Reverse brainstorming: find ideas by asking how to make it worse
Run this method in IdeaHone
Creates a space with the entities and scoring criteria of this method, ready for your next session.
Ask a room "how do we make onboarding better?" and you get silence, two safe suggestions and someone checking their phone. Ask "how do we make sure every new customer gives up in the first week?" and the room laughs and fills a wall in five minutes. Reverse brainstorming is that second question, plus a disciplined way to turn the answers back into ideas you can build.
Why it works
People find faults faster than fixes, because everyone in the room has already watched this product or process break. The negative question removes the fear of being wrong, since nobody can propose a bad way to fail. And every bad idea flips into at least one real one, because a precise description of how something breaks is half a description of how to prevent it.
Use it when the group is stuck, cynical or too polite: a normal round produced the same four ideas, the sceptics dominate, or hierarchy keeps the juniors quiet. Skip it when the group is already in flow.
How to run it
- Write the real question. One sentence on the wall, as a how might we question: "How might we get new customers to a first result in week one?" It stays visible all session.
- Write the reversed question underneath. Three forms work: how do we cause the problem, how do we make it worse, how do we guarantee failure. Pick the one that reads most naturally: "How do we make sure every new customer gives up in the first week?"
- Five minutes of silent writing. One failure per sticky note. Say out loud: "Nothing is too obvious and nothing is too silly." Silent writing is what gets the quiet people onto the wall.
- Read out and enjoy it. Everyone reads their notes aloud and sticks them up. The laughter is part of the method; it tells you the group has stopped guarding itself. Do not comment, do not filter.
- Cluster the ways to fail. Five to eight clusters, each named after the failure, not a fix: "nobody knows what to do first", "we ask before we give".
- Flip each cluster. Two questions per cluster, in this order: "So what stops this?" and "What is the opposite?" Write the answers on notes of a different colour under the real question.
- Keep the concrete, drop the mirror images. "Confusing first screen" flipped to "clear first screen" is a negation, not an idea. Keep only flipped notes that name a change someone could start on Monday.
- Score the survivors. Ten to twenty ideas usually remain. Run a quick impact and effort matrix or a dot vote and leave with three.
A worked example
Reversed question: "How do we make sure every new customer gives up in the first week?"
| Reversed idea | What it reveals | Flipped idea |
|---|---|---|
| Send the welcome email three days after sign-up | The first hour is when interest peaks, and we waste it | A first task the customer finishes within ten minutes of signing up |
| Ask for twelve fields before they see anything | Every form field is an exit | Show the product first, ask for data when the product needs it |
| Hide help behind a search box nobody uses | Help only works where the problem happens | Short inline help on the three screens that generate most tickets |
| Give everyone the same tour, whatever they bought | We do not know what the customer wanted on day one | One question at sign-up, and the first week branches on the answer |
| Never tell them what finished looks like | People quit when they cannot tell if they are doing it right | A visible "you are set up" moment with a defined first win |
The middle column is the one that matters. If you cannot fill it, the reversed idea was a joke with no fault behind it, and the flip will be weak.
One role, one risk
Ask the sceptic in the room to lead the reversed round. The person who normally kills ideas at step one is the best in the building at inventing failures, and it is the one round where that role shines. Give them the pen.
With a sensitive topic, for example a team that is already failing, the reversed round can turn into blame within two notes. Frame the question as the product or the process ("how do we make the release process fail"), never as people, and say so before the writing starts.
Four ways the flip yields no real ideas
- Stopping after the fun part. The reversed round is the easiest fifteen minutes of any workshop, and some groups end there, pleased and with no ideas. Book the flip before you start and protect its time.
- Flipped ideas that are only negations. "Do not confuse the customer" is the mirror of a failure, not an idea. No verb and no object: back to the cluster.
- Too much time on the reversed round. Five minutes writing, ten minutes reading. Longer and you get the same failures reworded, and the energy for the flip is gone.
- The facilitator filters the silly ones. "Make the login page play a trumpet" looks like a joke until it flips into "the login page is the loudest thing in the product and the quietest in our analytics". The silly ideas flip best, because they exaggerate a real fault.
Capture every idea while people talk
IdeaHone listens in your session and turns what people say into idea cards, questions and decisions, live on a shared board.
What is reverse brainstorming?
An ideation method where the group first answers the opposite of the real question (how do we cause the problem, make it worse or guarantee failure) and then flips each failure into a preventive or opposite idea. It works because people find faults faster than fixes.
When does reverse brainstorming work best?
With groups that are stuck, cynical or too polite: when a normal brainstorm produced the same safe ideas, when sceptics dominate, or when hierarchy keeps part of the room quiet. It is a poor choice when the topic is so sensitive that failures read as blame.
How do you turn reverse brainstorming results into real ideas?
Cluster the failures, then ask two questions per cluster: what stops this, and what is the opposite. Keep only flipped ideas that name a concrete change, drop pure negations, and score the rest with a quick impact and effort matrix or a dot vote.
