Article

The one thing only you know: getting it into the decision before the room moves on

Say the unshared fact early, name it as something the others do not have, and name the decision it changes. A 2026 experiment with 370 three-person groups found teams decided worse when the newest member held the deciding fact, and false expert credentials did not fix it. Timing and framing do the work. This page is for the person holding the one piece of information nobody else in the room has. It covers why groups talk past it, when in the meeting to say it, and what to do when saying it once changes nothing. Writing your contribution out beforehand is a separate job, and this site has a page on that.

Why does a room ignore the one thing only you know?

Four people agreeing on a launch date, and the fact that breaks it sitting in your head. The vendor contract runs out in March, which you know because you were copied on the renewal thread in August. Nobody else in that meeting was. The date gets agreed, and your fact never reaches the decision.

That pattern has a name in the research, which is the hidden profile. Researchers hand a group a decision, give everybody most of the information, then give one person the piece that changes the answer. To get it right the group has to pool what it holds. It usually does not.

Rajivan and Cooke ran that design on security analysts and published it in Human Factors in 2018. Thirty three-person teams worked two threat-detection missions, with security incidents handed out so that some were known to everyone in the team and some to a single member. When the authors analysed what the teams actually discussed, the incidents the majority already knew came up three times as often as the ones held by one person.

Unaided, the teams were inefficient at finding the links between the uniquely held pieces, which was the whole job.

The paper's own summary of the prior work states the mechanism plainly: when information is spread unevenly, people share what most of the team already knows and hold back what only they have. Nobody in your meeting is ignoring you on purpose, then. The room is behaving the way groups behave when the pieces are split up, and what you hold is on the wrong side of the split.

Carry the direction of that result and not the setting. Those teams were correlating fabricated security incidents in a synthetic task environment, with no budget and no boss attached to any mission, and the aids that reduced it were purpose-built analyst software. Nobody has measured this in your Tuesday standup.

One thing follows immediately, though. Holding a decisive fact is not the same as the room having it, and the gap between those two is yours to close rather than the chair's. If you walk out with the fact still in your head, the record will show a group that decided with everything it had.

When in the meeting should you say it?

Early, and earlier than feels polite. Waiting does not cost you the agenda slot; it costs you the difference between informing a position and unseating one.

How little it takes for a room to acquire a position got measured in 2026. Phlegar, Del Rosario, Dumitru and West put 370 three-person groups, 1,110 people in total, through a hidden-profile hiring task and published it in Communications Psychology. Two strangers met in an online chatroom and spent five minutes discussing the qualities of a job candidate. Then a third person joined them. One member of the finished group, sometimes one of the original pair and sometimes the newcomer, held the critical unshared information the team needed to pick the right candidate out of four.

Teams whose newcomer held that information were less likely to reach the correct decision than teams where an original member held the same thing. Same fact, same task, same three people. The only difference was which of them walked in late with it.

Five minutes of conversation built that gap.

The authors then tried to fix it, which is the most useful part of the study for anyone sitting in your seat. They gave some newcomers false positive feedback on a test of skills supposedly relevant to the decision, so the group would read them as the person who knew what they were doing. Status of that kind did not credibly overpower the newcomer position. The standard advice on getting heard tells you to establish your credibility first and say the thing second, and what got tested here was a borrowed version of that, a credential handed to the newcomer rather than earned. It bought them nothing the authors could measure, which is worth knowing before you spend a meeting building up to your point. Credibility first and fact second is not what this evidence supports.

Two other measurements in the same study describe what holding the fact did to the newcomers themselves. They showed weaker linguistic coordination, meaning they picked up less of the group's own phrasing and style, which is a measure of how aligned people in a conversation are. They also reported more task conflict than newcomers with nothing extra to contribute.

Both of those are associations rather than instructions, and the study does not show that sounding more like the room would have changed any decision. What they suggest is narrower and still worth knowing: holding the unshared piece makes you sound least like the room at the moment you most need it listening to you.

So the levers you actually have are timing and framing, and in practice the framing fits in one sentence. Name the fact, name that the others do not have it, and name the decision it changes: the contract ends in March, which I only know from the renewal thread, so a May launch has no vendor behind it. People drop the middle clause almost every time. It is the clause that stops your date being heard as an opinion about the launch.

Say it in the first pass round the table, not in the summary at the end.

What do you do when saying it once changes nothing?

You said it, somebody said that was interesting, and the discussion resumed from exactly where it had been. The repair for that is not volume.

