Brainstorm every way to cause the problem, then flip each answer into a fix.
Reverse Brainstorming turns the usual question upside down. Instead of asking “how do we fix this?”, you ask “how could we cause this problem, or make it worse?” Then you brainstorm as many answers as you can, the more outrageous the better.
It sounds like a joke, and it's meant to feel a little like one. People who freeze when asked for brilliant solutions are often superb at listing ways things go wrong. Once the list is up, you flip each cause back into a fix. The uncomfortable, valuable moment is noticing how many of the “bad ideas” are already happening.
Use it when a group is struggling to come up with ideas, when people lean toward critique more than invention, or when a problem keeps coming back and you suspect your own habits are part of it.
Flip the problem, list ways to cause it, find the ones already happening, then flip them back into fixes.
Choose a part of the diagram or a tab to explore each step.
How could you make this problem happen, or make it as bad as possible?
What are all the ways you could cause or worsen the problem?
Which of these “bad ideas” are already happening, even a little?
What is the opposite of each cause, and what would prevent it for good?
Which fixes would make the biggest difference, and which can you start this week?
Reverse Brainstorming works with a few natural habits of the mind instead of fighting them.
Most of us are quicker to spot what's wrong than to invent what's right. The reverse question puts that critical skill to creative use, so even skeptical or analytical people contribute freely.
Nobody worries about suggesting a “bad” idea when bad ideas are the goal. The playful framing relaxes the room, gets people laughing and draws out ideas they would normally keep to themselves.
When the list of ways to cause the problem matches what you already do, it's hard to ignore. Seeing your own habits framed as sabotage is a gentle, face-saving way to admit them.
Each cause comes with a built-in opposite. Turning “give them no one to ask” into “assign a buddy on day one” is a short, concrete step, so the session ends with practical fixes, not just a list of complaints.
Write your goal as a question, then reverse it. “How do we help new hires settle in?” becomes “How could we make new hires feel lost and unwelcome?” Make the reverse question vivid and specific.
List every way to cause or worsen the problem. Go for quantity and enjoy the absurd ones. Ask: what would a saboteur do? What small habits would quietly make it worse?
Read the list and mark every cause that already happens, even occasionally. Be honest. These circled items are often the most valuable output of the whole session.
For every cause, especially the circled ones, write its opposite and a way to prevent it. Push past simple negation: “don't ignore them” is weaker than “a named buddy who checks in daily”.
Group the fixes, then pick the few that would make the biggest difference with a quick NUF Test or Voting. Give each one an owner and a start date.
A whiteboard split into two columns (how to cause it / how to fix it), sticky notes, a timer and 45 minutes. With AI, start with the prompt below.
Subject: onboarding at a 40-person marketing agency. Reverse question: how could we make new hires feel lost, useless and unwelcome? Constraint: managers can give about two hours a week each.
Already real: laptops arrive on day three. Fix: IT sets up the desk, laptop and accounts the Friday before the start date.
New hires don't know who to bother with small questions. Fix: a named buddy from another team, with a 10-minute daily check-in for two weeks.
Already real: a link to a 200-page wiki. Fix: a one-page “first week map” with five things to read and five people to meet.
Already real: meetings full of client acronyms. Fix: a team glossary, and explicit permission to stop anyone mid-sentence and ask.
The first real feedback comes at the six-month review. Fix: short check-ins with their manager at days 7, 30 and 60.
34 ways to lose a new hire in 15 minutes, and an awkward 12 were already happening. Each one got a fix.
AI is surprisingly good at sabotage. Ask it how to make your problem worse and it will list dozens of causes, from the obvious to the sly, with no fear of sounding negative. What it can't see is which of those causes are already happening in your team. That honest step is yours, and it's where most of the value is.
Ask for a vivid list of ways to cause the problem, from obvious to subtle.
Write a saboteur's guide: 20 ways to make new hires feel lost and unwelcome in their first month.
Have AI find causes that look like normal, well-meant practice.
Which of these sabotage moves might a busy, well-meaning team make without noticing? Explain each.
Push past simple opposites to fixes that make the mistake hard to repeat.
For each cause I circled, give a fix that prevents it by design, not just by trying harder.
Turn your top causes into signals that the problem is creeping back.
For my top 5 causes, suggest an early warning sign I could check each month.
Describe the problem, pick a mode, then copy the prompt into Claude, ChatGPT, Gemini or Copilot. In coach mode the AI helps you flip the question, collects your ways to make things worse, then helps you turn each one into a fix.
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.