However, the variables don’t end at genre. A few years back, YouTube published figures showing people watched video for about two minutes on their phones and just over four minutes on tablets.

So, genre, device, and all sorts of other variables can affect how long people watch online video. So, what are we to do?

Key media principle—short and simple

I think aiming for short is exactly the right thing to do. However, we need to be smart about what short means because it depends on the kind of content and, for us in the learning world, it depends on the topic.

As a rule, broadcast professionals are taught to keep everything short and simple. As a talk show host in the nineties I was always told to keep my on-air interviews less than three minutes.

Talk to any newspaper journalist and they’ll tell you their formative newsroom experience was having an editor slash their 400-word article to 250. So, short is a traditional media discipline.

It dovetails nicely into today’s age of impatience. People want learning now, but not all of it. Only the bits that are relevant to their immediate needs. (This is why terms like microlearning—albeit a vaguely defined and evolving term—resonates with so many people.)

So, if short is variable, how do we define it?

As short as you can make it

Having had one foot in the world of learning for the last few decades, and the other foot in professional media production, I humbly suggest training videos should be as short as they possibly can be. Yes, that’s vague, but we can be more concrete.

Good training videos start, as does good instructional design, with a clearly defined learning objective. If you’re an instructional designer your instinct will be to construct the objective along the lines of Mager’s three-part objective: action, condition, and standard.

The answer to how long is short lies in the learning objective. The learning objective needs to be a yardstick for measuring whether the video is too long by asking the question of every element in the video, “Do I need this for my learner to be able to perform this task?”

Look at your script—what words can you take out without changing the meaning? Look at the shots and text graphics—are they necessary? You only know by asking whether they help achieve the learning objective.

Good videographers critique every element of their video. The shots, the graphics, the music, words, and the effects. And if they can’t honestly prove to themselves that an element will constructively achieve the objective, they remove it.

Short is about being disciplined and not cramming anything more into your video than you need to. And the learning objective is the device you use to determine if an element makes the cut.

I suspect that someone might be thinking right now, “Did I just read 800 words to find this out—it’s so obvious!” And they’d be right. It’s simple. But often the simple and obvious are overlooked.

How many training videos have you seen that are loaded with more shots, words, and graphics than are needed to convey the message? Crammed full of self-indulgent narrative and special effects?

So, when people ask me what’s the right training video length, I resist the temptation to offer a formulaic answer or specific time. We need to be smarter than that. Instead, I say it needs to be as short as you can make it to achieve the learning objective.