How to Make Educational Videos That Actually Teach

I have watched a lot of educational videos die in the first fifteen seconds. Not because the person on screen was wrong, and not because the lighting was bad. They died because the video tried to cover a whole subject instead of answering one thing the viewer actually came for. So before I talk about cameras or editing, I want to reset what “good” means here. A good educational video is not the one that looks the most polished. It is the one someone finishes, and then does the thing you were teaching.

I run a creative studio, and educational content is one of the requests I get most often. Product teams want tutorials. Marketing teams want explainers. Founders want to teach the thing only they understand well. The process below is the one I actually use with my embedded team, stripped of the parts that waste time.

What actually makes an educational video work

Everything good starts from how people take in information, not from production value. The research on this is unusually clear. In one of the largest studies of its kind, MIT and Harvard researchers analyzed 6.9 million video-watching sessions on edX and found that engagement drops off sharply once a video passes roughly six minutes, no matter how good the content is (Guo, Kim, and Rubin, 2014). That single finding should change how you plan almost everything.

There is a good summary of the underlying principles in Cynthia Brame’s paper on effective educational videos, which pulls together the cognitive-load work into three levers you can actually pull (Brame, 2016):

• Keep the load light. Cut anything on screen that the viewer does not need to understand the point. A music bed under a voiceover, a lower-third that stays up too long, a busy desktop wallpaper behind a screen recording. Each one competes with the lesson for the same attention.
• Signal what matters. Point the viewer at the important thing with a highlight, an arrow, a zoom, or your own words (“watch this part”). Do not make them guess where to look.
• Break it into pieces. One video should teach one idea. If you have five ideas, you have five videos.

I keep these three taped to the top of every script document. When a video feels bloated, one of the three is almost always the reason.

Start with one question, not a topic

“How to use our dashboard” is a topic. Nobody searches for a topic and feels a need. “How do I export a report as a PDF” is a question, and it has a person behind it with a specific frustration. I start every educational video by writing the exact question a real viewer would type or ask out loud. If I cannot write it in one sentence, the video is not scoped tightly enough yet.

Then I write the answer in one more sentence, before any script. If the answer needs a paragraph, I have probably bundled three questions together, so I split them. This sounds almost too simple, but it is the step people skip, and skipping it is why so many educational videos wander.

Where do the questions come from? Three places I trust:

• Your support and sales conversations. The questions customers ask over and over are your best briefs. Someone is already telling you what they do not understand.
• Your own experience using the thing. Sit down and do the task you are teaching. The moment you get stuck or confused is the moment worth filming.
• Search. The phrases people type into Google and YouTube are literal questions. Read them as scripts waiting to happen.

Script it so a person can follow it, not read it

A script for video is not an article read aloud. Written language is dense and formal, and it sounds strange coming out of a mouth. I write the way I would explain the thing to a colleague at my desk. Short sentences. One idea per line. A clear “here is what we are doing and why” before every step.

My structure almost never changes:

1. The hook (first 10 seconds). State the question and promise the answer. “You want to export a report as a PDF. Here is the fastest way.” No logo animation, no long intro. The MIT and Harvard data shows how little patience viewers have, so I spend the opening earning the next thirty seconds.
2. The steps. Numbered, in order, one action at a time. I say what to click before I click it.
3. The check. Show the finished result so the viewer knows what success looks like.
4. The next step. One thing to do now. Not five.

I read every script out loud before we shoot. If I stumble on a sentence, the viewer would too, so I rewrite it. Half the time the fix is just cutting a clause I only added to sound complete.

The length that holds attention

Here is the part people fight me on. Shorter is almost always better. Because engagement falls off around six minutes, I treat that as a ceiling, not a target. Most tutorials I make land between two and four minutes. If a lesson really needs more, I do not make it longer. I make it a short series, and each part answers its own question.

Length is not about attention spans being broken. It is about respect. A viewer who came to learn one thing does not want to sit through four unrelated things to get it. When I cut a seven-minute draft down to three, the video almost always gets better. The three minutes that survive are the three the person actually needed.

