Get a group calling out ideas fast, with judgment on hold, so one thought sparks the next.
Best for
Lots of ideas, fastany open problem a group cares about
Time
30–60 minutesplus 15 minutes to sort the ideas
People
Team4–10 people and a facilitator
Difficulty
Beginnereasy to start, harder to do well
Origin
Alex Osborn, 1940s–50sset out in his 1953 book Applied Imagination
01 What it is
Many minds, one question, and judgment on hold.
12345
Brainstorming is the classic group method for producing ideas. A small group faces one clear question, people call out ideas, and someone writes every one where all can see. At this stage the goal is quantity. You are filling the wall, not picking a winner.
It runs on four rules. Hold back criticism. Go for as many ideas as you can. Welcome the wild ones. Build on and combine what others say. The rules matter because the quickest way to kill an idea in a meeting is to judge it, out loud or silently, before it has been heard. Separate the making from the judging and a group becomes far more inventive.
Done carelessly, brainstorming lets the loudest voices take over while everyone else waits for a turn. A sharp question, a few silent minutes at the start and a firm facilitator fix most of that. If your group is shy or uneven, try Brainwriting instead.
Use it when you need a wide range of options quickly, the group brings varied knowledge, and people are comfortable thinking out loud in front of each other.
Where it comes from. Advertising executive Alex Osborn began running group idea sessions at his agency around 1939. He coined the word “brainstorming” and set out its rules in his books, most famously Applied Imagination (1953).
Steps 1–4 generate and step 5 judges. Keeping them apart is the whole trick.
Choose a part of the diagram or a tab to explore each step.
1
Frame the question
One clear question, written where all can see
What one “How might we…?” question do we want as many answers to as possible?
Is it narrow enough to answer, but open enough to surprise us?
Have we said who it's for and what “better” means?
Would everyone in the room read it the same way?
Café slow hours“How might we bring more customers in between 2 and 5 p.m. on weekdays?”
2
Set the four rules
No criticism, volume, wild ideas, build on others
Does everyone know the rules, and who will keep them?
Post the four rules where everyone can see them
Name a facilitator to guard the rules and a fast scribe
Warm up with two minutes on a silly practice question
Café slow hoursWarm-up: uses for an unsold muffin. Two minutes of laughter, then the real question.
3
Flood the wall
Call out ideas; the scribe captures every one
How many ideas can we get on the wall in the next 15 minutes?
Which idea on the wall could we stretch, flip or combine with another?
Say “Yes, and…” to someone else's idea
Merge two numbered ideas into a third
Push a tame idea further: bigger, stranger or cheaper
Café slow hoursBoard games plus the student discount: a weekday “study and play” pass for the college nearby.
5
Sort and shortlist
Cluster first, then choose with a separate method
Which themes stand out, and which ideas are worth testing?
Group ideas into themes before anyone votes
Use dots or a quick score to pick 3–5 to develop
Rescue one wild idea: what's useful inside it?
Café slow hoursThree themes emerge: events, deals and atmosphere. The team votes to try the study pass first.
02 Why it works
Why holding back judgment works
Brainstorming is easy to dismiss, but each of its rules targets a real habit that stops groups from thinking well.
1
Judgment shuts people down
In ordinary meetings people filter their ideas before saying them, for fear of looking foolish. Psychologists call this evaluation apprehension. A firm no-criticism rule lowers the risk of speaking, so more half-formed ideas reach the wall, where others can help them grow.
2
One idea triggers the next
Memory works by association. Hearing someone else's idea opens paths in your mind that you wouldn't have found alone. That's why the rule to build on others matters: the best idea often arrives third or fourth in a chain.
3
Quantity leads to quality
The first ideas in any session are usually the obvious ones everyone already knows. Pushing for volume forces the group past them. Studies of idea generation generally find that groups who produce more ideas also produce more good ones.
4
Know its weak spots
Only one person can talk at a time, and people forget ideas while they wait. Researchers have repeatedly found that people who think alone first, then pool their ideas, often produce more. So start with silent minutes, and keep voices balanced.
03 How to use it
Run a brainstorm in six steps
Write a sharp question 5 min
Phrase it as “How might we…?” and make it specific. “How might we cut car traffic at school pickup?” beats “Traffic ideas?”. Write it at the top of the board and check that everyone reads it the same way.
Gather the right mix before
Invite 4–10 people with different knowledge: someone close to the problem, someone new to it, someone who will have to make it work. Choose a facilitator who stays neutral and a scribe who writes fast.
Warm up and post the rules 5 min
Read the four rules aloud and pin them up. Then spend two minutes on a silly practice question so people get used to saying odd things out loud.
