How to Create Your Own Business Simulation: From Paper to Web App
You have a case or workshop exercise that gets people arguing over a decision. How do you turn it into something they can play in a browser, where today's choice changes the next round?
Conceptual illustration of a paper decision map beside a web prototype.
To create a business simulation, choose a decision your learners need to practise. Write down what each option changes, then run the exercise on paper. Once the rules hold up, build a web prototype. Test it with learners before you set up hosting, saved progress, and the controls a facilitator will need.
ChatGPT or Claude can help write the scenario and code the first version. Vercel can put a web app online. If you need to run it across cohorts, AdelieSim can help design and build the finished simulation.
1. Choose the decision learners will face
Give participants a choice they cannot make without giving something up. Show the consequence, then let that result shape their next choice. What judgement do you want them to practise?
Example: a leadership programme
A plant manager has limited budget and two urgent problems: a production delay and a safety warning. Learners choose what to fund first. Cash, output, and safety risk change. They make the next decision with those results in front of them.
An MBA instructor could use this structure for pricing or negotiation. It also works for stakeholder management. In an L&D programme, managers might have to weigh speed against quality with limited team capacity. Pick one audience and one learning objective for the first version. Two or three decisions are enough to test the idea.
Write down the brief
- Who will play, and what do they already know?
- What should they be able to judge or explain afterward?
- What role are they in? What pressure and incomplete information do they face?
- What can they choose, and what must they give up?
- What will you show in the debrief to discuss their reasoning?
2. Write the rules and consequences
Put the outcomes in a table before writing code. For each option, say what changes immediately and what happens in the next round. A developer or AI coding tool needs those rules in writing.
| Decision | Immediate effect | Next-round consequence |
|---|---|---|
| Fund the repair | Cash falls; safety risk falls | Production recovers more slowly |
| Prioritise output | Cash holds; output rises | Safety risk and employee concern rise |
These rules are invented for the example; they are no guide to running a real plant. Each option should have a cost that learners can defend in the debrief. Write down how the scores change, when a new event appears, and when the exercise ends so the result stays consistent from one playthrough to the next.
3. Test the loop on paper
Give the brief and decision sheet to three or four people who resemble your learners. Play the facilitator yourself. Watch where they ask for information you have not given them or assume a rule you never wrote. Ask whether the options felt distinct and whether the result made them reconsider their first choice.
Change the rules now, while they are still on paper. If the round produces little to discuss, change the decision or its consequences. In a class, leave time for learners to compare choices and connect them to the concept you teach.
4. Use ChatGPT or Claude to prototype
ChatGPT can help write a simple app and show you a preview. Claude can make small interactive apps called artifacts. Either can help you try a first version of the decision screen. Let learners play it before deciding the design is ready.
Give the tool your brief and decision table. Ask for one complete round first. A prompt for the plant-manager example could read:
Build a simple web demo for a 30-minute management simulation. The learner is a plant manager choosing between funding a safety repair and protecting this quarter's output. Use the decision rules below exactly. Show a briefing, a choice, its result, the next situation, and a summary for the facilitator. For this first demo, keep the learner's choices in the browser; do not add accounts yet. Make every button usable without a mouse. Do not invent extra scores or outcomes. Ask me about any missing rules before you build it.
Replace the plant-manager scenario with yours and paste the rules underneath. Play every path in the result. Ask the tool to explain how each choice changes the scores and what happens next. Have it check the important rules too. Keep confidential client, learner, or company information out of the prompt unless your organisation has approved its use.
See OpenAI's code preview guide and Claude's artifacts guide for the tools' current capabilities.
5. Turn the prototype into a usable web app
The prototype may work well until someone closes a laptop halfway through a round. Before you use it in a class or programme, decide what it must do when learners arrive late, lose their connection, or need help from the facilitator.
- How will learners join, sign in, and resume a session?
- Will different roles see different information or make decisions that affect one another?
- How will the facilitator start, pause, advance, or end a session?
- Which choices and outcomes will the facilitator need for the debrief or a downloadable report?
- Can learners use the app with a keyboard and read it comfortably? How will you handle their data?
For examples of browser-based simulations in different subjects, see AdelieSim's Stakeholder Management Simulation or Transfer Pricing Simulation.
6. Give learners a link they can open
A preview inside ChatGPT or Claude is useful while you test the idea. For a course or programme, learners need a link they can open in their browser. Hosting is the service that keeps your app online at that link. Vercel is one website that does this.
Make a first version
ChatGPT, Claude, or a developer turns your rules into an app you can try.
Put it online
A hosting website such as Vercel publishes the app at a web address.
Try the link
Learners open that address. Check what happens when they leave and return.
If Vercel asks for a “repository”
It means an online folder for your app's files, often kept on a website called GitHub. “Git” is the tool developers use to track changes to those files. A technical colleague can help with this setup, or AdelieSim can build and host the simulation with you.
If learners need accounts, saved decisions, or a screen where the facilitator can follow their progress, those features need to be set up and tested too. Check what happens when someone closes the browser and comes back later.
In a pilot, have learners join on their own devices and finish without hints. Ask them to close and reopen the app, then tell you where they got stuck. Give the facilitator only the data the app provides and see whether they can run the debrief. Fix confusing rules and delivery failures before adding more rounds.
Build it yourself or work with a specialist?
A solo exercise with a few decisions is a manageable place to try building with AI. When learners have different roles or their decisions affect one another, the build gets more demanding. So do tools for running a live session, downloadable results, connections to a learning platform, and repeated use across cohorts. You can direct the teaching and subject matter while a development partner builds the web app.
AdelieSim builds custom browser-based business simulations with professors and L&D teams. We can start with a case or workshop exercise and work through the simulation design, development, delivery, and debrief materials with you. A brief like the one above is enough for a first conversation.
Tell us the decision your learners need to make.
Include who will play and how long you have. We'll work out what a web version would need.
Frequently asked questions
Can ChatGPT or Claude build a business simulation?
Yes. Give either tool a scenario and explicit decision rules, and it can help you make a prototype or generate web app code. You need to play through every path, check the rules, and prepare the app for real participants.
Do I need to know how to code?
You can design and test the first round on paper without coding. AI tools can help make a web demo. For a live course or programme, you may need technical help to put it online, give learners accounts, save their progress, protect their data, and let a facilitator manage the session.
How do I put a simulation online?
The app needs a web link learners can open. A hosting website such as Vercel can publish it and provide that link. If learners need accounts or saved progress, someone must set those up as well. Test the live version with a few participants before using it with a full group.
What should I give a simulation developer?
A developer can start with your learning objective, audience, and session length. Add a short scenario, the participant roles, one or two difficult choices, what happens after each choice, and the questions you plan to ask in the debrief.
When should I work with a specialist?
Consider a specialist if learners have different roles or make decisions that affect one another. Running a live class, saving each learner's progress, showing results to a facilitator, or connecting the simulation to a learning platform also adds work.