Myth 15 · Myths
The brainstorm that produced fewer ideas
Brainstorming in a group is supposed to produce more and better ideas, but many experiments found that the same number of people working alone, then pooling ideas, came up with more.
- Diehl and Stroebe, Journal of Personality and Social Psychology
- Tübingen, Germany
- 1987
- 3 min read
On this page
Before you readWhat would you do?
Your team needs new ideas for a marketing campaign. Your manager plans a one-hour group brainstorm that starts with everyone reading a list of 50 ideas from a chatbot. What change would most likely improve the results?
Pick an answer now. The file tells you whether you were right.
01The file in one minute
1The myth
Brainstorming as a group always produces more and better ideas than people working alone.
2The tests
From 1958 on, experiments found people working separately, with ideas pooled later, produced more ideas.
3Why
In a group only one person speaks at a time, so others wait, forget ideas or stop generating them.
4The lesson for AI
Let people think alone first, then use AI to add or challenge ideas, so nobody anchors on its list.
02Timeline
1953
Alex Osborn promotes group brainstorming in Applied Imagination.
1958
A Yale study finds individuals working apart produce more ideas than groups.
1987
Diehl and Stroebe identify production blocking as a main cause.
1991
A meta-analysis confirms groups tend to produce fewer ideas.
03The myth
In his 1953 book Applied Imagination, the advertising executive Alex Osborn promoted brainstorming: a group shares ideas freely, nobody criticises, and wild ideas are welcome. He claimed groups working this way could produce far more ideas than people working alone.
Researchers soon tested it. In 1958 a Yale team compared real groups with the same number of people working separately, whose ideas were then combined. The separate individuals produced more ideas. Many later studies found the same.
In 1987 the German psychologists Michael Diehl and Wolfgang Stroebe tested why. The biggest cause was "production blocking": in a group only one person can speak at a time, so others wait, forget ideas or stop thinking of new ones. A 1991 meta-analysis of many studies confirmed that groups tended to produce fewer ideas than the same people working alone.
04What the evidence says
Group sessions have other costs too: people worry about being judged, a few voices dominate, and the group drifts toward the first ideas mentioned. Group discussion still has real value for choosing and improving ideas, and for getting people to agree on a plan.
Methods that work better often separate the steps: people write ideas alone first, sometimes called brainwriting, then the group shares, combines and picks.
The AI angle: if your whole team looks at an AI tool's list of ideas first, everyone may anchor on it. Have people think alone first, then use the AI to add or challenge ideas.
The idea behind this caseLesson: Ask it to think step by step →
05How to spot the next one
- Ask everyone to write down ideas alone for a few minutes before the group talks.
- Use the group meeting to combine, improve and choose ideas, where groups are strong.
- Collect ideas in a shared document so people do not have to wait their turn.
- Bring in AI suggestions after people have had their own ideas, not before.
07 — Sources
- Productivity loss in brainstorming groups: Toward the solution of a riddle · Journal of Personality and Social Psychology · 1987-09
- Productivity Loss in Brainstorming Groups: A Meta-Analytic Integration · Basic and Applied Social Psychology · 1991-03