The Brainstorming Origin Story
In 1948, advertising executive Alex Osborn published Your Creative Power, introducing a technique he called "brainstorming." The rules were simple: defer judgment, encourage wild ideas, build on others' suggestions, and aim for quantity. Osborn claimed that groups using his method could generate twice as many ideas as individuals working alone. The concept spread through American corporate culture like an airborne pathogen. By the 1970s, the brainstorming session was a fixture of organizational life, appearing in boardrooms, design studios, and strategic planning retreats.
There was one problem. Osborn never tested his claim experimentally. He based it on anecdotal observation at his advertising agency, BBDO. When researchers began testing the method in controlled settings, the results were immediate and unambiguous.
What the Meta-Analysis Found
Brian Mullen, Craig Johnson, and Eduardo Salas aggregated 20 studies that all used the same basic design. Participants were randomly assigned to either an interactive brainstorming group (following Osborn's rules) or a "nominal group" condition. In nominal groups, the same number of individuals worked alone on the same problem for the same duration, and their non-redundant ideas were pooled afterward.
Interactive groups produced fewer ideas than nominal groups across every productivity measure. The deficit held for total ideas, unique ideas, and categories of ideas. It held across group sizes, though the productivity loss was worse in larger groups. It held regardless of whether participants were students, professionals, or managers. The effect sizes ranged from d = 0.60 to d = 0.82, meaning the average nominal group outperformed roughly 73-79% of interactive brainstorming groups.
Larger groups showed larger losses. A four-person brainstorming group didn't just generate fewer ideas than four individuals working separately; the gap widened compared to a two-person group versus two individuals. Every additional person added to the group increased the per-person deficit, suggesting the mechanism was social, not cognitive.
Three Mechanisms That Kill Group Creativity
Production blocking is the dominant culprit. In a group, only one person can speak at a time. While waiting for a turn, other members must hold their ideas in working memory while simultaneously listening to someone else's contribution. Working memory decays rapidly under interference. By the time their turn arrives, people have forgotten ideas they were ready to share seconds earlier, or have redirected their thinking to follow the speaker's thread rather than their own.
Evaluation apprehension compounds the loss. Despite Osborn's "no criticism" rule, group members still modulate their contributions based on anticipated social judgment. Unusual ideas are self-censored before they reach the air. The more senior or authoritative the other group members, the stronger this filtering effect becomes.
Social loafing provides the third drain. In groups, individual effort is difficult to attribute, so each person contributes slightly less effort than they would working alone. This effect, documented across dozens of group task paradigms, represents the reliable human tendency to coast when accountability is diffused.
Yet People Believe Groups Are Better
The most perplexing finding in this research is that participants in brainstorming groups consistently rate their sessions as more productive, more creative, and more enjoyable than participants in nominal groups rate their solo work. The illusion of group productivity is robust. Paulus and Dzindolet (1993) found that group members overestimated their output by roughly 40%, while nominal group members were relatively accurate.
The illusion has at least two sources. First, social comparison within groups creates the impression of high productivity because members hear many ideas, even though they personally contributed fewer. Second, the experience of group work is genuinely more stimulating and pleasant than working alone, and people confuse enjoyment with effectiveness.
The Strongest Counterargument
Defenders of group brainstorming argue that idea quantity is the wrong metric. Groups may produce fewer total ideas but more diverse categories of ideas through cross-pollination. Some research on "brainwriting" (a hybrid method where individuals write ideas, then exchange and build on them) shows category diversity advantages. Group brainstorming may also serve organizational functions beyond idea generation: alignment, buy-in, shared understanding, and relationship building.
These points have merit, particularly the organizational functions argument. But on the core claim that brainstorming was invented to support, that groups produce more and better creative ideas, the evidence is not ambiguous. They don't. Hybrid methods that start with individual work and then combine it in structured group discussion consistently outperform pure group brainstorming.
What We Didn't Prove
The studies in this meta-analysis primarily used simple divergent thinking tasks (generate uses for a brick, names for a product). Real organizational problems are more complex and may benefit from group interaction in ways these tasks don't capture. Most participants were undergraduate students; professional teams with established working relationships and domain expertise may compensate for some of the documented losses. Electronic brainstorming tools that eliminate production blocking show improved performance, suggesting the problem is the format, not the collaboration.
The Bottom Line
The Bottom Line
Group brainstorming suppresses individual creativity rather than amplifying it. The productivity loss is large, consistent, and driven by well-understood mechanisms. People enjoy brainstorming sessions and believe they work. They are wrong on the second count. Thirty-five years of controlled research all points the same direction.
What You Can Do
Start every ideation session with individual work. Give people 10-15 minutes to generate ideas alone, in writing, before any group discussion. Then use structured sharing: round-robin presentation, anonymous submission via sticky notes or digital tools, or the "6-3-5" method (6 people, 3 ideas each, 5 rounds of building). Reserve live group discussion for evaluating and combining ideas, where social interaction actually adds value. If you manage a team, experiment with brainwriting versus traditional brainstorming on parallel projects and compare the output. The difference will be visible in a single cycle.