Rajivan and Cooke tested aids as well as analysts. Teams were randomly assigned different collaboration tools for the second mission. What reduced the pooling bias were displays built to support recognition memory and perceptual comparison, so an incident held by one person sat on a screen the others could see instead of living in one analyst's recollection. Their software is not your meeting chat. The shape of the fix still ports: anything only spoken has to survive everyone else's memory of it, and anything you write down does not.

Put the line in whatever your shared artifact is here. The agenda document, the channel, the thread the decision is actually being made in. One sentence, the decision it bears on attached, in writing before the meeting ends.

A recording or an AI summary is not that artifact, and the difference is worth stating. A transcript captures that you said it, which helps you afterwards and does nothing during. This site has a page on what to write down when a transcript already exists, and the line it draws applies here too: the record and the decision are two separate objects, and your fact has to reach the second one.

Then the harder version, which is a room weighing seniority instead of information. Tangen, Kent and Searston looked at how to combine judgments in fingerprint analysis, in Cognitive Research in 2020. Professional analysts, they confirmed, greatly outperform novices and still make mistakes. Pooling small groups of experts by taking the majority decision cut false positives by up to 8 percent and false negatives by up to 12 percent. Taking the majority also beat following the most confident rater, and it beat following the most experienced one.

Those were independent judgments on the same image, combined by a rule, which is nothing like six people talking for an hour. The caveat earns its space here, because pooling novices in the same study pushed false positives up by as much as 11 percent. More voices is not the finding.

What survives the trip is narrow and worth having to hand: among people who all know the work, the most confident voice and the most experienced one are both worse guides than a straight count of what each of them thinks. The novice result is why that condition has to travel with the claim. When a decision is converging on the seniority of people who do all know the work, that is an argument you can make out loud.

If the meeting closes with your line in the document and the decision made without it, you are in a different situation from the one this site covers under the meeting that ended with no decision. A group that settled on the wrong thing is harder to reopen than a group that settled on nothing. Your timestamp is what makes reopening cheap.

How does your working style change the answer?

Which part of this you get wrong tends to follow how you already work. The four names below are this site's own, from the working-style self-check on the home page.

Architect-leaning attendees write the fact down beforehand, in order, with the supporting detail underneath, and then wait for the agenda item it belongs to. The item gets cut for time. Preparing the sentence is the right instinct, and this site has a page on writing your contribution before the meeting; the change is to spend it in the first open minute instead of at the point in the plan where it fits.

Sprinters say it inside the first thirty seconds, which is the right timing, and say it as a bare fact with no decision attached. The room files it as a detail. Add the clause about what it changes and your same sentence stops being trivia.

Visionary-leaning attendees lead with the implication, so the room hears a position on the launch date and never gets the contract date underneath it. A position invites disagreement. A date nobody else has invites a question, which is the response you want.

Improvisers wait for the opening, read the room accurately, and do their best work live. The 2026 study is the problem with that: five minutes of discussion before you spoke was enough to change how a group handled the identical piece of information. The opening you are waiting for is the thing that closes.

Waiting is the one mistake all four share.

So decide your sentence now, before the next meeting where you hold something the others do not. Say it in the first pass, name that it is not already in the room, name the decision it lands on, and put the same line in the shared document before anyone leaves. If nothing you hold bears on the decision at all, that is the meetings pillar question rather than this page.

Frequently asked questions

How do you get your point heard in a meeting?

Say it early, name that the others do not have it, and name the decision it changes. In a 2026 experiment with 370 three-person groups, teams decided worse when the newest member held the critical fact. An original member holding the same fact did better. False feedback presenting the newcomer as skilled did not fix that. So credibility first, fact second, is not what the evidence supports.

Why does a group ignore information only one person has?

Groups discuss what most members already know far more than what one member holds. A study of 30 three-person security teams measured what they discussed. Incidents known to the majority came up three times as often as the ones a single analyst held. Unaided teams were inefficient at linking those single-holder pieces. That was a synthetic task environment, not a work meeting, so carry the direction and not the setting.

When in a meeting should you say the thing nobody else knows?

In the first pass, before the room converges on a position. Two people talked for five minutes before a third joined with the deciding fact. Those groups were less likely to decide correctly than groups where an original member held it. Once a position exists, your fact has to unseat it rather than inform it.

Should you put your point in writing instead of saying it in the meeting?

Say it and write it, in the place the decision is being made. Anything only spoken has to survive everyone else in the room remembering it. In the security-team study, the aids that reduced the bias kept uniquely held information visible rather than recalled. A transcript is not that. It records that you spoke, which helps after the meeting and not during it.