Article
Service DesignWorld Cafe: Method, Process, and Practical Guide for Large Groups
The World Cafe method step by step: process, 7 design principles, and practical guide for 20-200 participants.
The World Cafe is a structured dialogue method for large groups (20 to 200+ participants) that activates collective intelligence through rotating table conversations. In multiple rounds, small groups discuss one or more guiding questions at cafe-style tables. After each round, participants switch tables — only the table host stays and connects the ideas of the previous group with the new one. The method was discovered in 1995 by Juanita Brown and David Isaacs and systematically described in 2005 in The World Cafe: Shaping Our Futures Through Conversations That Matter [1].
What distinguishes the World Cafe from other large-group formats: it combines the intimacy of a table conversation with the reach of a large event. While classic conference formats produce a room full of passive listeners, the World Cafe activates every single participant — and generates through rotation a network of ideas that no plenary discussion could produce.
Search the web for “World Cafe method” and you will find dozens of superficial descriptions: tables, paper tablecloths, markers, rotate. No result explains why the quality of guiding questions determines success or failure. None shows the difference between a successful “harvest” and a chaotic plenary. And none systematically compares the World Cafe with other large-group formats like Open Space, Fish Bowl, or Design Sprint.
This guide closes those gaps — with the 7 design principles, a complete process, a service design example, and an honest analysis of the most common mistakes.
Where the method comes from: A discovery at the kitchen table
The origin story of the World Cafe is itself an example of emergent innovation. In 1995, Juanita Brown and David Isaacs invited a group of executives and academics to a strategic dialogue at their home in Mill Valley, California. When rain prevented the planned outdoor event, they improvised: small groups sat at kitchen tables, paper tablecloths served as note-taking surfaces, and after 20 minutes the guests switched tables. At the end of the evening, the group discovered that this improvised structure had produced deeper and more creative conversations than any planned conference [1].
Brown and Isaacs formalized this experience over the following years into a reproducible method. In 2002, they founded The World Cafe Community Foundation. In 2005, the seminal work The World Cafe: Shaping Our Futures Through Conversations That Matter appeared, codifying the method in 7 design principles [1].
The theoretical basis connects several strands: David Bohm’s dialogue theory (1996), which emphasizes the difference between discussion (arguments against each other) and dialogue (thinking together) [2]; Margaret Wheatley’s work on living systems and self-organization [3]; and research on collective intelligence showing that networked small groups under certain conditions make better decisions than individuals or plenary groups [4].
Reach: The World Cafe has since been used in over 70 countries — in companies, government organizations, NGOs, universities, and communities. The World Cafe Community counts over 200,000 documented applications worldwide [1].
The 7 design principles
Brown and Isaacs defined 7 principles that distinguish a World Cafe from informal table conversation [1]:
1. Set the Context
Clarify before the World Cafe: What is the purpose? What outcomes do we expect? Who needs to participate? A World Cafe without clear context produces pleasant conversations but no actionable results.
Practical rule: Formulate the context in one sentence to share with participants in the invitation: “We are conducting this World Cafe to achieve [specific goal].“
2. Create Hospitable Space
The physical environment influences conversation quality. Cafe atmosphere is not a decoration tip — it is a methodological requirement. People speak differently at a small table with coffee than in a conference room with fluorescent lighting.
Practical rule: Round or square tables for 4-5 people. Paper tablecloths and markers on every table. Beverages and snacks. Background music during arrival. If possible: natural light.
3. Explore Questions That Matter
The quality of guiding questions determines the quality of results. A good World Cafe question is open, inviting, has no obvious answer, and prompts thinking beyond what participants already know.
| Weak question | Strong question |
|---|---|
| ”How can we increase revenue?" | "What would need to change for our customers to actively recommend us?" |
| "What are our strengths?" | "If our customers were to describe what makes us unique — what would they say?" |
| "How do we improve our service?" | "What do our customers experience at the moment they need us most?" |
| "What should we do next?" | "What assumption about our customers have we never tested?” |
The three-question rule: Use progressively building questions across the three rounds, leading from assessment to vision. Example: Round 1: “What do our customers experience today?” — Round 2: “What stands in the way of an outstanding experience?” — Round 3: “What would our service look like if anything were possible?“
4. Encourage Everyone’s Contribution
A World Cafe works only when everyone speaks — not just the loudest. The table size (4-5 people) is methodologically grounded: in groups over 6, parallel conversations and silent participants inevitably emerge.
