How to Create Scenario-Based Training Videos (With AI)

Have you ever watched someone perform a task once, and thought to yourself, I can do that? Do you execute it flawlessly or do you get stuck on a specific step? Odds are the latter. 

That’s because there’s a difference between knowing how to do something and being able to do it consistently, especially under stress. If I take a First Aid course where I learn about CPR, but never practice doing it, I can guarantee that you wouldn’t want my first time performing it to be an actual emergency. 

That’s where scenario-based learning comes into play, to close the loop between knowing and doing. In the workplace, that means putting employees in realistic situations and having them make decisions where there are consequences. 

What makes scenario-based learning effective

Scenario-based learning is only as effective as the scenario. (Obvious, I know.) But I see this all the time. Someone creates a "scenario" where the right choice is obvious, and there are no consequences for a wrong answer. If a scenario is basically a multiple-choice question dressed up as a scenario, your employees won't learn anything from it.

So here's what you need to do: design a scenario that feels authentic, even if it is fictional. The goal is verisimilitude. It needs to be believable to your employees. The easiest way to do this is to find real examples that you can scrub of identifiable details. Then, identify opportunities to have employees use their problem-solving and critical thinking skills within that scenario.

In my experience, the best scenario-based learning involves ambiguity. There's no clear "right" answer, so employees have to make an informed decision and see the consequences of that decision paired with feedback. These scenarios work because they give employees a safe, risk-free environment to experiment. Ideally, this leads to them being more judicious in their decision-making on the job.

Scenario-based learning can happen in live workshops, immersive environments (e.g., virtual reality), and roleplay sessions. I'll be focusing on video-based training.

How to create a scenario-based training video

If you're using video as your format for scenario-based learning, you'll need a realistic situation, at least one meaningful decision point, and consequences that follow from it.

If your scenario is really just filling an information gap, then a scenario is the wrong format. You likely just want a knowledge check, not a full-blown scenario.

Step 1: Scope your scenario

Your first task is to properly scope your scenario. Here's what that looks like: defining the behavior you want to train, and how you'll measure whether or not the training worked. You can use a template like this:

This scenario is for [specific role] who currently [context or gap]. After completing it, [specific role] should be able to [observable decision or action] so that [business outcome].

Now you can begin to scope your scenario. I like to use the 3C model developed by Tom Kuhlmann. You start by presenting the challenge, then offer choices, and build in the consequences of each choice.

With video training, the complexity of your scenario should be determined by the target length of the video. Generally, I recommend training videos be between 3 and 5 minutes long. While there's no "right" length for scenario-based training videos, research backs this up (in a study of 6.9 million video-watching sessions, MIT researchers found that median engagement time topped out at 6 minutes).

That means if you're designing a particularly complex scenario, you'll need to be focused on the branching options. What drives the scenario forward, and what is adding complexity for complexity's sake?

Step 2: Write the script 

Your script will be guided by how you shaped your scenario. You'll want to document:

• The challenge: why it matters to your audience, and the business impact of getting it right or wrong
• The choices: each one should be plausible
• The consequences: keep feedback actionable and focus on the outcome for the people impacted by the decision
• The branches: how each choice and consequence connects back to the main narrative

If this feels like a lot, I recommend using a prompt like the one below in an LLM to get started. You can also use a template to get started.

‍Step 3: Build your video

At this point, you're ready to build your video or hand off production to a vendor. How you build your scenario-based training video really depends on the tools you're using. I'll walk you through how it works in Synthesia (but if you're looking for tool-agnostic guidance on how to create a training video, I've got you covered here).

With Synthesia, our AI Assistant guides you through the video creation process. You start by uploading your script into our video generator tool to get a first draft. Then, you can make decisions around if you want an avatar, what accent they'll have and how they pronounce any jargon or important words (like your company's name), the branding (colors, fonts, your logo), and more.

Another way to get started in Synthesia is to use a template. If you're new to designing training videos (or just to Synthesia), a template can help you focus on the content instead of the structure. Everything is editable, so you can still make all the design decisions you would otherwise.

With a template, you can see how a training video, like the scenario-based example below, builds around branching. The text is just a placeholder. Use your script to update the narration and what's on screen.

No matter which way you get started, you'll spend most of your time refining the branching components, ensuring the logic flows for each decision: a wrong choice should lead somewhere meaningful, whether that's a consequence scene or a chance to recover. I recommend avoiding a "try again" loop back that sends someone back to the same options.

Step 4: Publish and distribute your video 

When you're ready to publish and distribute your video, think through three things:

• Where will the video live, and how will people find it?
• Who will have access to the video, and what happens when it's shared?
• How will you update content without creating confusion (i.e., version control)?

These questions matter more for scenario-based training videos than for a standard training video, because the interactivity changes what's possible depending on how you distribute it.

Let's say you created your video with Synthesia. You have a few options for how to publish and share from the platform, including downloading the video as a SCORM package or as an mp4, or creating a shareable link directly from the platform. However, an mp4 loses all interactivity. So in this case, you'll want to avoid distributing it that way.

Another important consideration is measurement. If you want to track how people move through a scenario, you'll likely want to host it in a compatible LMS or LXP (Synthesia also offers built-in analytics).

No matter what you decide, I recommend piloting the video first. Send it to a SME familiar with the topic or a small group from your target audience. Have them run through the scenario to make sure everything is working as intended and that the data you're looking to capture is being recorded.

How to measure scenario-based training videos 

In the example I gave earlier about managers coaching high performers who want a promotion when there isn't a business case for one, I gave two measurement examples: engagement data and retention metrics.

Engagement data is notoriously fickle. Teams running engagement surveys frequently redesign them — changing the questions, the timing, or both — and results rarely account for context like a new manager or a recent re-org.

Therein lies the challenge of evaluating training and explaining whether it's actually changing behavior. Using the same example: let's say a group of new managers all went through the same scenario-based training. Over a quarter, you track engagement and retention, and notice that compared to the previous quarter, engagement has stayed the same but retention has gone down. Does that mean the training failed?

In this case, you should try to parse out the data. Maybe you'll find that a specific individual was recruited away for the title they'd been asking for. Can you say the training was effective? Probably not, but you also can't say it wasn't.

That's why the data you use to evaluate whether the scenario is working should be repeatable and consistent over time. You can combine two types of measurement:

1. Snapshot measurement — video completion rates, scenario outcomes, choice distribution. These tell you whether people are using the training and how they're moving through it.
2. Impact measurement — the numbers your stakeholders care about, like retention, QA scores, escalation rates, or time to proficiency. These tell you whether the training is changing behavior on the job.

The goal is good enough measurement that tells you whether the scenario-based training is having an impact, and if not, whether you should edit the video to improve the outcomes or reconsider the scenario altogether.

If you're ready to create a scenario-based training video, head to our free AI video generator tool to get started. See how the workflow comes together in minutes.