Start silent, then go loud 15–25 min
Give everyone three minutes to jot ideas alone. Then go around the room once, then open it up. The scribe writes every idea, numbered, in the speaker's words. Aim for a target, such as 60 ideas.
Push when it slows 5–10 min
Lulls are normal. Restart with a twist: “What would a hotel do?”, “What if money were no object?”, “Combine #12 and #30.” The ideas that come after the first lull are often the freshest.
Sort later, separately 15 min
Stop generating before anyone judges. Cluster the ideas into themes, then shortlist with a dot vote or a NUF Test. Keep the full list: today's reject may solve next month's problem.
✓Do
Give people a few silent minutes before anyone speaks
Write every idea down, in the words it was said
Keep ideas short, and number them
Thank people out loud for wild ideas
✕Avoid
Letting the most senior person speak first, or most
“That won't work” and other quick verdicts
Debating one idea while everyone else waits
Picking a winner in the same breath
What you need
A whiteboard or flipchart, markers, a timer, the question written large and the four rules posted. One facilitator, one scribe. With AI, start with the prompt below.
04 Worked example
Brainstorming in action: the school pickup jam
The challenge
“Every afternoon the streets around two of our elementary schools gridlock, and the neighbors call my office.”
Setting: a small city's transportation office. Goal: fewer idling cars and safer crossings by next fall. Constraint: almost no budget. Group: eight people from the city, the schools and the parent association.
8 people · 40 minutes
1
Frame
The group reworded “fix pickup traffic” as “How might we get kids home safely without a line of idling cars?” Focusing on kids, not cars, opened new doors.
2
Warm up
Two minutes on “uses for an orange traffic cone” had the city engineer and the parent association chair laughing and calling out ideas side by side.
3
Flood
First came the familiar: staggered dismissal, more crossing guards, idling fines. Then a bike train, pickup by text message and a “kiss and go” lane.
4
Build
“Walking school bus” (#14) plus “park at the church lot” (#31) became a staffed walk from a lot four blocks away. “Ban all cars” became a car-free school street pilot.
5
Sort
The facilitator stopped generating at 25 minutes. The group sorted 71 ideas into four themes: timing, walking and biking, parking, and talking to parents.
The haul
71 ideas, four themes and a shortlist of three after a dot vote. The wildest idea, closing the street, made the final three.
The shortlist, scored after the vote
Car-free school street pilot
IMPACTEASESUPPORT
Park-and-walk route
IMPACTEASESUPPORT
Staggered pickup times
IMPACTEASESUPPORT
05 ✦ Go AI
Brainstorm with a partner who never runs dry
A brainstorm stalls when the obvious ideas run out, and that is exactly when AI helps most: it can throw in fresh angles in seconds. But it doesn't know your people, your politics or what's been tried before. Let AI widen the list and restart the room; let the humans judge what fits.
YouYou bring
The question, and why it matters now
Local knowledge: what's been tried, who decides
The energy and trust in the room
The final call on what's worth testing
⇄
AIAI brings
Dozens of ideas in a minute to restart a stalled room
Angles from fields your group never considered
A neutral voice with no rank and no pet idea
Quick clustering of a messy list into themes
MOVE 01
Sharpen the question
Before the session, ask for several ways to word your question, from narrow to broad.
Reword this as 5 different 'How might we...?' questions, from narrow to broad: [problem]MOVE 02
Restart a lull
When the room goes quiet, ask for prompts that push the group somewhere new.
We're stuck at 40 ideas on [question]. Give me 5 'what if' prompts that push us somewhere new.MOVE 03
Build on numbers
Feed AI two ideas from the wall and ask for hybrids the group can react to.
Combine idea #14 (walking school bus) with #31 (church lot parking). Give me 3 hybrids.MOVE 04
Cluster the haul
Paste the raw list and let AI propose themes, which the group then checks and renames.
Group these 70 ideas into 4-6 themes. Flag duplicates and any idea that fits nowhere.
Watch out for
Human ideas first. Fill the wall before adding AI's ideas, or people anchor on them.
Generic in, generic out. Tell AI the place, the people and the limits.
Check the facts. AI may claim another city or school did something it never did.
Keep it a group decision. AI can sort ideas, but people need to own the choice.
06 Your prompt
Your brainstorm facilitator, ready to paste
Describe your question, pick a mode, then copy the prompt into Claude, ChatGPT, Gemini or Copilot. In coach mode the AI runs the session like a facilitator: you go first, it builds on your ideas, and judgment waits until the end.