Practical rule: The table host’s task is to actively include quiet participants: “[Name], you mentioned something interesting earlier — would you like to share it here?“
5. Connect Diverse Perspectives
The rotation principle is the methodological core of the World Cafe. When participants switch tables, they carry ideas and insights from the previous round — “cross-pollinating” the conversation at the new table. Over three rounds, a network of connected ideas emerges that could not have formed at any single table.
6. Listen Together for Patterns
During conversations, every participant — and especially the table host — watches for recurring themes, surprising connections, and emergent patterns. Writing and visualizing on the paper tablecloths supports this “listening together.”
7. Share Collective Discoveries
The harvest phase at the end of the World Cafe is the decisive moment — and the one most frequently mishandled. This is where collective insights are made visible. Without a successful harvest, the World Cafe produces 20 tables full of written paper but no shared understanding.
Step by step: Conducting a World Cafe
Preparation (2-4 weeks ahead)
Participants: 20-200 people. Below 20, the diversity of perspectives suffers — a workshop format is better suited. Above 200, logistics become complex but remain feasible.
Composition: As diverse as possible. The World Cafe derives its power from connecting different perspectives. A World Cafe with 40 product managers produces different results than one with 10 product managers, 10 customers, 10 service employees, and 10 developers.
Room planning:
- Round or square tables, 4-5 chairs per table
- Paper tablecloths (white or brown, writable) and markers in various colors on each table
- Enough space between tables for movement during rotation
- Microphone and flip charts for the harvest
- Beverages and snacks (authentic cafe atmosphere)
Prepare table hosts: Select one host per table and brief them 30 minutes before the start: their role is to (1) keep the question present, (2) encourage everyone’s participation, (3) at each table switch, summarize the core ideas of the previous round in 60 seconds, and (4) watch for patterns.
Execution
Opening (15 minutes): Welcome, set the context, explain the rules:
- Focus on listening and discovering, not debating
- Write and draw on the paper tablecloths
- Build on others’ ideas and develop them further
- At the signal, switch tables — the host stays
Round 1 (20-25 minutes): The first guiding question is introduced. Groups discuss, write, draw. The host ensures balanced participation.
Rotation (3-5 minutes): Signal (bell, gong, music). Everyone except the host stands up and distributes across other tables — ideally not staying with the same group. The host summarizes the core ideas for the new group (60 seconds).
Round 2 (20-25 minutes): The second guiding question is introduced (or the same question is deepened). New groups build on the ideas from the previous round.
Rotation + Round 3 (20-25 minutes): Same principle. The third question typically moves from analysis to vision or concrete action.
Optional Round 4: For complex topics or large groups, a fourth round may be useful. Three rounds are the standard.
Harvest phase (20-30 minutes): The most critical part. Three harvest formats:
Format A — Gallery Walk: Paper tablecloths are hung on walls. All participants walk around, read, and mark the most important insights with dot stickers.
Format B — Table host reports: Each host has 90 seconds to present the most important insight from their table. The moderator collects patterns on a flip chart.
Format C — Cluster Harvest (recommended for actionable results): All sticky notes and annotations are collected on a large wall. The full group clusters them by theme and dot-votes the top 3 insights or action areas.
Closing (10 minutes): The top 3 insights are presented. The moderator names next steps: Who processes the results? When will they be communicated? What happens next?
Timeline overview
| Phase | Duration | Cumulative |
|---|---|---|
| Opening | 15 min | 15 min |
| Round 1 | 20-25 min | 40 min |
| Rotation | 3-5 min | 45 min |
| Round 2 | 20-25 min | 70 min |
| Rotation | 3-5 min | 75 min |
| Round 3 | 20-25 min | 100 min |
| Harvest | 20-30 min | 130 min |
| Closing | 10 min | 140 min |
| Total | approx. 2.5 hours |
Practical example: Service vision for a hospital
Context: A large hospital in the DACH region plans to redesign its outpatient patient journey — from the first phone call to aftercare. The challenge: 14 different departments are involved, each with a different perspective on “good service.” The executive board wants to involve all stakeholders in the vision development before a service design project is launched.
