Storytelling in Learning Design: What the Evidence Actually Says

Insights from a recent Digital Learning Institute (DLI) webinar with Tom McDowall.

We tell ourselves stories are magic. Wrap any piece of training in a narrative and the thinking goes, retention skyrockets, behavior changes, and the content finally sticks. It's a comforting idea. It's also, according to the evidence, not quite true.

That's the starting point for a recent session on evidence-informed storytelling in learning design. Founder and principal consultant at Evolve, an exponent of evidence-informed practice and founder of the IDTX event series, Tom McDowall set out to separate the genuine power of storytelling from the myths that have grown up around it and to give instructional designers a practical, research-backed way to use stories well.

You've probably seen the claim: stories are 22 times more memorable than facts. It's the kind of statistic that shows up in LinkedIn posts and business books, promising a silver bullet for learning and development.

There's just one problem: no peer-reviewed, replicated study supports it. It's a lovely claim, but it's not true, in the same family as "a picture is worth a thousand words."

That doesn't mean stories don't work. It means the truth is more interesting, and more useful, than the marketing version. Research suggests a story may be understood and remembered better than the same content delivered as plain explanation but this is highly situational. It depends on the content, the audience, and how the story is told. Stories are powerful, but only when used in the right way, for the right audience, at the right time. In that sense, they're no different from any other instructional method available to a designer.

Strip a story down to its essentials and you get a small set of components: who it's happening to, when it's happening, where it's happening, what they're trying to achieve, and what caused them to do this. Spin those in a loop and you have a story.

These components matter because they're also what we remember. When people recall a story, they're not replaying it word for word they're reconstructing something called a situation model: the key beats, the cause and effect, the things that transcend the specific language used to tell it. Ask someone to write down the story of Cinderella and they'll skip the exact wording, skip whether the stepsisters have names, and land on the core: a downtrodden character, an intervention, a time-limited transformation, a happy resolution.

Of those components, two carry unequal weight. The who and the when matter. The goal and the cause, the causal connection between what happened and why matter enormously, arguably more than anything else. The where matters far less; it's why we're comfortable relocating a Shakespeare play to a modern city without losing the story. (Corporate training is sometimes an exception there are cases where the "where" carries real weight but as a rule, causal connection is doing most of the heavy lifting.)

This is worth testing on yourself. Two sentences: The report went out without the checks being run. On Monday morning, the client rang. Nobody assumes the client called to extend their contract. Our brains are bridging inference machines we automatically connect available facts into cause and effect, even across two sentences we've never seen linked before.

That automatic bridging is a double-edged sword. It means people will readily fill in gaps you leave them which is useful. But changing behavior in someone's actual job requires more than a bridging inference, it requires a predictive inference. Learners need to get from "that's what happened in the story" to "and therefore, when I go back to my job, I need to do X." That leap doesn't happen automatically, and it's a major source of learning transfer failure.

The fix is deceptively simple: say the forward consequence out loud. Ask directly, what have you learned from this story? How will you apply it? What does tomorrow look like? Don't trust learners to figure that out on their own; making that predictive leap explicit is what turns an enjoyable story into one that actually changes behavior on the job.

Retrieval isn't playback. Just as storytellers around a campfire changed small details with every retelling, a goat becomes a cat, a he becomes a she the details in your training stories that aren't load-bearing will fade, get forgotten, or get changed by the people trying to recall them, simply because those details are perceived as unimportant.

Take the moment in some versions of Cinderella where a stepsister cuts off her toes to fit the shoe. Great detail, not essential some tellings include it, some don't, and the story survives either way. Now remove the fairy godmother. The story collapses. One detail is decoration, the other is structural.

This has a direct design implication: anything you want remembered must be tied to cause or consequence. Everything else is at risk of being dropped, altered, or replaced by whatever the learner's brain decides is the "real" causal chain which may not be the one you intended.

It also means resisting what's sometimes called seductive detail, the beautifully written tangent that adds atmosphere but nothing else. Tolkien can spend five pages on the grass of the Shire and it's glorious. A GDPR e-learning module describing the grass is not. Only include details that are relevant and explicitly useful; anything else is something a learner will have to actively discard, and every discard is a small tax on the story's effectiveness.

We tend to think emotion "elevates" a story evenly, making the whole thing more memorable. It doesn't. Emotion works like a spotlight: it focuses attention intensely on one moment, at the expense of everything else around it.

Think of the Disney Cinderella's slow, orchestral first half against the frantic, emotionally charged countdown to midnight. That shift in intensity draws attention precisely to that moment which is exactly why placement matters so much. If the emotional peak of your training story lands anywhere other than the teaching moment, you've built a genuinely great story and pointed all its power at the wrong thing. A phenomenal story can completely fail to help people learn, simply because the spotlight was aimed at the wrong beat.

Handing someone 100 pages of information and telling them the answer to question one is on page 46 is not helpful. Handing someone a filing cabinet with a sensible structure, then adding information piece by piece, is. Stories work the same way: where possible, set up the context before delivering detail, rather than dumping information and hoping a story at the end will retroactively organize it.

This connects to schema, the organized structures of prior knowledge everyone brings into a room, even on their first day in a new job. A good story signals which existing schema new information belongs to, and whether that information agrees with what's already there or challenges it. Information that fits an existing schema slots in easily. Information that contradicts it is harder to absorb which is also why "unlearning" isn't really a cognitive process. You can't delete a schema, you can only reinforce a new pattern until it becomes stronger than the old one, through repeated practice across different contexts.

