A mind map your whole workshop edits at once.

Alex Osborn made brainstorming famous in the 1950s, promising that a group would produce more ideas together than its members would alone. Decades of research found the opposite.

In a 1987 study, Michael Diehl and Wolfgang Stroebe tested why groups fall short. The main cause wasn't laziness or fear of being judged. It was simply taking turns: since only one person can speak at a time, everyone else waits, and ideas fade or never form. Psychologists call this production blocking.

That doesn't make workshops a bad idea. You still need people in the same conversation to build on each other and agree. It means the part where people come up with ideas shouldn't depend on taking turns.

• People wait. With one person talking, everyone else is holding an idea or has already lost it.
• One person holds the pen. Whoever shares their screen decides what gets written and how it's worded.
• Rank wins. Once the most senior person speaks, the discussion tends to line up behind them.
• The result gets lost. A photo of a wall or a board nobody reopens can't carry a decision forward.
• Nothing happens next. Everyone leaves feeling it went well, and nobody knows who is doing what.

The team below is made up for this page: a small invoicing app whose new users sign up and then never send an invoice. Every screenshot is a real Shiko map we built by following these steps.

Before: one question and a code

The facilitator creates a map from the Brainstorm template and writes the challenge at its centre: how might we get new users to a first invoice on day one? They turn on the Decision matrix plugin, which only a map's owner can do, and create a room code that lasts until the session.

During: everyone writes at once

The template has three branches on purpose. Wild ideas gives people permission to be silly, Practical ideas covers what could ship, and Borrowed ideas asks what works elsewhere. For ten minutes nobody talks. Nine ideas land on the map, from “We send their first invoice for them” to “Sample invoice they can edit”.

Shiko/Website relaunchMJPShare

Website relaunch

Goals

Content

Progress2 / 3

Final copy

Redirects

Announce

Case studies

Maya

Jonas

Priya

Then: answer what you don't know

Before scoring anything, the group asks where new users actually give up, and a question node records the answer they settle on: adding a line item. Now the ideas that fix that step look more attractive.

Deciding: scores, not volume

Three ideas make the shortlist. The facilitator types them into a Decision matrix with three criteria, and the scores are debated out loud. It's all plain text, so changing a score takes a second, and the totals update as you type.

Guided first invoice wins by a single point, on Impact. It's close enough that the team can see it was a real choice, and anyone who wasn't there can see how it was made.

After: one owner, one date

Before anyone leaves, Maya takes the next step: build the guide, and test it with five new users. The task node sits right beside the matrix that justified it.

The session ends with the whole conversation on one map: every idea that was considered, what the group learned, what it chose and who is doing it. Nobody has to write it up afterwards.

A whiteboard gives people a place to put ideas, and the sorting usually happens later. In Shiko, ideas, decision and next step are one structure, built as the session goes.

People wait their turn and lose their ideas

Everyone types into the same map at once, so nobody has to hold a thought until there’s a gap in the conversation.

One person holds the pen

There is no scribe to filter what gets written down. Each person adds their own words, under their own name.

The most senior opinion decides

A Decision matrix makes everyone argue about the scores. The result, and how it was reached, stays on the map for all to see.

The output is a photo of sticky notes

The map is the output. It stays editable after the session, with every idea still where the group put it.

Nobody owns the next step

Next steps are written on the map before people leave, as a task node with an owner and a date.

Most workshops hit at least one of these. Here's what to do in the moment.

The room goes quiet

Point at an empty branch and ask for three bad ideas. A deliberately bad idea is easy to write, and it often leads to a good one. The Wild ideas branch is there for this.

One person keeps talking

Say "let’s all write for three minutes" and stop the discussion. Written ideas are harder to talk over, and quiet people get the same space.

Everyone debates the first idea

Remind them that judging comes last. The template says so on its first node: fill each branch first and save the judging for the end.

Two options tie

Check the criteria, not the options. A tie often means the criteria don’t capture what the group really cares about. Change one and the scores update.

Time runs out

Skip anything but the last step. Agreeing on one next step, with an owner and a date, matters more than finishing the agenda.

1. 5 minutes, frame it. Read the question aloud and say what a good outcome looks like.
2. 10 minutes, write in silence. Everyone adds ideas at once, with no discussion.
3. 15 minutes, read and react. Go through the map together. Add comments and fill the gaps.
4. 5 minutes, shortlist. Pick two to four options worth scoring.
5. 15 minutes, decide. Agree on the criteria first, then score the options in a Decision matrix.
6. 10 minutes, assign. Write the next steps as task nodes, each with one owner and a date.

• Diehl and Stroebe, Productivity Loss in Brainstorming Groups (1987)
• Production blocking, Wikipedia