Participants: 80 people — physicians, nurses, administrative staff, reception personnel, IT, quality management, and 15 former patients (as experts on the patient perspective).
Setup: 16 tables with 5 people each. Each table is deliberately cross-functionally composed (no all-physician or all-nursing tables). Paper tablecloths, markers, coffee. Three table hosts are external moderators, 13 are internal leaders who were briefed beforehand.
Three rounds with progressive questions:
Round 1 — Assessment (25 min): “What do patients experience today on their journey through our hospital — from the phone call to aftercare? What works well, what does not?”
Tables fill up quickly. Patients report orientation problems (“I was sent to the wrong corridor three times”), physicians report information gaps (“I cannot see what my colleague discussed last week”), reception staff report overload at peak times.
Round 2 — Obstacles and causes (25 min): “What stands most in the way of a seamless, empathetic patient journey — and why?”
Patterns emerging across multiple tables: (1) Departmental silos — each department optimizes for itself, no one owns the entire patient journey. (2) IT systems that do not communicate. (3) Time pressure that prevents empathetic interactions. A patient puts it pointedly: “Every individual here is friendly. But the system is not.”
Round 3 — Vision (25 min): “What would the ideal patient journey look like if anything were possible?”
Tables become visionary: “A single point of contact who guides me through the entire process.” “I do not have to tell my story five times.” “I always know where to go and what happens next.” “Waiting times become information times.”
Harvest (Cluster format, 30 min): All paper tablecloths are hung on walls. Participants walk through and mark the most important ideas. The moderation team clusters in real-time on a large board. Result: 5 action areas — patient navigation, information continuity, waiting time utilization, point-of-contact model, cross-departmental process ownership.
Outcome: The 5 action areas became the foundation of the subsequent service design project. The executive board reported that acceptance of changes was significantly higher than with previous top-down initiatives — because 80 stakeholders felt they had co-created the vision.
Note: This example is illustratively constructed to demonstrate the method in a service design context.
Comparison: World Cafe vs. Open Space vs. Fish Bowl vs. Design Sprint
| Dimension | World Cafe | Open Space Technology | Fish Bowl | Design Sprint |
|---|---|---|---|---|
| Inventor | Brown & Isaacs (1995) | Harrison Owen (1985) | Origin unclear, widespread since 1960s | Jake Knapp / Google Ventures (2016) |
| Participants | 20-200+ | 5-2,000 | 10-50 | 5-7 (core team) |
| Duration | 2-3 hours | 0.5-3 days | 30-90 minutes | 5 days |
| Structure | High (fixed rounds, questions, rotation) | Low (participants set own topics) | Medium (inner circle discusses, outer circle listens) | Very high (fixed 5-day plan) |
| Control | Moderator steers questions and timing | Self-organized (Law of Two Feet) | Moderator steers transitions | Sprint master controls all phases |
| Output | Collective insights, patterns, action areas | Topic landscape, action plans | Deepened discussion of one topic | Validated prototype |
| Strength | Activates all participants, connects perspectives | Highest self-organization, emergent topics | Deep discussion with simultaneous transparency | Concrete, testable result |
| Weakness | Results often remain at insight level | Requires high trust in self-organization | Limited participant count for active discussion | Not suitable for large groups |
| Best for | Stakeholder alignment, vision development, idea networking | Complex topics with many perspectives, conferences | Controversial topics with expert discussion | Concrete product development with user validation |
Our recommendation: Use the World Cafe when you want to activate and connect many stakeholders before a concrete project starts. Use Open Space when participants should determine which topics matter. Use Fish Bowl for controversial topics with expert discussion. Use a Design Sprint when you want a testable prototype at the end of the week.
