Give everyone $100 of pretend money to spend across the ideas, and see where it really goes.
Ask a group to rate ideas from 1 to 5 and almost everything ends up a 3 or a 4. Give each person $100 of pretend money to spend on the ideas they believe in, and the picture changes. The budget is fixed, so every dollar placed on one idea is a dollar not placed on another. People suddenly have to choose.
Each person divides their $100 however they like: all of it on one idea, or spread across several. Add up the totals and the group's real priorities appear, including how strongly people feel, which a show of hands can't capture. Then talk about the surprises before you decide.
Use it when you have a manageable shortlist and need a quick read on priorities, when ratings keep coming out the same, or when you want quiet people's preferences to count as much as the loudest voices.
A community garden's eight members spent $100 each. The top two ideas drew more than half the money.
Which options are on the table, and does everyone understand each one?
What does a dollar stand for, and are there any limits?
If this were your own money, where would you put it?
Which ideas attracted the most money, and how spread out is it?
What does the spread tell you that a simple vote wouldn't?
What will you take forward, and why?
The Hundred Dollar Test is quick and a little playful, but it fixes several problems with ordinary rating and voting.
When you rate ideas one at a time, you can like them all. A fixed budget makes every choice a trade-off: funding one idea means not funding another. That's how real decisions work, so the result is closer to what people would actually choose.
A show of hands records only yes or no. Dollars show intensity: one person putting $70 on an idea tells you something a single vote never could. You see both how many people care and how much.
Everyone gets exactly the same budget, and spending in private stops people simply following the boss. Quiet members' priorities show up on the tally as clearly as anyone else's.
Everyone already knows how to spend a budget. Framing the decision as an investment makes people ask would I really back this? instead of do I like it?, and that question tends to produce more careful, honest answers.
Write each option as one clear line and make sure everyone understands it. Merge duplicates. The test works best with four to twelve options; thin a longer list first with Negative Selection.
Everyone gets $100. Say what a dollar means, such as “where you'd invest if it were your own money”, and agree on any limits, such as whole $10 amounts or a cap per idea. Settle this before anyone spends.
Each person divides their $100 on paper, sticky notes or a quick online form. Silent, simultaneous spending stops people copying the most senior person in the room.
Add up the total for each option and show the results as a simple bar chart. Note how many people backed each idea, as well as how much money it drew.
Ask what surprised people, and listen to anyone who bet big on an outsider. If no idea stands out, run a second round with tighter rules, such as backing only two ideas each. Then decide.
Option cards on a wall, play money or sticky notes for each person (or an online form), a calculator and 20 minutes. With AI, use the prompt below.
Group: a community garden run by eight volunteer members. Goal: choose the projects for next season. Rules: $100 each, in $10 steps, any split allowed, spent in private.
A tool library, a kids' growing club, rain barrels, a pollinator strip, a compost workshop and a pizza oven, each written on its own card.
Everyone got $100 in $10 steps. A dollar meant: “I'd give up my own Saturdays to make this happen.”
Members filled in their slips at the same time. One put all $100 on the tool library; most spread theirs across three ideas.
Tool library $230, kids' club $180, rain barrels $150, pollinator strip $110, compost workshop $80, pizza oven $50.
The pizza oven had been the loudest idea in meetings, yet only two members funded it. The kids' club was a quiet newcomer's suggestion.
The tool library and kids' club go ahead. Rain barrels wait for a grant, and the pizza oven becomes a one-off fundraiser idea.
AI can't spend your group's money for you, and it shouldn't: the point of the test is to reveal what your people value. Where it helps is before and after. It can turn a messy list into clear options, model how different people might spend, and question the result so you notice what the tally hides.
Rewrite every option at the same size and detail, so none wins just by sounding better.
Rewrite these 8 options as one neutral line each, at the same level of detail, without selling any of them.
Ask how people who aren't in the room might spend, then check with them.
How might a parent, a retiree and a new member each split $100 across our list? Give a reason for each.
Get help seeing the shape of the result: broad agreement, split camps or a passionate few.
Here are our totals and how many people backed each idea. What patterns do you see, and what should we ask?
Design a sharper second round when the money is spread too thinly.
Our top four are within $30 of each other. Design a second round with rules that force clearer choices.
List your options, pick a mode, then copy the prompt into Claude, ChatGPT, Gemini or Copilot. In coach mode the AI helps you prepare the list and the rules, waits while your group spends, then helps you read the results.
The “Open in” links put the prompt in the web address. If your challenge includes anything sensitive, copy and paste instead.
After the session, click one to copy it, then paste it into the same chat.