Key Takeaways
Scale conversations by helping everyone consider ideas, not merely submit them.
Count resonance, not repetition. At a table, people may initially order beer because that is what someone else ordered. When one person mentions a margarita, the group can reconsider and choose it instead. The important idea was not the first or most repeated one. MacLeod calls such late-emerging, influential suggestions “margarita thoughts.” Most large-group methods miss them by counting responses or clustering common words rather than letting people consider one another’s ideas.
Make the exchange possible. Online, people can contribute across distance and time, then evaluate others’ thoughts. In Connecticut, more than 40,000 people joined a school-safety conversation over a weekend, giving state leaders a much wider view of concerns than a small meeting could provide. Scale matters when it helps overlooked ideas reach decision-makers.
What’s striking is the distinction between collecting opinions and creating conditions for opinions to change. The margarita example captures deliberation: people learn from one another, rather than simply register a preference. This is closer to a conversation than a poll. Yet a high rating is still a proxy for importance, not proof of truth or feasibility. Strong processes should reveal both what resonates and where disagreement remains. Participation also needs scrutiny: a large response can still omit people without reliable access, time, or trust in the process.
Build relationship, process, and ownership capital before asking people to change.
Earn trust in three ways. MacLeod uses “capital” to mean assets leaders build and can draw on when decisions get difficult. Relationship capital grows when leaders listen without approaching people only to get something. Process capital comes from involving people affected by a decision and making the decision process understandable. Ownership capital grows when people have agency in matters that affect their work, resources, or future.
Ask for advice, not allegiance. A leader planning a budget might ask what resources people need, then explain how their ideas informed the choice. The aim is not to promise that every suggestion will be adopted. It is to show that people’s knowledge mattered. MacLeod argues that this kind of inclusion helps unify people’s existing effort with organizational direction, rather than treating them as disengaged people who need to be motivated.
This framework connects participation to procedural justice: people often judge a decision partly by whether they were treated respectfully and could understand how it was made. That can preserve trust even when the outcome disappoints. The distinction between voice and control is crucial. A request for advice is not a referendum, and leaders should say so plainly. Otherwise, participation can become tokenistic: people speak, leaders retain authority, but nobody knows what influence the input had. The process earns credibility when its limits and consequences are visible.
Make room for every voice, then close with shared understanding.
Structure the exchange. A useful conversation begins with clear intent, gives people room to speak, and makes listening visible. In a small group, that may mean arranging chairs in a circle and giving each person a turn. MacLeod describes doing this during a tense meeting about compensation: the circle changed the room from a leader addressing an audience into colleagues hearing one another.
Close without forcing agreement. Listening includes trying to understand what matters to someone, even when their experience is unfamiliar. A conversation also needs a summary of what people understood, what remains unresolved, and whether any action was agreed. Closure does not require a decision. The group may simply have a clearer picture of its disagreements, which is more useful than a list of promises nobody intends to keep.
The circle is more than a visual symbol. It changes turn-taking and makes unequal access to the floor easier to notice. This resembles good facilitation in deliberative settings, where silence, sequencing, and summaries are part of the work rather than pauses around the real work. Still, equal speaking time does not guarantee equal influence. Status can shape whose words are taken seriously even after everyone has spoken. Leaders need to attend not just to who contributed, but to how the group responded to each contribution.
Give ideas equal, anonymous consideration before judging their merit.
Protect the idea from the speaker’s status. MacLeod’s early activity, called “35,” invited participants to write an idea anonymously, pass ideas around, and rate them. In the original version, each idea received five ratings on a seven-point scale, producing a maximum score of 35. The author later recommends a simpler one-to-five agreement rating. The point is not the arithmetic. It is that every person can contribute, and every idea gets considered.
Avoid premature sorting. If a facilitator groups comments before people evaluate them, the facilitator’s judgments shape what participants see. MacLeod’s online method instead presents thoughts individually and aims to give each an equal chance to be rated. Anonymity can increase candor: his small company saw participation nearly double when names were removed from contributions and ratings.
The design addresses two different problems: social risk and cognitive overload. Anonymity can help people focus on an idea rather than the author, while one-at-a-time evaluation is easier than comparing a wall of comments. But anonymity is not a complete cure for bias, and equal exposure is not the same as equal understanding. A fair process also needs moderation, accessible language, and clear rules for interpreting ratings. The strongest contribution here is treating the evaluation method as part of the conversation, not as neutral plumbing.
Treat surveys as measuring instruments, not substitutes for discovery.
A number needs context. An employee gave a very low score to a question about recommending their workplace, then explained they thought it asked whether their friends were suitable job candidates. The rating said little about their satisfaction. Closed-choice surveys can constrain answers, while open-ended boxes often leave participants unsure whether anyone will read or act on their words.
Beware false representation. A school district’s calendar poll showed a 55-to-45 split. Leaders treated it as research, but some parents interpreted the result as a vote and protested when the less popular option was chosen. Focus groups can also overstate how well a small, self-selected group represents everyone. Town halls may feel staged when leaders mostly present information and manage questions. These methods can still help, but none automatically creates a conversation.