Combining World Cafe + Design Thinking: A World Cafe is excellently suited as the opening of a design thinking project. The World Cafe generates breadth (What are the relevant topics and perspectives?), design thinking generates depth (How do we concretely solve the most important problems?).
5 common mistakes with World Cafe
1. Weak guiding questions — the chat hour
Symptom: Table conversations are pleasant but superficial. Results are platitudes: “We should communicate better,” “The customer is at the center.”
Why it hurts: The quality of results is directly proportional to the quality of questions. Vague questions produce vague answers. Closed questions produce yes/no answers instead of exploration.
Fix: Invest at least 2 hours in question development. Test the questions with 2-3 people beforehand: Do they provoke thought? Is there no obvious answer? Do the three rounds build on each other?
2. Too many people per table — the mini-plenary
Symptom: 7-8 people sit at a table. Two people speak, the rest listens. The cafe intimacy is lost.
Why it hurts: The World Cafe works through intimacy. At a table with 4-5 people, everyone speaks. From 6 people, subgroups form; from 8 people, you have a mini-plenary — exactly what the format is designed to prevent.
Fix: Strictly 4-5 people per table. With 80 participants: 16-20 tables, not 10-12. Better more tables than overfull tables.
3. No harvest — the paper tablecloth wasteland
Symptom: The three rounds are over. 20 tables are full of notes. The moderator says: “Thank you very much, that was a wonderful World Cafe.” No one knows what the collective insights are.
Why it hurts: Without a harvest, the World Cafe is a pleasant afternoon but not an instrument. Participants leave feeling they “talked a lot” but without clear insight or direction for action.
Fix: Plan at least 20-30 minutes for the harvest. Use the cluster format (Format C): collect, cluster, prioritize. Name 3-5 concrete insights or action areas at the end.
4. Homogeneous table composition — the echo chamber
Symptom: All marketing people sit together, all developers sit together, all customers sit together. Tables confirm each other instead of generating new perspectives.
Why it hurts: Connecting different perspectives is the core principle. Homogeneous tables produce homogeneous results — insights that each department already had.
Fix: Assign table seats in advance (e.g., colored name badges determining starting tables). Ensure that each table has a mix of functions, hierarchy levels, and — if possible — external perspectives (customers, partners).
5. No follow-through — the flash in the pan
Symptom: The World Cafe was inspiring. Three weeks later, no one remembers the results. Nothing happens.
Why it hurts: A World Cafe that produces no change is more expensive than no World Cafe — because it disappoints participant expectations and undermines trust in participatory formats.
Fix: Document results within 48 hours. Communicate the top insights to all participants. Name those responsible for follow-up. Schedule a follow-up (workshop, project kickoff) within 4 weeks.
When the World Cafe does NOT work
1. Fewer than 20 participants: Below 20, the critical mass for networking effects is missing. Rotation across 3 tables does not produce real perspective diversity. Use a classic workshop or brainstorming instead.
2. When a concrete decision must be made: The World Cafe generates insights and perspectives — not decisions. If a yes/no decision must stand at the end of the day, a decision workshop is the right format.
3. With hierarchical blockage: If a board member sits at the table and everyone else falls silent because they feel the hierarchy, the format does not work. Solution: either deliberately mix hierarchy (no all-executive tables) or place management in an observing role.
4. With highly controversial topics and entrenched positions: If two camps face each other (e.g., union vs. management during restructuring), the World Cafe does not produce rapprochement — it produces argumentation at small tables instead of the large one. Mediation formats or Fish Bowl are better suited here.
5. Without follow-through: If the organization is not prepared to translate World Cafe results into concrete projects or decisions, do not use the format. A World Cafe without follow-up is worse than none — it demonstrates to participants that their voice is heard but not taken seriously.
Frequently asked questions
What is the World Cafe method?
The World Cafe is a structured dialogue method for large groups (20-200+ participants) where small groups discuss guiding questions at cafe tables and rotate after each round. A table host stays and connects ideas across rounds. At the end, collective insights are made visible in a harvest phase. The method was developed in 1995 by Juanita Brown and David Isaacs [1].
How many people do you need for a World Cafe?
