Brainstorms that don't reward the loudest extrovert, and produce three times the ideas

Group brainstorming has a problem that researchers at Yale and Wharton have been pointing out since the 1980s. It produces fewer ideas than the same people working alone, and the ideas that do surface skew toward the early-mentioned and the loudly-defended.

Most product teams know this and still run group brainstorms, because the alternative, "work on it alone and come back," feels lonely and unstructured. The team gathers, the whiteboard fills up, the extroverts dominate, the group converges on a direction within fifteen minutes, and the actual range of possible ideas gets compressed before the introverts have decided whether to risk the second idea they were considering.

The fix is not to skip the group session. It's to do the ideation alone first, then converge as a group.

Async ideation, before the meeting, on a shared document where everyone drops three to five ideas, in their own time, without seeing anyone else's contributions until they've posted theirs. Twenty minutes of focused individual thinking, structured by a clear prompt ("how could we reduce trial-to-paid friction"), produces more ideas per person than two hours of group whiteboarding. The cognitive science term is anchoring; the practical effect is that the first idea spoken in a group session pulls every subsequent idea in its direction, and the team converges on the neighborhood of that first idea instead of exploring the full space.

Once the async ideas are posted, the group session does work it's actually good at: ranking, combining, killing. The 30-idea list gets read in the first five minutes. The team spends the next 40 minutes asking "which of these are actually different ideas and which are variations of the same one," "which two of these combine into something better," and "which can we just decide to not pursue right now." The output is three to five real candidates, not 30 stickies that nobody knows what to do with.

The teams that ship interesting product work tend to use this pattern, often without naming it. The async ideation matches what individual researchers do anyway, sitting with the problem, generating options, refining. The group session matches what teams are uniquely good at, comparison and decision. The mistake most teams make is collapsing both into one whiteboard session where neither happens well.

There's a second-order benefit that's worth naming. Async ideation surfaces ideas from people who never get airtime in group sessions. The senior engineer who thinks before she speaks, the designer who needs a quiet hour with the problem, the CS lead whose insight is rooted in a specific customer conversation she'd rather write up than perform. Their best work shows up on the page before the meeting, and the room they walk into is one where their thinking is already on the table.

The most common objection is that async ideation feels like homework. It does, if the ask is unbounded. Three to five ideas, twenty minutes, one shared document, posted within 24 hours of the prompt going out, is a constraint that lifts the burden. The team that adopts it stops doing the second-rate group whiteboarding and gets a better output from less total time.

If your team is about to run a brainstorm, the test is whether the people in the room are going to leave with an idea they didn't get to share. If yes, the room is doing the wrong work. The async window before the meeting is where the missing ideas live.