How to Convert Training Materials to Video with AI

If you're in L&D, you probably have a content library that's underutilized. And I'm not just talking about eLearning courses that no one finishes. I'm talking about content sitting in PowerPoint decks, PDFs, Word docs, and internal wikis.

When teams come to us, it's often because they're sitting on a gold mine of institutional knowledge that's stuck in formats employees aren't referencing. So they want to turn their training materials into engaging videos, and quickly.

But here's the reality. Converting a text-heavy, boring PowerPoint into a video will give you a bad training video. That's because engagement is correlated with effective learning design. And even AI can't turn poorly designed content into an effective training video.

Here's how to do it well.

Step 1: Confirm the content is worth converting

You've probably heard "garbage in, garbage out" applied to AI. If you put in poorly designed content, you're going to get a well-produced version of the same problem.

That's why, before you convert anything into a video, you need to do a quick quality check. Ask one question of every piece of content: does this have a clear learning objective and measurable outcome?

Every video you create should be grounded in a learning objective and performance outcome. You can use a template like this: 

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

If your content doesn't pass that bar, ask yourself: is it worth redesigning before converting? If yes, proceed. If no, that's okay. Not everything needs to be a video.

Step 2. Generate an outline

Once you've confirmed your content is worth converting, it's time to generate an outline. Go to our AI video generator and upload a file, like a one-page document outlining the key reminders from an SOP.

If you're interested in following along, here's an example of a PDF training that I've converted into a video below.

After you upload your file and click generate, you'll land in our AI Assistant, which builds an outline like the one below. You can edit the text or delete entire scenes before moving on.

Step 3: Create your first draft

When you click "Continue to editor," you'll land in Storyboard, Synthesia's script-focused view within the editor. This shows you what the AI has generated based on the input you've provided. In my example, there are no visuals, only text, so the AI will recommend an avatar as well as images, motion graphics, b-rolls, and more for my scenes.

I can give the Assistant feedback by chatting in the sidebar, or switch to Canvas to directly modify my scenes.

Step 4: Choose your avatar

With a draft ready, it's up to you what to focus on next. Some people like to start with their avatar, since they guide the viewer through the video. You can customize their appearance (branded puffer vests, anyone?), pose, and even location to align with your workplace. I've selected Inna for my training video, and am using her voice, which is described as "Encouraging".

Pay close attention to their tone, accent, and pronunciation. If your company has a name like ours, you'll want to standardize how that gets said across your content.

You can also go avatar-free and use narration only. Having an avatar onscreen adds to cognitive load, so for dense procedural content it's worth considering.

Step 5. Customize your scenes

Next, you can direct your attention to customizing the scenes. Think one idea per scene. If you want to add any visualizations, like motion graphics or B-roll, or interactivity, remember to keep attention focused.

Preview your video before generating, and double-check that everything from the pronunciation to the flow makes sense. 

Some generated assets may appear as placeholders or low-fidelity previews until you generate the final video. Here's an example of how I converted the PDF to video.

Step 6. Publish and share

When you've generated your video, you can localize it into over 160 languages with a few clicks. If you’re localizing videos, be sure to include native speakers when you conduct any pilots. 

Once you're satisfied with your video, you'll want to publish it and work through a few decisions.

1. Do you need to localize the video? (If so, follow these localization instructions.)
2. Where will the video live (e.g., an LMS), and how will employees find it?
3. Who will have access, and what happens when it's shared?
4. How will you update content without creating confusion or outdated versions?
   

Each of these decisions will impact your measurement goals (more on that next). 

If you want to see how a company transformed their content library, read about how BD Rowa, a pharmacy automation and robotics company, cut content production time by 75% by transitioning from slide decks and emailed updates to training videos.

Step 7: Measure the impact

Before you try to convert an entire content library into video, I highly recommend evaluating whether the transition to video is worth it for your organization. Often, I find users are so excited by the possibilities of quickly creating training videos that they forget to measure if the transition made any impact.

So here's how to do that. First, go back to your learning objective and performance outcome statement (the bolded one I told you to fill in earlier). Then, identify a few signals that you're meeting expectations.

Remember, you don't want to replace one content library that no one uses with another one. The entire point of converting content into video is to explore whether another medium is more impactful for your employees.

If you want to talk to our team about transitioning a content library, or any other content challenges you're facing, book some time.