Surface the beliefs hiding inside a problem, then ask what becomes possible if they're false.
Every problem comes wrapped in assumptions: things we believe so firmly that we never think to check them. Meetings happen in person. Customers want the cheapest option. We can't move the deadline. Most are reasonable. A few are quietly blocking the best ideas.
Assumption Busting makes those beliefs visible. You list everything you're taking for granted about the problem, then challenge each one: what if this weren't true? Each challenged assumption opens a new direction for ideas, and some turn out to be false the moment anyone checks.
Use it when a team is stuck, every idea looks like the last one, or someone says “that's just how it is”. It's also a fast way to restart a session that has run out of energy.
Name the assumption on the left, then flip or drop it on the right. Every flip opens a new line of ideas.
Which rules, requirements or standards are you treating as fixed?
What do you believe you can't do, afford or get permission for?
What are you assuming about what people want, think or will do?
Which methods do you use simply because they're the usual way?
What do you believe is causing the problem, and how do you know?
What are you assuming the end result has to be?
Assumptions save us time every day. In creative work they also hide the options we've stopped seeing.
Assumptions are invisible because they feel like facts. Writing them down, especially the obvious ones, turns them into statements you can question, test or drop.
If a team silently agrees that “it has to be cheap” or “it must happen in person”, every idea lands inside that box. Removing a single false limit can open up a whole new family of solutions.
Asking “how do we know this is true?” is far less threatening than telling someone they're wrong. It turns disagreement into shared curiosity, and makes it easier to question the ideas nobody dares to touch.
Situations change faster than beliefs. A rule written for a reason that no longer applies, or a picture of customers that's ten years old, can often be checked in an afternoon, and the payoff is immediate.
Write the challenge in one sentence, the way you'd normally describe it. The wording often holds the first assumptions. “How do we get residents to come to our meetings?” already assumes that meetings are the answer.
Write down everything you're taking for granted: rules, limits, beliefs about people, methods, causes and the goal. Listen for the signal words: must, can't, they, always, only. Aim for at least 15.
Mark the assumptions that, if false, would change the most. These are usually the ones that feel too obvious to question.
For each starred assumption ask: what if the opposite were true? What if it didn't apply? Write down the ideas each challenge opens up. When new assumptions appear, bust those too.
Some assumptions can be tested fast: ask the people involved, read the actual rule, run a tiny trial. Turn the best busted assumptions into experiments with a clear signal of success.
A wall or sheet split into two columns (we assume / what if not?), sticky notes, a timer and 20–30 minutes. With AI, start with the prompt below.
Subject: public input on a small city's annual budget. Goal: hear from a wider mix of residents, including renters and young families. Constraint: state law requires one formal public hearing.
Assumed: input means the formal hearing. Busted: the law requires one hearing, not one channel. Add pop-up tables at the farmers market.
Assumed: residents don't care about the budget. Busted: they care about parks, potholes and libraries. Ask about those, not “the budget”.
Assumed: input means three minutes at a microphone. Busted: let people split a pretend $100 across city services, on paper or by text.
Assumed: we can't afford more outreach. Busted: libraries, schools and the youth council already reach residents. Borrow their channels.
Assumed: success means a fuller hearing room. Busted: success means more voices in the decision, wherever they're heard.
18 assumptions listed and 7 busted. The legal rule turned out to be far looser than everyone believed.
A pretend $100 budget game at three pop-up tables (farmers market, library, school pickup) plus a text-in option, with results presented at the formal hearing. They'll run it for one budget cycle and compare who took part with previous years.
Your own assumptions are hard to see, because to you they look like facts. AI doesn't share your history, your office culture or your habits, so it's good at naming what you take for granted. What it can't know is which assumptions are actually true in your world. That's your job, and often a quick check with real people settles it.
Paste how you usually describe the problem and let AI find what's hidden in the wording.
Here's how we describe our problem. List 15 assumptions hidden in it, grouped by type.
Turn every starred assumption into its opposite and see what ideas it unlocks.
For my top 5 assumptions, state the opposite and give two ideas each opposite makes possible.
Find organizations that already work without your assumption.
Which cities or organizations gather budget input without public hearings? What do they do instead?
Rank assumptions by how easy they are to test, then design the quickest checks.
Rank my assumptions by how cheaply I could test them, and suggest a one-day test for the top 3.
Describe the problem the way you normally would, pick a mode, then copy the prompt into Claude, ChatGPT, Gemini or Copilot. In coach mode the AI helps you dig up your assumptions, then challenges them one at a time with you.
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.