Idea brainstorm, run.
Also called: Ideation session · Solution brainstorm · Divergent thinking workshop · Concept generation
A structured session that turns the evidence gathered in Discover into a spread of solution directions, then narrows them to the few worth pursuing.
Diverge first, converge second, both on purpose. Feed the room the evidence, generate widely, then cut hard to a handful of directions. Done well it ends Discover with real options. Done as sticky-note theatre it ends with a wall of nothing.
What running an idea brainstorm is
A brainstorm is two movements in one session, and most people only run the first. You diverge: generate as many distinct ways of solving the problem as the room can stand. Then you converge: hold those options against the evidence and cut them down to the few worth taking into Stage 03. Skip the converging half and you have a wall of sticky notes and a vague feeling of progress, which is not the same thing as a decision.
The popular version of brainstorming has done real damage here. “There are no bad ideas” is a fine warm-up rule and a terrible operating principle, because by the end of the session some ideas are demonstrably worse than others and pretending otherwise just postpones the cut. The point of diverging is not to be nice. It is to widen the search before you narrow it, so the direction you commit to has actually beaten some alternatives rather than being the first thing anyone shouted.
The thing that separates a useful session from a feel-good one is the evidence. By the time you reach this activity you have a persona, a market context, an IP scan and stakeholder interviews behind you. That is the raw material. A brainstorm that ignores it (which is most of them) generates clever solutions to a problem nobody confirmed exists.
Why this gets run badly
- The room rewards volume, not relevance. Whoever fills the most sticky notes feels productive, so the session optimises for quantity and quietly forgets to ask whether any of it answers the actual problem.
- Nobody owns the cut. Diverging is fun and converging makes enemies, so the session ends at the fun part and the hard part gets emailed round later, where it dies.
- The loudest voice wins. Without a structure, the most senior or most confident person anchors the room in the first ten minutes, and you spend the rest of the hour decorating their idea.
The session, on one canvas
The clearest way to see a good brainstorm is to look at one that worked. Here is the proofing box’s session, the one we ran near the end of Discover, so you can see the shape rather than a generic template.
Eleven directions in, two out the other side. The session did its job not because it generated a lot but because it threw most of it away against criteria the room agreed before it started arguing.
- “No bad ideas” held as a rule for the whole session, not just the warm-up.
- Generate widely, then photograph the wall and call it a day.
- No evidence in the room, so no way to tell a good idea from a loud one.
- The cut quietly happens later, alone, in someone’s inbox.
- Diverge freely, then converge openly, in the same session with the same people.
- Agree the filters before you generate, so the cut is fair not political.
- Bring the persona, market and IP evidence into the room as raw material.
- Leave with a short list of directions and the reasons each one survived.
How it fits the bigger picture
Run Idea Brainstorm is the last activity of Stage 02 Discover. It gathers up everything the stage produced (the value goals, the persona, the market context, the IP scan and the stakeholder interviews) and turns that evidence into a handful of solution directions. Those directions are exactly what Stage 03 Innovate then develops, refines and concept-tests. Discover finds the problem; this activity opens the door to solving it.
What it can do
It widens the search before you commit, so the direction you carry into Innovate has beaten some alternatives rather than being the first idea anyone liked. It also surfaces the runners-up, which become useful fallbacks the moment the favourite hits trouble in Stage 03.
What it can’t do
It can’t prove a direction is right. A brainstorm produces options and a defensible shortlist, not validated answers. That is what Stage 03 Innovate’s concept work and Stage 04 Evaluate’s commercial assessment are for. Treat what survives the session as the best current bet, not a verdict.
See the full 10-stage process →
Try it yourself
Book one hour. Write the problem as a single “how might we” prompt on the board. Spend the first half diverging: everyone generates directions, quietly first so the room doesn’t anchor on the loudest voice, then out loud to build on each other. Spend the second half converging: agree three filters out loud, score every direction against them, and cut to the two or three that survive. Park the rest in a dated notebook, not the bin.
Or run the guided version. The Free Sprint walks you through framing the problem and weighing directions before you commit. Start the Free Sprint →
Your brainstorm checklist
Project notes: eleven directions, one box
▸ From the notebook · optional reading
An hour in Stockport, eleven ways to warm a dough, and the three filters that did the cutting for us.
3 min read · click to open
Dan wanted to start sketching a box on day one. I asked him to wait. We had a persona, a market context and a stack of stakeholder notes, and it seemed a waste to skip straight to the first idea anyone fancied. So we booked an hour and put one prompt on the wall: hold dough at 26°C overnight in a cold UK kitchen.
The diverge half
We worked quietly first, each person writing directions before anyone spoke, which kept Dan’s very confident “it’s obviously a heated box” from anchoring the room before we’d looked. Then out loud, building on each other. Eleven directions went up: heated box, heated mat under a cloche, an insulated cool-box with a small warmer, a smart-plug oven hack, a water bath, a fridge run in reverse, and a few that were, charitably, warm-up material. (One involved a hot-water bottle and a tea towel. It did not survive.)
The converge half
Then the part most rooms skip. I asked the baker on the call what would actually live on her worktop overnight, and we turned her answer into three filters everyone agreed before we started cutting: holds 26°C give or take half a degree all night, looks like something you’d be glad to own, and fits the sub-30W no-app brief. The oven hack failed on temperature stability. The cool-box failed on the worktop test, nobody wants a camping box by the kettle. The water bath failed on both.
What survived was the heated ceramic box with a low-voltage element. The heated mat and the reverse-fridge went into a dated notebook as fallbacks, which mattered later when we hit a tooling question in Stage 03 and could glance back at a real alternative rather than starting cold.
The session cost an hour. It gave us a direction the whole room could defend and two fallbacks we never had to invent in a panic. Worth far more than the sketch Dan wanted to start with.
— Discover stage, project notes, 2026
— Next stage → Stage 03 · Innovate