The central distinction is between measurement and inquiry. Surveys work well when the question has a defined answer, such as benchmarking a rate. They are weaker when leaders do not yet know what concerns or explanations matter. A response percentage also needs a denominator and a description of who did not participate. The calendar example shows that the perceived promise of a method matters as much as the leader’s intention. If participants think they are voting, calling the exercise consultation afterward will not restore trust.
Start with people’s immediate needs before inviting them to envision change.
Meet the present first. The book’s question hierarchy moves from immediate safety and practical conditions toward connection, improvement, shared culture, a common vision, and radical change. It is a guide to sequencing, not a script. If people fear losing their jobs, a prompt about an ambitious future may feel detached. If a team is still struggling with basic systems, asking it to celebrate its culture can miss the problem.
Use the right rung for the issue. During the pandemic, leaders first needed to hear fears about health and employment, then learn what people needed to work or study, and only then explore how to move forward together. The sequence is not strictly linear: one person or organization can face a safety concern on one topic while imagining the future on another. The practical test is whether the question fits people’s current circumstances.
This is a useful antidote to leaders’ preference for aspirational language. A polished vision cannot substitute for basic clarity or security. The framework also resembles needs-based models of motivation, but it should be used as a diagnostic, not a rigid ladder. People do not all move through needs in the same order, and practical safety can remain unsettled even amid success elsewhere. Asking a check-in question, then taking the answer seriously, is more trustworthy than assuming a whole organization occupies one shared emotional state.
Ask candid questions that invite advice without disguising a vote.
Find the real question. The Five Whys technique means repeatedly asking why a problem matters until the underlying concern becomes clearer. When executives wanted to understand employee departures, their real question was how to keep good people from leaving. A direct request for advice about retention was more honest than a broad questionnaire about happiness and workplace conditions.
Design for the next step. At scale, each response should express one idea that others can assess individually. A question asking for three priorities creates bundles that are hard to rate. Wording also sets expectations: asking what leaders “should” do can sound like a promise to follow the group’s choice. Asking what options might help invites perspectives while leaving decision authority clear. MacLeod calls the change rationale especially important: explain why a change matters far more fully than what is changing.
Question wording is a form of institutional design. It establishes who is expected to decide, what kinds of answers count, and whether participants are advising or choosing. The author’s emphasis on candor is valuable because evasive prompts often conceal the very problem people already discuss privately. One nuance: sometimes a genuine vote is appropriate, and the word should is not inherently misleading. The ethical requirement is alignment between the question, the decision rule, and the authority participants actually hold.
Treat uncomfortable feedback as information before treating it as an attack.
Notice the defensive reflex. MacLeod recounts asking colleagues whether the company should hire a senior people leader. Public responses were supportive, but an anonymous exchange surfaced concerns about fit, budget, and timing. His first reaction was irritation. After pausing, he recognized that colleagues lacked the context he had and that their questions were reasonable.
Show what listening changed. A leader can assemble a diverse group to help interpret and share results, acknowledge what was learned, and distinguish understanding from agreement. Some suggestions may lead to immediate action, others to investigation, and some to no change. The important step is to explain that choice. Leaders can also learn from participants’ language: in one school district, calling a planned new building a renovation obscured what the community actually supported.
The distinction between hearing criticism and being condemned by it is psychologically important. A pause creates room for reflection before a defensive response hardens into policy. This is also a form of feedback literacy: information becomes useful when the receiver can separate its content from the emotional discomfort of receiving it. Yet leaders should not frame every objection as a personal growth opportunity. Some criticism identifies a genuine power imbalance or harm, which requires more than composure. Demonstrated learning means changing an interpretation or decision when the evidence warrants it.
Make diversity productive by changing how ideas are heard and weighed.
Difference needs inclusion. MacLeod argues that recruiting varied people is not enough if their ideas are ignored or filtered through bias. He cites a crowd experiment in which about 17,000 people estimated a cow’s weight: the diverse crowd’s estimate was closer than the estimate from self-described experts alone. The broader lesson is not that crowds always outperform expertise, but that varied judgments can reduce shared blind spots when they are aggregated well.
Look for barriers in ordinary rules. In one school district, students needed access to a car to join job placements. That requirement excluded students whose families had less access to wealth, including many Black families affected by past discrimination. Equity requires examining who can participate and how policies affect them. Inclusion also means giving people safe ways to raise concerns and ensuring those contributions receive serious consideration.
The strongest claim is conditional: diversity can broaden a group’s information, but only if interaction does not erase or punish difference. This connects to the distinction between representation and participation. A room can look diverse while its decisions remain unchanged. The cow example is memorable, but a single crowd estimate cannot establish that diverse groups are wiser in every domain; expertise, task structure, and aggregation all matter. For organizations, the more actionable test is whether different perspectives alter the questions asked, the evidence considered, or the barriers built into policy.