Not everything needs a story, and using one indiscriminately can actually undermine the technique's credibility over time.

A useful distinction here, borrowed from social anthropology, is transparent versus opaque knowledge. Transparent knowledge is visible, you can watch people enact it just by looking around the workplace. Opaque knowledge is the stuff nobody would know unless someone specifically told them. Stories are strong at conveying principles, ways of working, and opaque, nuanced, or contextual knowledge. They're a poor fit for "fill out this form and click submit."

Two other filters are worth applying:

• Recall versus application. Meta-analyses of storytelling in training consistently show its real benefit is in application, not recall. If the goal is a straightforward compliance tick-box exercise, a beautifully crafted story is often wasted effort. If the goal is genuine behavior change, storytelling earns its place.

• The expertise reversal effect. Training that works well for novices can be not just unhelpful but actively counterproductive for experienced staff. Where possible, tailor story-based content to the audience's experience level, or make it optional for people doing a refresher rather than learning something new.

The blunt test: if in doubt, delete the story and ask whether the training still works. If it does, the story wasn't adding value it was just taking up the learner's time. That's true of nearly any technique, not just storytelling: strip it out, and if nothing is lost, leave it out.

Once there's a genuine case for using a story, the construction process is straightforward, if disciplined:

1. A single objective. One thing the protagonist is trying to achieve, not six. Six objectives is a novel, not a training story.

2. A single causal chain. Map out what the protagonist needs to do, and why, step by step, to reach that objective. This is the spine of the story.

3. Only the cast that earns its place. Every character and detail should serve the causal chain. Irrelevant color, that two coworkers both happen to own cats, say weakens the story rather than enriching it.

4. A clear peak moment. Identify where the emotional or informational high point sits, and make sure your pacing and structure highlight it deliberately.

5. Strip it back. Once it's built, go through and try removing things. If removing something doesn't hurt the story, it didn't need to be there.

6. Support transfer. This is the step most often skipped, and the one most responsible for story-based training falling flat. What happens immediately after the story? Learners need to be asked to reflect on what they learned and set out concrete next steps, the predictive inference has to be prompted, not assumed.

Stories should also come from a trusted source. The evidence is clear on this point: whether the storyteller is a respected figure within the business or the L&D function itself, if learners don't trust where the story is coming from, they'll question its content rather than absorb it, undermining even a well-constructed narrative.

One reassuring note for anyone worried about their creative-writing skills: most workplace stories don't need to be invented from scratch. They already exist. Asking people what actually happened "walk me through it", usually surfaces a true story that's more specific, more credible, and more useful than anything invented for the purpose. Generalized, invented stories tend to lose the domain-specific detail that makes them land, real ones already have it built in.

Storytelling in L&D is often sold as a creative pursuit, and it is, but the strongest work in this space treats it as an engineering problem as much as a creative one. It draws on scientific research, professional experience, and organizational insight together, rather than leaning on any one of them alone.

Good instructional storytelling isn't about following a template like the hero's journey. It's about identifying a single objective, mapping the causal chain that gets a protagonist there, stripping away anything that doesn't serve that chain, placing emotional weight exactly where the teaching moment sits, and critically helping learners make the leap from story to action once it's over.

Used with that kind of discipline, storytelling doesn't need exaggerated claims to justify itself. The evidence for what it can do, used well, is compelling enough on its own.

Are stories really 22 times more memorable than facts? No. There's no peer-reviewed, replicated study behind that claim. Stories can be understood and remembered better than plain explanation, but the effect is highly situational, it depends on the content, the audience, and how the story is delivered.

What are the core components of a story? Who it's happening to, when it's happening, where it's happening, what they're trying to achieve, and what caused them to do it. The who, the goal, and the cause carry the most weight in what people actually remember; the where usually matters least.

What does "front-loading" a story mean in practice? It doesn't require putting the entire story up front. A story can be opened, characters and situation introduced, then paused to deliver more formal or detailed instruction, before returning to close out the narrative. As long as the context is established early, learners have a structure ("a filing cabinet") to organize the detail that follows.

Can you give an example of storytelling used well? One example given was early cybersecurity awareness training, told from a first-person, "you don't know anything yet" perspective: learners discover they've been hacked, then go back in time through the events of that day to spot what they missed. This let people re-explore a familiar-feeling scenario and make different decisions the second time around, without the exercise feeling punishing.

Where should someone start if they're new to story-based learning design? Stay tool-agnostic at first. Write the story out in a plain Word document before touching any authoring tool. Once it holds together as a story on paper, look for natural decision points, where could the protagonist make the wrong call, and what would that look like? Only move into branching or more complex tools once that core narrative spine is solid.

Do stories work equally well for everyone? No, this is the expertise reversal effect. Training that's highly effective for novices can be unhelpful or even counterproductive for experienced staff. Where possible, tailor story-based content to audience experience level, or make it skippable for people doing a refresher rather than learning something new.

Is storytelling more useful for recall or for application? Application. Meta-analyses show storytelling's clearest benefit is in helping people apply knowledge, not in helping them recall facts. For pure compliance tick-box training, a story is often not worth the investment; for genuine behavior change, it is.

Do I need to invent stories from scratch? Usually not. Most useful workplace stories already exist, ask people what actually happened and they'll tell you. Real, specific stories tend to be more effective than invented, generalized ones, because the domain-specific detail that makes a story land is already built in.