Brainwriting: how the 6-3-5 method beats a group brainstorm

Brainwriting is silent, written idea generation: instead of calling ideas out in a discussion, everyone writes them down at the same time and then passes them on to be built upon. The best-known format is 6-3-5, where six people each write three ideas in five minutes, then pass their sheet along, five times. It removes the two things that most reliably suppress a group brainstorm: waiting your turn, and worrying who is watching.

The technique came out of German design and innovation practice in the late 1960s, credited to Bernd Rohrbach, and it has quietly outlasted a lot of louder methods. The reason is unglamorous: it fixes a mechanical problem with a mechanical solution.

What is brainwriting?

Brainwriting is a structured brainstorming technique where participants generate ideas in writing, in parallel, without speaking. Everyone works at once on the same challenge, ideas are recorded before any discussion happens, and each round builds on what someone else wrote rather than on what the loudest person said.

The mechanics vary, but three properties are always present. Generation is simultaneous, so nobody waits. It is written, so nobody is performing. And it is iterative: your sheet does not stay yours, it moves to the next person, who has to extend the ideas already on it.

How does the 6-3-5 method work?

6 participants, 3 ideas, 5 minutes, repeated six times. Each person gets a sheet divided into a grid of eighteen boxes: three columns for ideas, six rows for rounds.

1. State the challenge as one question and write it at the top of every sheet. "How might we cut first-month churn?" works. "Churn" does not.
2. Round one, five minutes. Everyone silently writes three ideas in the top row. One box, one idea, a sentence each. No talking, no judging, no editing.
3. Pass the sheet to the person on your left.
4. Round two, five minutes. Read the three ideas in front of you, then write three more in the next row that build on, extend, or react to them. Not three fresh unrelated ideas: three that stand on the previous row's shoulders.
5. Repeat until every sheet has been round the table. Six rounds, thirty minutes, and the arithmetic gives you 108 written ideas.
6. Then converge, which is a separate exercise and the one that gets skipped.

The 108 figure is the method's marketing, and you should not take it too seriously. In a real session, many boxes will be near-duplicates and a handful will be blank because someone stalled. Expect maybe thirty to fifty distinct directions from a good session with six engaged people. That is still far more than a talk-out-loud brainstorm of the same length will produce, and the ideas are better distributed across the room rather than clustered around two confident voices.

Why is brainwriting better than brainstorming?

Because classic verbal brainstorming has a design flaw that no amount of enthusiasm fixes: only one person can talk at a time.

Researchers call it production blocking. While you wait for your turn, you are holding your idea in working memory, listening to someone else's, and rehearsing what you will say. The idea decays, or you drop it because it now sounds too close to what was just said. Diehl and Stroebe demonstrated in 1987 that this waiting mechanism accounts for a large share of the gap between real groups and "nominal" groups, which is what you get when the same number of people generate ideas separately and you pool the lists afterwards. The separated individuals win, and they have been winning in study replications since Taylor, Berry and Block first ran the comparison in 1958.

Two other effects pile on. Evaluation apprehension keeps people from floating the weird idea in front of a boss, and the weird idea is the one you were paying for. Social loafing lets the quiet half of the table coast, because in a discussion nobody notices who contributed nothing. Brainwriting takes a swing at all three at once:

What brainwriting does not fix is the ceiling on the room itself. Six people can only write down what six people know. If everyone in the room shares a background, you get a very efficient exploration of a very narrow space, and the sheets fill up with three variations of the same instinct. That limit is structural, not procedural, and it is why the method pairs so well with an idea generator that supplies angles nobody in the room would have reached.

How do you run brainwriting remotely?

Better than you run it in a room, honestly, because the format was always secretly asynchronous. The passing of paper is just a way to serialize contributions, and software does that without a table.

The remote version: post the challenge as a single question, give everyone a private space to write their first three ideas without seeing anyone else's, then reveal the pool and ask each person to extend three ideas that are not their own. You can run it across a day instead of across thirty minutes, which suits distributed teams and lets people think at their own pace. An online brainstorming tool with a shared wall and private first-round entry handles this cleanly, and it solves the other half of the problem too: the sheets end up as a sortable pile rather than a stack of photographs.

One rule matters more than the rest. Keep round one blind. The moment people can see each other's first ideas, anchoring comes back through the front door and the whole point of the exercise evaporates.

Is there a brainwriting template?

The template is genuinely trivial, which is part of the method's charm: a grid, three columns wide and six rows deep, with the challenge written across the top. You can draw it on paper in ten seconds, and there is no reason to pay for one.

What is worth setting up carefully is the challenge statement itself. "How might we..." phrasing, one sentence, specific enough that a stranger could write an idea against it. A vague prompt produces vague sheets and no template fixes that. Our guide on how to run a brainstorming session covers the framing step in more detail.

What do you do with 100 ideas?

This is where most brainwriting sessions quietly fail. The generation phase works, everyone feels productive, and then someone volunteers to "type these up," and six weeks later the sheets are in a drawer. A pile of ideas is not an outcome. A decision is.

Converge in a second, separate sitting, and treat it as real work rather than a wrap-up. Group the ideas into themes with affinity clustering, which usually collapses a hundred boxes into five or six honest directions. Score each idea on impact against effort, out loud, so the trade-offs are visible. Then pick one, write down why you picked it, and write down what would have to be true for the runner-up to win instead. That last sentence is what makes the decision reviewable in three months when the world has changed. The mechanics are on our idea prioritization page.

Where the ideas go next depends on what you were brainstorming. Product bets become experiments with a success metric. Campaign angles become something you can actually test, and the fastest honest test is usually to turn the two strongest angles into real ad creative and see which one people click, rather than arguing about them for another week.

Brainwriting versus a generated wall

Brainwriting is a fix for a coordination problem: it stops the room from getting in its own way. It does nothing about the knowledge problem, which is that the room only contains what the room already knows. Both problems are real, and they need different tools.

The combination that works best in practice is to generate first and write second. Put the challenge into an idea generator, get two dozen genuinely different directions on the wall in thirty seconds, each tagged with the angle that produced it. Now run the brainwriting round with those as seeds rather than from a blank grid. Round one is no longer six people staring at an empty box trying to be clever on demand; it is six people reacting, extending and arguing with material that is already in front of them, which is the thing humans are unambiguously better at than machines.

Then converge as a group, because judgment is the part you should not outsource. The sequence is the point: the machine supplies breadth, the written round supplies depth and ownership, and the room supplies the decision. That is a brainstorm that ends with something you can fund.