How the Best Leaders Shape Conversations
Alison Wood Brooks, Andy Atkins · article · 2026-08-17 source ↗ Reader 🖥️
Because every organizational outcome — strategy, decisions, innovation, relationships — originates in conversation, leaders should treat conversation as a skill to be deliberately designed and stewarded, by making a group's goals, roles, and psychological 'soul' explicit, rather than leaving conversational quality to chance.
practical · mixed — Primarily practical/argumentative — it argues leaders ought to deliberately steward team conversation via a named framework — with expository definition of terms and three narrative case anecdotes used as evidence-by-example, closing in an explicitly prescriptive how-to section.
Structure
| Part | Role |
|---|---|
| Opening problem statement: meetings that meander, leave concerns unspoken, or revisit the same decisions. | context |
| Reframe: the real object of leadership attention isn't meeting logistics (agendas, calendars) but conversation itself, which the authors name 'conversational stewardship.' | thesis statement |
| Survey evidence: 1,000+ participants, 100+ teams, ~70 organizations; how teams use time predicts effectiveness more than how often they meet. | evidence |
| 'Rethinking Team Dialogue' section: distinguishes one-on-one conversational skill from the added coordination demands of group conversation. | context / problem elaboration |
| Goals section (purpose, products, process) with the Fortune Global 25 anecdote. | evidence / framework part 1 |
| Roles section (facilitator, decision-maker, participant expectations) with the acquisition-bid chairman anecdote. | evidence / framework part 2 |
| Soul section (relational vs. transactional trust, psychological safety) with the failed-ISP anecdote. | evidence / framework part 3 |
| 'Steward Your Next Conversation' closing section: three concrete practices (state the goals, clarify the roles, protect the soul). | conclusion / call to action |
Key terms
- Conversational stewardship Conversational Stewardship
The leader's practice of intentionally caring for the quality of a group's dialogue — creating clarity about why the group is meeting, who is responsible for what, and how to make honest contribution feel safe. - Goals (as a stewardship element)
A composite of three sub-questions the leader must answer before a conversation starts: purpose (why are we here), products (what should we walk away with), and process (how will we have this conversation). - Roles (as a stewardship element)
Explicit assignment of two specific functions — facilitator (steward of how the conversation unfolds) and decision-maker (owns the final call) — plus clarified participant expectations, deliberately decoupled from organizational hierarchy or title. - Soul (as a stewardship element)
The felt psychological quality of a conversation — whether people believe they can say what needs to be said. Decomposed into relational trust (benevolence-based, built dyadically) and psychological safety (a group-level shared belief, following Edmondson, that candor won't be punished). - Relational trust vs. transactional trust
Two distinct forms of interpersonal trust: relational (benevolence-based — confidence others care about and respect you) and transactional (competence-based — confidence others are capable, reliable, aligned in intent).
Propositions
- explicitConversation, not technology, is the operating system of collaboration — every strategy, decision, innovation, and relationship begins as a conversation, and unproductive conversations cost organizations in wasted time, poor execution, and missed opportunities.conf: high
- explicitHow teams use their meeting time, not how often they meet, predicts effectiveness; teams that consistently leave conversations with clear accountabilities and next steps are roughly four times more likely to be rated effective, per a survey of 1,000+ participants across 100+ teams in nearly 70 organizations.conf: highextendsAmy Gibson's Post · p1 — Gibson claims meeting effectiveness is about 'what people walk away with,' not agenda/room/tech/timing, but never specifies what that should be; this article's survey supplies the specific, quantified content (clear accountabilities and next steps, ~4x effectiveness) that fills in Gibson's vaguer claim.supports5 Things High-Performing Teams Do Differently · p3 — An independent survey converges on the same finding: structural meeting practices (that piece's prework/agenda/check-in; this piece's explicit next-steps/accountabilities) predict team-effectiveness ratings.
- explicitEffective team conversation results from jointly stewarding three elements: goals (why/what/how), roles (who does what), and soul (whether people feel safe to speak candidly).conf: high
- explicitMaking the facilitator and decision-maker roles explicit — independent of organizational hierarchy — prevents conversations from drifting toward decisions no one has authority to make, or from being derailed by an undisclosed veto.conf: high
- explicitPsychological safety is a shared, group-level belief distinct from interpersonal trust, which is built one relationship at a time; a team can have strong one-on-one trust among its members while still lacking the group-level safety needed for candor.conf: highsupportsHow to Build a Superteam That Keeps Getting Better · p4 — Both pieces treat psychological safety (via Edmondson / the 2012 Google study) as a distinct, central driver of team performance, not reducible to individual relationship quality.supportsPresentation: The Human Scalability Problem: Why Your Teams Don't Scale Like Your Code · p4 — Near-identical claim structure: both state that psychological safety requires interpersonal trust as a foundation but is a distinct, further, group-level property, not reducible to trust itself.
- explicitConsequential truths that go unvoiced because a team lacks psychological safety can be catastrophic, not merely uncomfortable — the most consequential leadership failures typically trace to important truths that were never said aloud.conf: high
- explicitThe ability to intentionally shape group conversation is a learnable skill, not a fixed trait or a natural byproduct of group size or culture.conf: high
- implicitThe article treats participants' own self-rating of their team's effectiveness as a valid stand-in for the team's actual performance, without addressing that the same people who report the conversational behaviors also produce the effectiveness rating.conf: medium
- implicitThe three-part goals/roles/soul decomposition is treated as a sufficient account of what makes team conversation effective, without considering whether other factors (group size, physical vs. remote setting, cultural context, power distance) are independently necessary.conf: medium
Arguments
Leaders should deliberately steward team conversations — via explicit goals, roles, and soul — rather than leave conversational quality to chance. strength: medium
Explicit premises:
- Every organizational outcome (strategy, decisions, innovation, relationships) originates in conversation.