Ask about wellbeing before assuming people are ready to perform.
Make concerns speakable. After his father died unexpectedly, MacLeod experienced a panic attack and arrived at a clinic fearing a heart attack. A doctor helped him understand what was happening. Later, when he shared the experience, other people disclosed similar episodes they had kept private. The story shows how shame and isolation can conceal needs that affect people at work and beyond.
Offer care without pretending to provide therapy. Leaders can ask what is weighing on people and what support would help them do their work. During the pandemic, a company asked parents about the challenges of school closures, then invited suggestions for workplace support. Participants said the conversation helped them feel less alone. The purpose is not to diagnose anyone or demand personal disclosure. It is to make support and practical needs discussable, with appropriate privacy and professional help when needed.
The practical contribution is to treat wellbeing as relevant to collective work without turning managers into clinicians. A psychologically safer environment can make it easier to name strain before it becomes a crisis, while a specific invitation to disclose can also feel intrusive. Participation must remain voluntary, and leaders should be clear about confidentiality and available support. The balance is care without surveillance: ask what would help, listen without demanding a personal history, and avoid promising solutions the organization cannot provide.
Analysis
Scaling Conversations makes a strong case that participation is not a headcount. A process becomes meaningfully inclusive only when people can contribute independently, encounter the contributions of others, and see how their input was interpreted. The book’s central example, the margarita that beats the more frequent beer order, gives this distinction memorable form. Its practical value is especially high where leaders rely on surveys, town halls, or self-selected representatives to understand people affected by decisions.
The method also exposes a tension: structured ratings can reduce the influence of status and first-mover advantage, but they cannot make a process fully unbiased. The prompt frames what can be said. The platform determines whose contribution is visible. Access, time, literacy, trust, and willingness to participate shape the resulting sample. A large number of responses can create confidence without establishing that the participants represent everyone. Leaders should therefore disclose participation limits and interpret ratings as evidence about contributors, not as a definitive measure of an entire population.
The book is most persuasive when it treats listening as both a relationship and a design problem. Anonymity, equal exposure, and clear decision boundaries are not interchangeable, but together they can help people take part without being forced into public confrontation. The distinction between voice and vote is essential: asking for advice does not transfer authority, and pretending otherwise can produce the very distrust consultation is meant to prevent.
A further obligation follows from the method. Leaders need to explain what they learned, what changed, what did not, and why. Without that return path, scaled input risks becoming another collection exercise. With it, the process can create a more informed decision, surface overlooked concerns, and help participants understand the limits of collective agreement. The strongest measure of success is not unanimity, but whether more relevant knowledge reaches the decision and the reasoning becomes more accountable.
Glossary
Margarita thoughts
Late ideas that resonateThe book’s metaphor for ideas that may appear late or only once but become important after people consider one another’s perspectives. It contrasts these ideas with frequent first responses, which can dominate surveys and word counts even when they are not what the group ultimately values.
35
Anonymous idea-rating activityA face-to-face activity for finding what matters to a group. Participants anonymously write an idea, pass ideas around, and rate them. In the version described, each idea is rated five times on a seven-point scale, giving it a possible total of 35. The author also recommends a simpler one-to-five agreement scale.
Relationship capital
Trust built through relationshipsOne of three forms of leadership capital in the book. It develops when leaders build genuine relationships by listening, showing interest in people’s concerns, and offering support without approaching every interaction with an immediate request. This trust can help people work together during difficult decisions.
Process capital
Trust in fair decisionsThe confidence people gain when they understand how decisions are made and have a meaningful opportunity to contribute before decisions affect them. The book argues that a fair process can preserve trust even when people dislike the result, provided leaders explain how input informed the decision.
Ownership capital
Agency in shared decisionsThe sense of agency people gain when invited to advise on matters that affect their work, resources, or organization. MacLeod distinguishes this from trying to engage supposedly disengaged people. The aim is to bring people’s existing knowledge and effort into closer alignment with shared goals.
Scaled conversations hierarchy
Sequence for matching questionsA guide for choosing questions that fit people’s current circumstances. It moves from immediate safety and practical structures through connection, improvement, culture, and shared vision to radical transformation. The author cautions that people and organizations may occupy different levels on different issues at the same time.
Representational dissemination group
Group that interprets and shares findingsA group that helps leaders understand and communicate results from a large conversation. Unlike a representative committee, it is not expected to speak on behalf of everyone. Its role is to bring varied perspectives to interpretation and help make the findings accessible to the people affected.
Five Whys
Repeated questions to find causesA questioning technique that repeatedly asks why a problem matters in order to move from a surface request toward its underlying cause. In the book, executives concerned about employee departures use this approach to reach a more candid question about retaining staff and learning why people leave.
Golden ratio of change
Explain why more than whatMacLeod’s rule of thumb for communicating significant change: leaders should explain the reasons for a change far more fully than the description of what will change. The purpose is to help people understand the need and context, especially when the change is unfamiliar or disruptive.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.