Topic 2.2

Good Scenarios Take Forever. That's the Problem.

AI accelerates scenario creation. You still make them real.

⏱️ 12 minutes 📋 Prompt Templates ✓ Quality Checklist

The Problem

Good scenarios have names, numbers, realistic tension, and specific decisions. They feel like situations your learners actually face.

Which takes forever to write.

You're not blocked on creativity. You're blocked on volume—generating enough scenarios fast enough to give learners variety and practice.

AI solves that. But if you just ask for "a customer service scenario," you get generic situations no learner recognizes.

The fix: give AI specifics. It generates realistic scenarios fast. You add organizational truth.

The Basic Prompt Template

📋 Copy this template
Generate 5 realistic scenarios for [role] dealing with [situation]. Context: - Setting: [work environment/industry] - Constraints: [time pressure, resources, policies] - Learner level: [junior/experienced/expert] - Decision required: [specific choice or action] Requirements: - Include specific names, numbers, timeline - Force a real trade-off (no obvious answer) - Reflect actual work constraints - Keep setup under 150 words

Example filled in:

Generate 5 realistic scenarios for retail store managers dealing with inventory shortages. Context: - Setting: Mid-size retail chain, sporting goods - Constraints: Can't exceed weekly reorder budget, must maintain floor display standards - Learner level: New store managers (first 6 months) - Decision required: Which items to reorder vs. delay Requirements: - Include specific product names, stock levels, reorder costs - Force a real trade-off (no obvious answer) - Reflect actual work constraints (budget, customer demand patterns) - Keep setup under 150 words

What Makes Scenarios Work

🎯 Specificity over generic ▼

Generic: "A customer is upset about a delayed shipment."

Specific: "Maria Chen ordered custom cycling jerseys for a charity ride happening Saturday. It's Wednesday. The shipment is stuck in customs. She's calling to cancel unless you can confirm Friday delivery."

Names, timeline, stakes—those make it feel real.

⚖️ Force trade-offs ▼

Good scenarios don't have an obvious right answer. They have competing priorities.

Example: Approve overtime to finish a project on time (hits budget) vs. push deadline (damages client relationship).

If one option is clearly best, it's a quiz question—not a scenario.

🏢 Add organizational reality ▼

AI doesn't know your company culture. Add it:

  • Approval hierarchies ("Your director must sign off on refunds over $500")
  • Unwritten rules ("We never escalate to legal without trying the account manager first")
  • Peak periods ("It's Black Friday week—your regional manager is unavailable")
  • Resource constraints ("IT is on a hiring freeze, no new tool requests")

These details separate generic scenarios from ones your learners recognize.

Build a Scenario Bank

One good scenario + variations = reusable training asset.

The approach: Generate a strong base scenario, then create variations by changing one element at a time.

🔄 Variation prompt
Take this scenario: [paste scenario] Create 3 variations by changing: 1. The stakeholder's personality (from collaborative to hostile) 2. The time constraint (from urgent to long-term) 3. The resource availability (from limited to abundant) Keep the core situation the same.

Result: Four related scenarios testing different facets of the same decision-making skill. Less work. More practice opportunities.

Generate Realistic Data, Too

Some scenarios don't need dialogue. They need numbers.

Training spreadsheet skills? Financial analysis? Dashboard interpretation? The scenario is the dataset itself—and AI can generate realistic dummy data faster than you can make it up.

Example: Teaching Excel to employees at an electric utility.

Instead of fabricating outage data by hand, prompt AI:

Generate a realistic dataset of power outages for a mid-sized electric utility. Include: - 50 outage records from the past 6 months - Columns: Date, Time, Duration (minutes), Customers Affected, Cause, Substation, Crew Response Time - Realistic distribution: mostly weather and equipment failures, occasional vehicle accidents and animal contact - Seasonal patterns: more storm-related outages in summer - A few outliers (major events affecting 5,000+ customers) Format as CSV.

Now learners practice pivot tables, lookups, and analysis on data that looks like their actual work—without exposing real operational information.

Works for: Sales data, patient records (de-identified patterns), inventory logs, call center metrics, project timelines, budget spreadsheets—any structured data your learners will encounter on the job.

💡 The Advantage

Real data has privacy and compliance issues. Made-up data looks fake. AI-generated data threads the needle: realistic patterns, zero risk.

Four Traps to Avoid

1. Too much setup ▼

Problem: 500 words before the decision point

Fix: Get to the choice fast. Learners lose focus after 150 words of setup.

Test: Can you identify the key decision in the first paragraph? If not, trim.

2. Obvious answer ▼

Problem: One "right" choice = quiz question, not scenario

Fix: Add real trade-offs. Both options should have legitimate pros and cons.

Test: Would a reasonable expert pick differently than you? If not, add complexity.

3. Unrealistic stakes ▼

Problem: "Building on fire, 30 seconds to decide"

Fix: Dial back unless training firefighters. Match stakes to actual job pressure.

Test: Ask your SME: "Does this really happen?" Adjust based on their answer.

4. Missing culture ▼

Problem: AI doesn't know your org's unwritten rules

Fix: Add hierarchy norms, communication style, approval processes

Example: "In our culture, you email the client directly. At this company, you loop in your manager first."

The Refinement Loop

Don't restart. Direct revisions:

💬 Refinement prompts
Need more detail? "Add specific numbers, names, and a timeline to this scenario." Too complex? "Simplify for entry-level learners—reduce the number of complicating factors." Weak tension? "Make the trade-offs harder. Right now the best choice is obvious." Wrong context? "Rewrite to reflect [specific organizational norm or constraint]."

Key Takeaways

  1. AI drafts, you contextualize. Generate scenarios fast. Add organizational reality.
  2. Specificity beats generic. Names, numbers, timeline, setting—details make it real.
  3. Force decisions. Good scenarios have no obvious answer—just trade-offs.
  4. Build a bank. One good scenario + variations = reusable assets.

Try It Now

🎯 Your task:

Pick a topic from your current project. Generate 3 scenarios with the basic prompt. Select the best one. Add specific organizational details.

The test: Would your learners recognize this situation from their actual work?

📥 Download: Scenario templates and quality checklist (PDF)

Ready-to-use templates with prompt variations and evaluation criteria.

Download PDF