- Survey data show specific conversational practices (e.g., leaving meetings with clear accountabilities/next steps) correlate strongly with participants' effectiveness ratings.
- Conversational stewardship is a learnable skill.
Implicit premises:
- Self-reported team-effectiveness ratings validly proxy real organizational performance.
- The observed correlation reflects stewardship practices causing effectiveness, rather than effective teams naturally developing better conversational habits, or a shared upstream cause (e.g., strong leadership) producing both.
- The goals/roles/soul decomposition is an exhaustive or at least sufficient account of what conversational quality consists of.
Inferences:
- If stewardship is learnable and commonly absent, most organizations are leaving a large, correctable performance gap on the table.
- Formal authority/title should matter less to who plays which conversational role than does explicit, situational assignment.
Building interpersonal trust among team members is not sufficient to produce a psychologically safe team; group-level safety requires its own, distinct cultivation.
Explicit premises:
- Trust develops relationship by relationship, in two forms (relational and transactional).
- Psychological safety is a property of the group as a whole, not of any one dyad within it.
- The authors report having observed teams with strong individual trust but weak group candor.
Implicit premises:
- Individual-level trust does not aggregate additively into group-level psychological safety — the mechanisms are qualitatively different, not just quantitatively separable.
Inferences:
- Leadership development that only builds one-on-one rapport (mentoring, skip-level check-ins) may leave group meetings just as guarded as before, and needs a distinct, group-facing intervention.
Narratives
- A Fortune Global 25 leadership team was stalled for months in a standoff between functional and client-facing units, each frustrated the other didn't grasp its priorities. When Andy Atkins refocused the conversation's goal on what would best serve the whole enterprise, the team reached a novel solution within 90 minutes.
→ Supports the 'Goals' pillar: shows that reframing a conversation's stated purpose — not new information or more time — is what unblocks a stalled group decision. - A U.S. firm's senior executive team, deciding whether to solicit an acquisition bid rather than risk a hostile offer, held a multi-day meeting in which the chairman deliberately stayed mostly silent to avoid swaying the group, and made explicit up front that the final decision required every member's agreement.
→ Supports the 'Roles' pillar: shows that explicitly clarifying who decides, how, and when senior voices restrain themselves changes the quality of group deliberation. - A global internet service provider with a commanding market position appeared to be thriving from the outside, but its senior leadership team never surfaced known forecasts showing expenses would exceed revenues indefinitely. The silence persisted, and the business failed within two years.
→ Supports the 'Soul' pillar: shows the cost of a team that lacks the psychological safety to voice a known, consequential truth, even among senior leaders.
Questions addressed
- ✅ How can leaders transform team conversations from sources of frustration into engines of collaboration?
By practicing conversational stewardship: making goals (purpose/products/process), roles (facilitator/decision-maker/participant expectations), and soul (trust/psychological safety) explicit, operationalized into three closing practices. - ⭕ Does the goals/roles/soul framework generalize beyond small, senior, in-person leadership-team settings to large, cross-functional, remote, or lower-stakes meetings?
Gaps
- incompleteThe headline effect-size claims (e.g., 'roughly four times more likely to be rated as effective') come from the authors' own proprietary practitioner survey, in which the same participants who report their team's conversational practices also self-rate the team's effectiveness — the same common-method / self-report concern already on record in this corpus against a structurally similar HBR piece.
Read next: Peer-reviewed team-effectiveness research using outcome-anchored (not self-rated) performance measures, to test whether the goals/roles/soul practices predict effectiveness independent of a self-report halo. - incompleteAll three supporting anecdotes come from Andy Atkins's own consulting engagements with senior/executive leadership teams (a Fortune Global 25 company, a firm's chairman weighing an acquisition bid, an ISP's leadership). No evidence is given that goals/roles/soul stewardship works the same way in frontline, large cross-functional, or asynchronous/remote conversations.
Read next: Look for treatments of conversational or meeting design specifically tested at scale (large all-hands, distributed teams) rather than in small executive rooms. - incompleteThe article establishes that psychological safety and interpersonal trust are distinct and that having one doesn't guarantee the other, but never explains how a leader who has built neither should sequence the work — trust first, safety first, or both in parallel — despite this being exactly the practical question its own distinction raises.
Read next: Look for research or frameworks that address the build sequence or interaction between dyadic trust-building and group-level psychological-safety interventions.
Closes gaps elsewhere
- Amy Gibson's Post · g1 — Gibson's meeting-cadence prescriptions were assertion-only with no evidence, data, or named case. This article supplies the research-grounded treatment that gap called for — and gives a direct, negative answer on cadence specifically: a survey of 1,000+ participants across 100+ teams in ~70 organizations found meeting frequency does NOT predict effectiveness, only how the time is used (quantified via the next-steps/accountabilities finding, ~4x effect). Note this document's own g1 flags that the underlying survey still shares Gibson's self-report limitation, so the closure resolves the evidentiary gap but not the methodological one.
Related documents
- 5 Things High-Performing Teams Do Differently
- AI Enthusiasts Are in a Race Against Time, AI Skeptics Are in a Race Against Entropy
- Amy Gibson's Post
- Anders Aukland's Post
- A Leading Restaurateur on What It Takes to Scale a Winning Culture
- Every Organization Has Two Realities
- Frontline Workers Know How to Solve Your Organization's Biggest Problems
- The Hidden Realities of AI Adoption
- Our Favorite Management Tips on Giving Feedback
- When Persuasion Is (and Isn't) Manipulation