Branching
scenarios.
A branching scenario is a decision-driven story. The learner makes a choice, sees the consequence, and makes another. Done well, it teaches better than any knowledge check. Done badly, it's an expensive multiple choice question.
Teaching judgement under pressure.
Branching works when there's a real decision to teach — one with consequences, no perfect answer, and where seeing the outcome is the point. Compliance scenarios, customer service, leadership conversations.
- Teaching judgement, not facts.
- Situations where the consequence matters as much as the choice.
- Scenarios with no single "right" answer.
- Practice for high-stakes conversations or decisions.
Anything that's really a quiz.
If your "scenario" has one right answer and two wrong ones, it's a knowledge check dressed up. Use the real thing — it's cheaper to build and learners won't feel patronised.
- Single-correct-answer compliance questions (use radio buttons).
- Information delivery — branching doesn't teach facts efficiently.
- Short modules where a five-decision tree is too much overhead.
- Topics where you can't show meaningful consequences.
Play through it.
A short customer service scenario. Three decisions, each with a brief consequence before the next. No "correct" answer — different paths land in different places. Reset and try again to see the alternatives.
A customer is on the line. They're frustrated — the order they were expecting today hasn't arrived, and they need it for a meeting tomorrow. They've already spoken to two other people. What do you say first?
Scenario complete
Three decisions × three choices = up to 27 possible paths. In a real course, you'd typically converge after each branch (so it's nine, not 27, scenes to write). Branching that doesn't converge gets expensive fast.
Get the structure right.
Converge between branches
A "full" branching scenario with 3 choices at every step explodes combinatorially — 3 decisions gives 27 endings. Most well-designed scenarios converge after each decision so different paths reach the same next scene, with a "your last choice means..." callback. Far cheaper to build, just as effective.
3 choices per decision is the sweet spot
Two choices feels binary, like a forced "good vs bad" framing. Four or more makes each scene heavy and increases authoring cost. Three feels like a genuine choice with trade-offs.
Visible progress
Learners need to know how long they're in for. A subtle progress indicator (3 dots, "Scene 2 of 4", or similar) reduces drop-off and lets them commit to the experience. Hidden progress feels like a maze.
Allow restart
A branching scenario that doesn't let the learner try a different path is just a particularly slow multiple choice. Surface a "try again" or "explore other options" CTA at the end — and make the alternative paths feel worth exploring.
Accessibility — semantic structure
Each scene should be a self-contained block of content, with the question and choices clearly associated. Choices should be real <button> elements, focusable and keyboard-activatable. The progression should be announced to screen readers — visible progress isn't enough.
Accessibility — pacing
Avoid time pressure that forces learners to respond in seconds. The point of a branching scenario is reflection. Time-pressured scenarios exclude learners who need longer to read or process text, and they reward speed over judgement.
How to design a scenario that teaches.
Write the choices first, the scenes second.
The decisions are where the learning happens. If your three choices feel like "correct / wrong / silly", you don't have a scenario — you have a quiz with characters. Good choices are all plausible. They reflect real trade-offs. They each have something going for them.
Show consequences, not corrections.
The point of branching is that the learner sees what happens. "Correct! ✓" or "Wrong! ✗" wastes the format. Instead: "The customer goes quiet. They expected acknowledgement first." Let the consequence teach the lesson without grading it.
Make every path land somewhere meaningful.
If two choices end with the same generic "course over" screen, the learner learns that their choices don't matter. Different paths should lead to different end states — different lessons, different outcomes, different next steps. The variety is the value.
Keep scenes short.
Each scene should be readable in 15 seconds. A paragraph of context, the dilemma, three choices. Anything longer turns the scenario into homework. The momentum of "what happens next" is what keeps learners engaged — long scenes break that.
Give context, then complication.
Open each scene with a sentence of context (who, where, what's happening) then the complication (what makes this hard). Don't make the learner read three paragraphs to find the actual decision point. Front-load the situation.
Use real language, not training language.
"The stakeholder is exhibiting non-constructive communication patterns" is training language. "She's snapping at the team" is real language. Branching scenarios live or die on whether the dialogue feels real. Read it aloud — if it sounds like a manual, rewrite.
End with reflection, not a score.
"You took the cautious path — here's what that gets you, and here's what the other paths might have taught." Reflection-based endings respect the learner's choices. Scored endings ("you got 2 out of 3 right") flatten the whole thing back to a quiz.
When you'd reach for branching.
Difficult customer conversations
Service training where the goal is judgement, tone, and recovery. Branching shows the cost of different openings without graded "correct" feedback.
Leadership and feedback conversations
Performance discussions, behaviour conversations, holding the bar. Branching lets new managers practice the conversation before having it for real.
Compliance edge cases
Not "is this a hazard yes/no" — but "you've spotted something ambiguous, what do you do next, and what happens when you do." Where judgement matters, branching teaches it.
Sales discovery and objection handling
How you respond to "we're just exploring options" or "your price is high" — there's no perfect answer. Branching lets reps see the consequences of each approach.
We build branching
that teaches.
Branching scenarios are some of our favourite work. Three-week delivery, transparent pricing, zero drama.