Meet each side alone and carry proposals between them

When joint committee calls stall, meet each person alone, ask what they will not say in the room, carry the offers between them, and bring the group back only to confirm.

Evidence: documented practice. Sellers and trainers publish how they do it. No study has tested it.

What it is.

The joint committee calls go nowhere. Finance says one thing in the room, IT another, and nobody moves while their colleagues are watching.

Stop running the deal through the joint call for a while. See each person on their own, ask what they will not say with their colleagues listening, and carry the offers, the concessions and the exact wording between the rooms yourself. Bring the group back together when there is something to confirm. Mediators call the private meeting a caucus. The popular name, shuttle diplomacy, comes from reporters following Henry Kissinger's short flights between Middle East capitals in 1974, as the US State Department's historian records.

If two groups inside the buyer have stopped moving towards each other, mediation has a last-resort version: one written package, sent privately to both, with a deadline and two permitted answers, accept or reject, and a promise not to reveal either answer unless both say yes. Keep in mind that a mediator is a neutral both sides hired, and you are not.

What it looks like.

The sales version is a pre-call, not a replacement. On episode 453 of 30 Minutes to President's Club, David Rosenstein, a senior account executive at LinkedIn, prepares a large meeting this way: divide the room strategically, prep with your champion, "pre-call each stakeholder", then personalise the questions in the meeting itself. The joint meeting still happens.

The mediator's version. Gregorio Billikopf, writing in Group Facilitation in 2002, describes meeting each party separately before bringing them together at all. His test of whether a joint session is safe is to ask one party what they value in the other, and not to treat the first negative answer as final. He publishes no script.

A written proposal, word for word. The wording of Professor Eric Green's proposal in a 2012 dispute is public because the parties later fought over whether it bound them, and the Ninth Circuit's opinion in Sony v. HannStar reproduces the emails. He proposed settling for $4.1 million, paid on 30 March, and asked each side to tell him privately, by 5pm on 27 March, whether it would ACCEPT or REJECT. The answers were double-blind: a side that did not accept would never learn what the other had said. The next day he reminded both that no negotiation was permitted. Josh Fruchter's account for Merge Mediation draws the lesson from what followed: say in the document that it is binding.

What to try before it. Stephen Hochman tells mediators to exhaust the conditional offer first, asking one side in private whether, if the other side reached a stated figure, they would move to a figure of their own, and to keep the proposal for the end, when everything else has failed.

No sales source publishes a seller carrying offers between the rooms of one buyer, or tabling one package to two departments on accept-or-reject terms.

Where it has been tested.

In B2B sales

Nobody has measured it. No study compares sellers who broke a stalled committee into separate meetings with sellers who kept the joint call.

The one B2B finding is a warning. Gartner's 2025 survey of 632 B2B buyers found unhealthy conflict in 74% of buying groups, and that content aimed at the whole group raised consensus while content aimed at individuals cut it sharply; the figures are under After a group call, write to each person separately. That measures what the seller sends, not how the meetings are arranged, but it cautions against letting separate conversations become the whole process. Gartner sells research and advice to sales organisations.

In other disciplines

Meeting apart changes what people say, not whether they agree. Gary Welton, Dean Pruitt and Neil McGillicuddy analysed 51 community mediations in 1988. In private meetings, people were less hostile, gave more information and proposed more new options, and also talked up their own side and ran the other down more. The same group then studied 73 cases in 1992 and found no link between caucusing and whether agreement was reached or how good it was. Roselle Wissler and Art Hinshaw surveyed more than 1,000 mediators in eight US states in 2024 and found few differences in final outcomes between cases that started apart and cases that started together, once the amount of real discussion was taken into account. Jill Tanz and colleagues added an early private meeting to a community mediation programme in 2023; people felt better understood, and the study does not report whether more cases settled.

Nobody concedes in front of their own side. Peter Carnevale, Dean Pruitt and Scott Britton found in 1979 that negotiators watched by the people they represented turned to threats and hard positions, which left everyone worse off. A committee member watched by colleagues is a different relationship, and nobody has tested it.

The proposal, from inside mediation. Daniel Klerman and Lisa Klerman analysed one mediator's notes on more than 400 employment cases: over 94% settled, a mediator's proposal was used in almost 90% of them, and over 99% of those settled. One mediator, no comparison group, and the proposal is by design the last thing tried. Debra Berman's 2022 survey of 167 mediators found 84% had used one and remembered an average acceptance rate of 73%, and most urged keeping it for a last resort.

The objections. Judge Wynne Carvill, writing for JAMS in 2020, says that when parties came to expect his proposals they stopped negotiating with each other and spent the session trying to influence his number. Eric Galton, Lela Love and Jerry Weiss argue in The Decline of Dialogue that the drift to separate-rooms-only mediation is a loss.

Caveat.

Everything measured comes from mediation, where the person running the separate meetings is a neutral with no stake in the result, while a seller who does it wants one particular answer. The closest direct test found that starting apart made little difference to final outcomes, and nobody has compared a double-blind proposal with the same package put openly to both sides.

Takeaway.

Use separate meetings to find the blockage, not to close the deal: see each person alone, ask what they will not say in the room, then bring the group back together for the decision. If two groups inside the buyer are stuck, a single written package with a deadline and two permitted answers is the mediator's last resort, and it has never been run by a party with a stake in the outcome, which is what you would be.

Sources.

