How to make a training video people actually finish

1. Start with the learning goal and who it is for

Before anyone opens a camera, write down what a person should be able to do after watching. Not “understand the expense system”, but “submit an expense with a receipt and send it for approval”. A goal you can watch someone do is a goal you can teach in a video.

Then write who is watching. A new hire on their first week, an experienced colleague moving to a new tool and a contractor who needs the safety rules are three different audiences. They know different words, need different amounts of context, and watch in different places: at a desk with sound, on a phone on the shop floor, between calls.

• The goal: one sentence starting with a verb. “Set up…”, “Handle…”, “Check…”.
• The audience: their role, what they already know, and where they will watch.
• What they should do right after: try it in the system, pass a short quiz, ask their manager.

2. Split the topic into short modules

One long training video is hard to watch and harder to come back to. Someone who needs to remember step four should not have to scrub through an hour to find it. Split the topic into modules, each with one goal and a title that says what it teaches.

1. List every task the learner has to be able to do.
2. Put them in the order they will need them, not the order the expert thinks of them.
3. Make each task its own module. If a task has a long setup, the setup can be its own module too.
4. Name each module as an action: “Approve a request”, not “Approvals”.

Short modules are also cheaper to keep up to date. When the system changes, you re-record one module instead of the whole course.

3. Write a script or an outline for each module

Not every trainer wants to read a script, and a word-for-word script read badly is worse than a good outline spoken naturally. Either way, write something down. A simple template that works for most modules:

1. Title: the action this module teaches.
2. Why it matters: one or two sentences. What goes wrong if people skip it.
3. Before you start: what the learner needs open or ready.
4. Steps: numbered, one action per step, in the words the learner will see on screen or on the equipment.
5. Common mistake: the one thing people usually get wrong, and how to spot it.
6. Recap: the steps again in one breath.
7. Next: what to watch or do after this module.

Read the script out loud before recording. Sentences that look fine on paper often run too long when spoken. If the trainer stumbles on a line, shorten it. Keep the vocabulary the learner's: if the button says “Submit”, say “Submit”, not “send it off”.

4. Record a trainer, or use a session you already have

There are two common starting points: you record a trainer explaining each module on purpose, or a live training session was recorded and you want to turn it into modules. Both work. The first gives cleaner material; the second already exists.

If you are recording on purpose:

• Sound first. A clip-on microphone close to the trainer matters more than the camera. People forgive a soft picture, not a voice they can't hear.
• Light the trainer's face from the front or the side, with the window in front of them rather than behind.
• Record each module as its own take, and let the trainer start again from the top of a step when they stumble rather than patching mid-sentence.
• Film the hands, the machine or the whiteboard separately as well, so you have something to show while the trainer explains.
• Leave a second of silence before and after each step. It makes the cut much easier.

If a live session was recorded, check what you actually have before planning anything: is the trainer audible, are the questions from the room audible, and can you see what is being demonstrated? That decides which parts can carry a module and which parts need a title, a screenshot or a short re-record.

5. Turn a recorded training session into modules, transcript first

A recorded session has all the steps in it, but in the wrong order and between questions, tangents and “can everyone hear me?”. Watching it end to end to find them is what makes this job slip for weeks. Read it instead: get a transcript from your editor or recording tool and work on the text.

1. Put your module list next to the transcript.
2. For each step, find where the trainer explains it most clearly. It may be in the Q&A rather than the main talk; an answer to a real question is often the plainest version.
3. Mark the start and end of each explanation at a sentence break, and note the timecode.
4. Arrange the marked passages in your module order, as plain text, and read them through. If a step doesn't make sense on its own, find a better passage or plan a title to bridge it.
5. Note any step nobody explained on camera. Re-record it or cover it with on-screen text; don't stitch words together to make the trainer say it.

This is the same idea as a paper edit: decide the film in text, where changes are free, then cut the picture to match.

6. Add visuals, B-roll and on-screen text

A trainer talking for the whole module is fine for a minute and tiring after that. Show the thing being taught whenever you can, and keep the trainer's voice running underneath.

• B-roll of the task itself: hands on the equipment, the form being filled in, the place where it happens.
• Close-ups or zooms on the detail that matters: the button, the gauge, the label. Hold them long enough to read.
• A title at the start of each module with its name, and a short title for each step number so people can follow along and find their place.
• A lower third with the trainer's name and role the first time they appear.
• A recap card at the end with the steps as a short list, held for a few seconds.

Keep on-screen text short: a few words, not sentences. If it needs a paragraph, it belongs in the written material that goes with the video.

7. Captions and accessibility

Captions are not optional for training. People watch on the shop floor, in open offices, in a second language, or with a hearing impairment, and some organisations have accessibility rules that require them. Captions also make the exact wording of a step easy to read twice.

• Generate captions from the transcript, then read every line. Product names, internal terms and acronyms are where automatic transcription goes wrong.
• Keep to two short lines on screen, broken where the speaker pauses.
• Export a subtitle file (SRT or WebVTT) as well as, or instead of, burning captions into the picture. Most learning platforms accept one, and viewers can turn it on and off.
• Keep a plain-text transcript of each module. It helps people who prefer reading and makes the content searchable.
• Don't rely on colour alone to show something important, and say out loud what is being shown on screen.

8. Keep modules short, and review with a subject expert

There is no correct length, but a useful rule is one task per module and nothing that doesn't help with that task. If a module runs long, check whether it is really two. Watch it yourself as a learner: the moment your attention drifts is usually a place to cut.

Before publishing, have someone who knows the subject watch every module. Video editors and trainers are the wrong people to catch a wrong step; they know what was meant. Ask the reviewer to check:

• Every step is correct and in the right order for the current version of the system or procedure.
• Nothing safety-critical or legally required was cut for length.
• Names, numbers and terms on screen and in captions are spelled right.
• The trainer and anyone else on camera are fine with the video being shared this way.

Then show one module to somebody from the actual audience and watch them try the task afterwards. Where they hesitate is what to fix.

9. Publish in your LMS or intranet, then use the checklist

Most organisations publish training videos in a learning management system, on the intranet or in a shared video library. The details depend on your platform, but the same things help everywhere: one video per module, a clear title starting with the action, a one-line description of what it teaches, the subtitle file uploaded beside it, and the modules grouped in learning order. Add a date or version so people know whether it is current.

Before each module goes out:

1. The module teaches one goal, and its title says which.
2. The first seconds say what the learner will be able to do.
3. Every step is shown, not only described, wherever that is possible.
4. Captions are checked line by line, and a subtitle file and transcript are ready to upload.
5. The sound is clear and at a steady level from start to finish.
6. A subject expert has signed off on the content.
7. The module ends with a recap and the next step.
8. Someone is named to update it when the system or procedure changes.

For a recorded session that should become a short video for new hires, see turning a recorded training session into a training video. The same approach works for internal updates and for webinar recordings.