The owner of a small catering company gathers her team of six for a brainstorm about new corporate lunch packages. She writes the question on the whiteboard and asks for ideas. Within a minute, the head chef and the sales manager are bouncing suggestions off each other enthusiastically. The others nod along. The newest team member, who handles deliveries and hears what clients actually complain about, says nothing at all.
After forty minutes, the whiteboard holds twelve ideas, mostly variations on the chef’s first suggestion. Everyone agrees it was a productive session.
It probably felt that way. But decades of research suggest that the classic group brainstorm, despite its popularity, is one of the less effective ways to generate ideas. The good news is that a simple change to the format can make a big difference.
One Like insight: The best ideas in the room are often the ones nobody got a chance to say.
Why the group brainstorm disappoints
Brainstorming was popularised in the 1950s by the advertising executive Alex Osborn, who argued that groups generating ideas freely, without criticism, would produce far more than individuals. It was an appealing idea, and researchers soon began testing it.
The results were consistently awkward. Study after study compared “brainstorming groups” working together with “nominal groups”: the same number of people working alone, whose ideas were then pooled. The nominal groups usually produced more ideas, and often better ones. In 1987, the German psychologists Michael Diehl and Wolfgang Stroebe set out to understand why, and identified a key culprit they called production blocking. In a group, only one person can speak at a time. While waiting their turn, people forget ideas, lose their train of thought or decide their idea is no longer relevant.
Other researchers have added further explanations. Some people hold back for fear of being judged, especially when the boss or a confident colleague is in the room. Others coast, knowing someone else will keep the conversation going. And early ideas tend to anchor the discussion, pulling everyone towards variations of what was said first.
Group discussion isn’t useless. It’s good at combining, refining and choosing ideas. The research suggests it’s simply not the best way to generate them in the first place.
Alone first, then together
In 2010, Karan Girotra, Christian Terwiesch and Karl Ulrich, then of INSEAD and the Wharton School, published a study in Management Science that asked a slightly different question. Most businesses don’t need lots of average ideas. They need a few excellent ones. So the researchers measured the quality of the best ideas each group produced, not just the number.
They compared two approaches. In the team approach, everyone worked together from the start. In the hybrid approach, individuals first worked on their own, then came together to share, discuss and choose. The hybrid groups were more productive, came up with stronger ideas, and were better at judging the quality of what they had produced. They also found that the much-recommended practice of building on each other’s ideas didn’t help: ideas that built on earlier suggestions weren’t better on average.
The hybrid approach works because it captures the strengths of both. Working alone protects each person’s thinking from being crowded out or anchored by others. Coming together afterwards lets the group combine perspectives and judge which ideas are worth pursuing.
[Infographic placement] Two ways to run a session. Left, Classic brainstorm: a circle of people with speech bubbles overlapping, two large, four small; caption “Fewer ideas, anchored early.” Right, Hybrid: step 1, six people writing alone (10 minutes); step 2, ideas posted on a wall; step 3, group discusses, combines and votes. Caption: Girotra, Terwiesch and Ulrich (2010): hybrid groups generated more and better ideas, and were better at picking the best.
Running a better idea session
The practical change is simple. Before your next idea session, send everyone the question a day in advance, or give them ten quiet minutes at the start to write down as many ideas as they can, on their own, without discussion. Then put every idea up where everyone can see it, ideally without names attached. Only after that should the group discuss, combine and choose.
A few other habits help. Ask the most senior person to speak last, so their opinion doesn’t anchor everyone else’s. Make it clear that rough, half-formed ideas are welcome. And when choosing, consider letting people vote privately before discussing, so quieter team members’ preferences count as much as the loudest.
A few questions can help:
- Do our idea sessions give everyone time to think alone first?
- Whose voices tend to dominate, and whose do we rarely hear?
- Do we judge ideas by their merit or by who suggested them?
- How do we choose which ideas to pursue?
Which brings us back to the catering company and its whiteboard of twelve similar ideas. With ten silent minutes at the start, the delivery driver might have written down what clients keep asking for: lunches that arrive labelled for different dietary needs, so nobody has to ask awkward questions in front of colleagues. It’s the kind of idea that only comes from someone close to the customer, and it’s exactly the kind a traditional brainstorm tends to lose.
Sources
- Michael Diehl and Wolfgang Stroebe, “Productivity Loss in Brainstorming Groups: Toward the Solution of a Riddle”, Journal of Personality and Social Psychology 53(3), 1987.
- Karan Girotra, Christian Terwiesch and Karl T. Ulrich, “Idea Generation and the Quality of the Best Idea”, Management Science 56(4), 2010.