Recommended by

US Department of State, Office of the Historian diplomacy · Shuttle Diplomacy and the Arab-Israeli Dispute, 1974-1975 (Milestones in the History of U.S. Foreign Relations) · official document
The State Department's historian records Kissinger's shuttle diplomacy between Middle East capitals in 1974 and 1975. It is where the popular name for the card's move comes from.
30 Minutes to President's Club (David Rosenstein) · #453 - How to Nail Large Sales Meetings | David Rosenstein · 2025 · podcast
Rosenstein prepares a large meeting by pre-calling each stakeholder and prepping with the champion before the group meets. It is the nearest published sales version, a step before the joint meeting, not a replacement for it.
Gregorio Billikopf mediation · Contributions of Caucusing and Pre-Caucusing to Mediation (Group Facilitation, Spring 2002, reprinted on Mediate.com) · 2002 · post
Billikopf makes the practitioner case for meeting each party alone before bringing them together, and tests whether a joint session is safe by asking what each values in the other. It is the mediator's version of the card's separate meetings, with no script and no data.
Josh Fruchter law · The Mediator's Proposal: If Both Sides Say Yes, Make Sure It's Binding · 2019 · post
Fruchter walks through the same mediator's proposal and the fight that followed over whether it bound the parties. It is the source of the card's rule to say in the document that it is binding.
Stephen A. Hochman (Mediate.com) mediation · A Mediator's Proposal - Whether, When, and How It Should Be Used · 2016 · post
Hochman tells mediators to try the conditional offer first and keep a single proposal for the end, when everything else has failed. It is the order the card follows for the written package.

From other disciplines

United States Court of Appeals for the Ninth Circuit law · Sony Electronics, Inc. v. HannStar Display Corp., 835 F.3d 1155 (9th Cir. 2016) · 2016 · official document
The court's opinion reproduces a mediator's proposal word for word: one figure, a private answer by a deadline, accept or reject, and answers kept secret unless both accepted. It is the exact wording behind the card's last-resort package.
Gary L. Welton, Dean G. Pruitt and Neil B. McGillicuddy social psychology · The Role of Caucusing in Community Mediation · 1988 · paper
In community mediations, people meeting the mediator privately were less hostile, gave more information and proposed more new options, and also talked up their own side more. It shows what the card's private meetings change: what people say.
Roselle Wissler and Art Hinshaw empirical legal studies · Comparing Joint Session and Caucus Outcomes: Factoring in Substantive Discussions and Case Characteristics, 25 Cardozo Journal of Conflict Resolution 491 (2024) · 2024 · paper
A survey of more than 1,000 mediators found few differences in final outcomes between cases that started apart and cases that started together, once the authors allowed for the amount of real discussion. It is the closest direct test, and it says starting apart makes little difference on its own.
Jill S. Tanz, Martha K. McClintock and Rae Kyritsi conflict resolution · Can early caucus improve a community mediation model? Counteracting stressors prior to joint session · 2023 · paper
Adding an early private meeting to a community mediation programme left people feeling better understood, and the study does not report whether more cases settled. It supports the card's first step, seeing each person alone, on perceptions only.
Peter J. D. Carnevale, Dean G. Pruitt and Scott D. Britton social psychology · Looking Tough: The Negotiator Under Constituent Surveillance · 1979 · paper
Negotiators watched by the people they represented turned to threats and hard positions, which left everyone worse off. It is the mechanism behind the card's point that nobody concedes in front of their own side, though nobody has tested a committee member watched by colleagues.
Daniel Klerman and Lisa Klerman empirical legal studies · Inside the Caucus: An Empirical Analysis of Mediation from Within · 2015 · paper
One mediator's notes on employment cases show a proposal used in almost all of them, and almost every case where the mediator used it settled. It shows how often the last-resort package works in practice, from one mediator, with no comparison group.
Debra Berman dispute resolution · The Use of Mediator Proposals in Practice: What the Data Tell Us · 2022 · paper
Most mediators surveyed had used a mediator's proposal, remembered the parties accepting most of them, and urged keeping it for a last resort. It describes how the card's written package is usually built, from the mediators' own reports.

Who says do not

Gartner vendor · Gartner Sales Survey Finds 74% of B2B Buyer Teams Demonstrate Unhealthy Conflict During The Decision Process · 2025 · vendor data
Gartner found that content aimed at the whole buying group raised consensus, while content aimed at individuals cut it sharply. It is the warning against letting separate conversations become the whole process, though it measures what the seller sends, not how the seller arranges the meetings.
Gary L. Welton, Dean G. Pruitt, Neil B. McGillicuddy, Carol A. Ippolito and Jo M. Zubek social psychology · Antecedents and Characteristics of Caucusing in Community Mediation · 1992 · paper
The same group's follow-up found no link between meeting apart and whether the parties reached agreement or how good it was. It is the most important negative finding for this card: separate meetings change behaviour, not the result.
Wynne S. Carvill mediation · The Danger of Mediator's Proposals · 2020 · post
A retired judge explains why he stopped using mediator's proposals: once parties expected one, they stopped negotiating with each other and tried to influence his number. It is the warning that the package can replace the negotiation it was meant to finish.
Eric Galton, Lela Love and Jerry Weiss mediation · The Decline of Dialogue: The Rise of Caucus-Only Mediation And the Disappearance of the Joint Session · 2021 · paper
Three experienced mediators argue that running everything in separate rooms, with no joint session, is a loss. It is the standing objection to letting the card's separate meetings become the whole process.
Only the opening paragraph is free to read.