100%

Under 6 min

50%

9 to 12 min

20%

12 to 40 min

Viewers finish short lessons but abandon long ones, so a video that runs past six minutes gives back most of its own length in lost attention.

Source: Guo, Kim & Rubin, edX study of 6.9M sessions, ACM L@S 2014

Watch how the teaching choices show up on screen

Reading about this only gets you so far. This short piece from Edutopia walks through the small production and structure decisions that make a learning video land, and it is a useful reference for the ideas above.

Notice how much of it is about clarity and pacing rather than gear. That is the point I keep coming back to.

Filming without overthinking the gear

The equipment question is where costs run away and results do not follow. For most educational videos, the picture matters far less than the sound and the clarity. I would rather watch a plainly lit talking head with clean audio than a beautifully shot video where I cannot hear the person.

So my order of priority when filming:

1. Sound first. Get a lavalier mic or a decent USB mic close to the speaker. Record in a quiet room with soft surfaces (a room with a rug and curtains beats an empty office, which echoes). Bad audio makes people leave faster than anything else.
2. Light second. Face a window during the day, or use one soft light in front of the subject. Avoid a bright window behind the person, which turns them into a silhouette.
3. Framing third. Eyes in the top third of the frame, a little space above the head, camera at eye level. That is most of it.

For software tutorials, skip the camera entirely and record the screen. A clean screen recording with a calm voice explaining each step teaches better than a fancy production. What you are demonstrating is the star, not you.

Editing: the parts people skip, and the parts they need

Editing is where a decent recording becomes a video people finish. My editing pass is boring on purpose:

• Cut the dead air. Every pause, every “um,” every moment of nothing. Tighten it. An educational video should move.
• Add captions. A large share of people watch with the sound off, and captions keep them watching. Facebook found that the majority of video views on its platform happen without sound (Digiday, reporting on Facebook data). Captions are not optional anymore, and good editing software can generate a first draft of them for you.
• Signal the important moments. Zoom in on the button you want people to click. Put a text label on the key term. This is the “signaling” lever from the research, and it is the fastest way to make a lesson clearer.
• Trim the intro and outro. No thirty-second animated open. Get to the answer, and end with one clear next step instead of a slow fade.

If you want to go deeper on the mechanics of teaching-focused editing, I wrote a companion piece on how to make instructional videos that covers the step-by-step formats in more detail.

Where each format fits

Not every lesson is a tutorial. Picking the right format upfront saves you from over-producing something that should have been simple. Here is how I think about the common types, what each one is good for, and roughly how much effort it takes.

When someone tells me they want “one big video that covers everything,” I steer them toward the series row. It is easier to make, easier to update when the product changes, and far easier to watch.

Publishing so the video gets found

Making the video is half the job. The other half is putting it where the question gets asked. A tutorial that answers “how do I export a report” should sit next to the export button in your product, in your help center, and on the search result for that exact phrase. I map each video to the moment its question comes up, then place it there.

My short publishing checklist:

• Embed it on the relevant page of your site or product, not just a YouTube channel.
• Write a title that matches the real question word for word.
• Add a text transcript on the page, which helps both search and the people who would rather read.
• Point to it from your onboarding emails and your support replies.

If your goal is reaching new people rather than helping existing ones, the placement and framing change, and I got into that side of it in educational video marketing.

A closing thought before you hit record

The best educational videos I have seen were not the most expensive ones. They were the ones made by someone who understood both the subject and the person struggling with it. That judgment, knowing which detail to cut and which confusion to address, is what separates a video people finish from one they close. Tools can record and caption for you now, and they are useful for the mechanical parts. The teaching decisions are still yours.

If you reach the point where you are making educational video often enough that it needs a real system behind it, that is the kind of ongoing work my studio at Moonb takes off people’s plates. But most first videos do not need that. They need a clear question, a short script, clean sound, and the discipline to keep it short. Start there.