Running a Productive Workshop or Brainstorming Session

What research says about group brainstorming, and how to plan, structure and facilitate workshops that turn everyone's ideas into decisions.

Collaboration Center · Last reviewed October 23, 2026 · 6 min read

Workshops and brainstorming sessions promise something every team wants: more ideas, better ideas and shared ownership of what comes next. Too often, though, they turn into a few confident people talking while everyone else waits, followed by a whiteboard photo that nobody looks at again.

Research on group idea generation helps explain why, and it points toward simple design changes that make sessions more productive. This guide covers what the research says, how to plan a workshop, structures that get everyone contributing, and how to turn a pile of sticky notes into decisions.

What research says about group brainstorming

Brainstorming has been promoted as a method of group problem solving since at least the 1950s, and researchers have tested it for decades. The results are more mixed than its popularity suggests.

A key comparison in this research is between "real" groups, who generate ideas together out loud, and "nominal" groups, whose members work alone and whose ideas are pooled afterward. In a 1991 meta-analysis in Basic and Applied Social Psychology, Brian Mullen, Craig Johnson and Eduardo Salas integrated earlier studies and found that brainstorming groups were generally significantly less productive than nominal groups, in both the quantity and quality of ideas. The productivity loss was stronger in larger groups.

Why? In a 1987 set of four experiments published in the Journal of Personality and Social Psychology, Michael Diehl and Wolfgang Stroebe tested three explanations: free riding, fear of being judged (evaluation apprehension) and production blocking, which is the fact that only one person can talk at a time while everyone else waits and may forget or abandon their ideas. They concluded that production blocking accounted for most of the productivity loss in real brainstorming groups.

The picture changes when people can contribute in parallel. In a 1993 experiment in the Journal of Applied Psychology, Alan Dennis and Joseph Valacich found that 12-member groups using computer-based electronic brainstorming generated more ideas than 12-member nominal groups, while 6-member electronic and nominal groups did not differ. They attributed this to electronic brainstorming reducing process losses such as production blocking while still letting people build on each other's ideas.

The practical lesson is not "never brainstorm in groups." It is: let people generate ideas individually and in parallel first, then use the group to combine, build on and evaluate them.

Plan the workshop around a clear question and outcome

Most weak workshops fail before they start. Before you book a room, write down:

The question: one specific prompt, such as "How might we cut the time new customers wait for their first delivery?"
The outcome: what the group should leave with, such as "three ideas ranked, each with an owner to test it."
The decision method: who will decide what happens to the ideas, and how.
The participants: people who know the problem from different angles, including those who will live with the result.
The time: a realistic agenda that alternates between generating options and narrowing them down.

Share the question and any background material ahead of time so people arrive having thought about it. For a short session of an hour or two, a facilitator and a note-taker are usually enough. Longer workshops benefit from a co-facilitator to watch the room and track time.

Structures that get everyone contributing

A few well-tested structures put the research into practice.

Silent writing first. Before any discussion, give everyone a few minutes to write ideas on their own, one idea per note or line. This is often called brainwriting. It sidesteps production blocking and gives quieter people the same chance as the loudest.

1-2-4-All. This structure from Liberating Structures, a freely licensed collection of facilitation methods developed by Henri Lipmanowicz and Keith McCandless, moves from individual to whole-group thinking in under 20 minutes:

Introduce the question (1 minute).
Each person reflects silently and writes ideas (1 minute).
Pairs share and build on ideas (2 minutes).
Pairs join into groups of four to refine and note differences (5 minutes).
The whole group reconvenes and a few people share key ideas (7 minutes).

Nominal group technique. For a more structured approach, especially when you need to prioritize, the nominal group technique works well. In a 2016 overview in the International Journal of Clinical Pharmacy, Sara McMillan, Michelle King and Mary Tully describe its four classic stages: silent generation, round robin, clarification and voting, or ranking. People write ideas alone, share them one at a time in turn without debate, clarify what each means, then rank them privately.

Facilitating the room

Talk-based sessions have predictable failure modes. In a June 2019 Harvard Business Review article, Steven Rogelberg and Liana Kreamer noted that traditional talk-based meetings can be "subject to a host of problems: one person dominating, others checking out and multitasking, side conversations, straying off course, and pressures to conform to the boss's ideas."

A few facilitation habits counter these:

Separate generating from judging. During idea generation, ask the group to hold criticism and focus on adding or building.
Have the most senior person share their ideas last, or contribute in writing like everyone else.
Use round-robins so each person speaks in turn.
Keep a visible "parking lot" for important but off-topic points.
Time-box each activity and announce the remaining time.

Converging: from many ideas to a few decisions

Generating ideas is the easy part. Narrowing them takes structure:

Cluster similar ideas and give each cluster a short, descriptive name.
Agree on two or three criteria, such as impact, cost and time to test.
Prioritize with a quick method, such as giving each person a fixed number of votes or using private ranking as in the nominal group technique.
Discuss the top few in depth, including the risks.
Decide using the method you named during planning, and record who decided what.

For example, imagine a 20-person dental practice running a workshop on reducing missed appointments. Fifty ideas might cluster into six themes, and private ranking might surface two to test: text reminders two days ahead and an easier online rescheduling link. Each gets an owner and a check-in date.

Remote and hybrid workshops

Remote sessions can use the research to their advantage. A shared document, form or digital whiteboard lets everyone add ideas at once, which is the parallel contribution that the Dennis and Valacich study points to. Keep activities shorter than you would in person, use breakout rooms for pair and small-group steps, and ask in-room participants in hybrid sessions to use the same digital tools as remote colleagues so nobody becomes a second-class participant.

After the workshop

Send a summary within a day or two with the question, the ideas, the decisions, the owners and the dates. Later, tell participants what happened to their ideas, including the ones that were not chosen and why. That follow-through is what makes people willing to come to the next one.

Key takeaways

Classic out-loud brainstorming tends to underperform individuals working alone, largely because of production blocking.
Start with silent, individual idea generation, then use the group to combine and build.
Plan around one specific question, a clear outcome and a named decision method.
Use structures such as 1-2-4-All or the nominal group technique to include everyone.
Converge with clear criteria and private voting or ranking, then follow up with owners and dates.

Sources

Spotted something out of date? Let us know.