At least 20, ideally 30-80. Below 20, perspective diversity is insufficient — a workshop format is better. Above 200 is possible but requires professional facilitation, multiple hosts, and a large venue. Table size is always 4-5 people, regardless of the total group size.
How long does a World Cafe take?
A standard World Cafe with 3 rounds takes 2-2.5 hours (15 min opening, 3x25 min rounds, 2x5 min rotation, 25 min harvest, 10 min closing). With 4 rounds or a more elaborate harvest: 3 hours. Never plan less than 2 hours — below that, the harvest suffers.
What is the difference between World Cafe and Open Space?
The World Cafe works with predefined guiding questions and fixed rotation — the moderator steers topic and timing. Open Space Technology (Harrison Owen) lets participants set their own topics and decide which discussion to join (Law of Two Feet). World Cafe suits focused exploration; Open Space suits emergent topic discovery.
What are good World Cafe questions?
Good World Cafe questions are open, inviting, and have no obvious answer. They often begin with “What if…,” “What would need to change so that…,” or “What do … experience when…” They build progressively across rounds: from assessment (Round 1) through obstacles (Round 2) to vision (Round 3). Question development is the most important preparation step.
Related methods
- Design Thinking: World Cafe as the opening of a design thinking project — first breadth (World Cafe), then depth (Design Thinking)
- Service Design: The World Cafe as a stakeholder alignment method in the Discover phase of a service design project
- Multi-Stakeholder Design: World Cafe as one of several formats for involving many stakeholders
- Brainstorming: For smaller groups (under 20) as an alternative to the World Cafe
- Service Design Methods Overview: Positioning the World Cafe within the overall methodological landscape
Research methodology
This article synthesizes insights from the seminal work by Brown and Isaacs (2005), David Bohm’s dialogue theory, research on collective intelligence, and the analysis of 10 German-language publications on the World Cafe. Sources were selected by methodological rigor, practical relevance, and currency.
Limitations: The academic literature on the World Cafe is predominantly descriptive — controlled efficacy studies are lacking. Evidence for superiority over other large-group formats is anecdotal, not experimental. The practical example (hospital patient journey) is illustratively constructed, not a documented case study.
Disclosure
SI Labs provides consulting services in the area of service innovation and uses the World Cafe as a stakeholder alignment method in the early phase of the Integrated Service Development Process (iSEP) — particularly when many stakeholders need to be included in problem definition. This practical experience informs the positioning of the method in this article. Readers should be aware of the potential perspective bias.
References
[1] Brown, Juanita, and David Isaacs. The World Cafe: Shaping Our Futures Through Conversations That Matter. San Francisco: Berrett-Koehler, 2005. ISBN: 978-1576752586 [Foundational Work | Standard Reference | Citations: 3,000+ | Quality: 85/100]
[2] Bohm, David. On Dialogue. London: Routledge, 1996. ISBN: 978-0415336413 [Foundational Work | Dialogue Theory | Citations: 8,000+ | Quality: 90/100]
[3] Wheatley, Margaret J. Leadership and the New Science: Discovering Order in a Chaotic World. 3rd Edition. San Francisco: Berrett-Koehler, 2006. ISBN: 978-1576753446 [Foundational Work | Living Systems | Citations: 6,000+ | Quality: 82/100]
[4] Woolley, Anita Williams, Christopher F. Chabris, Alex Pentland, Nada Hashmi, and Thomas W. Malone. “Evidence for a Collective Intelligence Factor in the Performance of Human Groups.” Science 330, No. 6004 (2010): 686-688. DOI: 10.1126/science.1193147 [Empirical Study | N=699 People | Citations: 2,500+ | Quality: 92/100]
[5] Prewitt, Vana. “Working with the World Cafe.” Collectivit (2011). [Practitioner Guide | Application Reports | Quality: 65/100]
[6] Fouche, Christa, and Gillian Light. “An Invitation to Dialogue: ‘The World Cafe’ in Social Work Research.” Qualitative Social Work 10, No. 1 (2011): 28-48. DOI: 10.1177/1473325010376016 [Journal Article | Methodological Reflection | Citations: 200+ | Quality: 75/100]