In this guide
THE CONVERSATION BOARD
To discuss 03What slows us down?A better handoverWhere to start?
Discussing 01One topic.
Everyone’s attention.
Discussed 01Our next experiment
Original example board. Propose topics, vote, then discuss one at a time. At each time check, the group decides whether to continue.
Lean Coffee conversation board

An original worksheet. Print it or save it as a PDF.

Open worksheet

Lean Coffee is a participant-led conversation in which the group proposes topics, chooses what to discuss, and makes the queue visible. It is useful when people have shared interests but different questions. The output is usually understanding, connections, and a few agreed follow-ups.

The originators’ website credits Jim Benson and Jeremy Lightsmith with starting Lean Coffee in Seattle in 2009. It describes a simple board and participant voting. Lean Coffee™ is identified there as a trademark of Modus Cooperandi. This guide is an independent explanation, and its downloadable board is an original layout. Origin and format

When to use it

Use it for a community of practice, a cross-team learning session, an informal retrospective, or an exploratory conversation about a proposed change. A useful invitation gives people a boundary without supplying every topic: “Bring questions about how our teams handle customer feedback.”

Do not use it as the only mechanism for a decision requiring accountable approval, balanced evidence, or consultation with absent people. It is a poor fit for a confidential grievance, an emergency briefing, or a meeting in which every item must be covered. Choosing conversation topics does not give the room authority to change policy.

For a first session, plan for four to eight people and 60 minutes. These are practical starting choices, not validated limits. With a larger group, split into tables and reserve a shared closing discussion.

Expected outcome and materials

Leave with a visible record of topics discussed, unresolved questions, and any actions people explicitly accepted. A session without actions can still be worthwhile if learning was the purpose. Do not invent tasks just to make discussion appear productive.

Use a board with three columns: To discuss, Discussing, and Discussed. Add a separate area for follow-ups. Supply readable cards, pens, a timer, and voting marks. Remote groups can use a shared document with the same headings; a whiteboard is optional. Provide keyboard-accessible ways to contribute and do not rely on color to indicate status.

Open the printable Lean Coffee board.

A whiteboard with coloured idea notes grouped inside hand-drawn outlines.
Figure 1. An idea board at a Wikimedia Deutschland design thinking workshop, July 2018. This is an illustration of visible notes, not a photograph of Lean Coffee. Use the guide’s printable board for the actual conversation columns.Design Thinking Workshop WMDE 1 — Corinna Schuster (WMDE) · CC BY-SA 4.0. Resized and converted to WebP; no crop. Copies retain this license.

Prepare the invitation

Write the purpose, finish time, topic boundary, and expected output. Explain whether notes will be shared and whether names will appear. Ask participants to avoid personal or customer information on a board that may be photographed.

Choose a facilitator who will protect the process, a timekeeper, and someone to capture agreed follow-ups. One person can cover these roles in a small group, but separating them makes participation easier. If a senior leader attends, ask them to avoid opening each discussion with their preferred answer.

Confirm that participants may pass. Silence is not consent, and an invitation to speak should not become a demand for disclosure. Research on psychological safety links the perceived safety of interpersonal risk taking with team learning behavior; it does not establish that a meeting format creates that safety. Edmondson’s study

A 60-minute facilitation plan

The schedule below is this guide’s adaptation. Protect the closing five minutes even if a conversation is lively.

0–5 minutes: establish the purpose

Explain the board and how the group will choose topics. State that votes decide where to spend discussion time, not whose concern matters. Agree how to handle a time check: a quick simultaneous indication to continue or move on, followed by a short extension if the group wants one.

Say: “We can leave a topic unresolved. Before moving on, we will record what remains open and whether anyone wants a follow-up.”

5–10 minutes: write topics independently

Ask for one specific question per card. “How can we get timely feedback from support?” is easier to discuss than “Communication.” Encourage contributors to describe the question without proposing the answer inside its title.

Silent writing is a facilitation choice intended to give people time before public discussion. University of Wisconsin–Extension’s description of nominal group technique also separates individual generation from group sharing; Lean Coffee is a different format and should not be presented as equivalent to that structured consensus method. Wisconsin Extension guidance

10–17 minutes: clarify and choose the queue

Give each author a brief opportunity to explain their card. Clarify meaning without debating the substance. Combine duplicates only with the authors’ agreement: two cards that mention “feedback” may concern different problems.

For this session, give each person two votes and allow both on one topic. This matches the basic voting arrangement on the originators’ page. Order cards by votes and use a simple agreed tie-break, such as the order in which tied cards were proposed.

Ask: “Which conversations would be most useful to you today?” Keep a visible note of low-vote topics; their position is about today’s queue, not their inherent importance.

17–55 minutes: discuss in short rounds

Move the first card to Discussing. Invite its author to give enough context to start, then open the discussion. Use an initial five-minute timer as a local convention. At the time check, ask everyone to indicate whether more time would help.

If the group wants to continue, grant a three-minute extension and check again. If preferences are mixed, ask what one remaining question would justify an extension. Avoid allowing repeated extensions to consume the entire session without an explicit group choice.

The facilitator can say: “We have heard two suggestions. What evidence or experience would help us compare them?” Or: “We have not heard from everyone; anyone who wants to add a different perspective is welcome.”

Before moving a card to Discussed, record a brief takeaway. Distinguish an idea, a decision, and an accepted action. “Try shorter forms” is an idea. “Maya will draft a shorter form by Friday” is an action only if Maya accepts it and has permission to do so.

A CLOSER LOOK

At the timer: make the next choice explicit.

  1. ContinueThe group chooses one short extension.
  2. Move onRecord the useful point and take the next topic.
  3. Follow upA willing owner accepts a separate action or conversation.

A discussion does not automatically become a decision.

Figure 2. Original facilitation aid for the adapted agenda in this guide. Agree the continuation rule with participants before the first round; these are practical options, not prescribed official stages.Original illustration · Innovation & Change

55–60 minutes: close and confirm

Read the follow-ups aloud and confirm owners and review dates. Ask participants for one useful insight or one improvement to the format. Explain what will happen to unaddressed cards. Participants may bring them to a later session, arrange a smaller conversation, or decide they no longer need discussion.

Take a board snapshot only under the agreed note-sharing arrangement.

Hypothetical example: support and product teams

Six colleagues meet to discuss feedback moving between support and product development. Their cards include unclear escalation rules, requests without context, duplicate reports, and uncertainty about release dates.

The group votes to discuss escalation rules first. During the first round, a support colleague explains that urgent reports are sent to several channels because nobody knows who is responsible. A developer describes receiving the same report without a customer impact description.

The facilitator records a shared uncertainty: the group has no agreed definition of urgency. They do not settle an organization-wide policy in the meeting. Instead, two volunteers accept a follow-up to examine recent cases and propose a draft escalation guide for the accountable manager.

A low-vote release-date question remains on the board. Its author arranges a separate conversation with the release coordinator. That is a legitimate outcome; covering every card was never promised.

Common mistakes and adaptations

The highest-ranking person explains every card. Return topic clarification to its author and invite other perspectives before the leader responds. Private or simultaneous voting can make preferences less visible during selection, but does not remove power differences.

Voting becomes a claim of consensus. Say what the vote means. A short preference vote does not establish informed agreement, expertise, representativeness, or permission. CDC describes nominal group technique as a more structured prioritization process and notes its constraints, including preparation and limited discussion. Choose such a process when structured prioritization is the actual task. CDC evaluation brief

The timer becomes punitive. Use it to check whether the group wants more time, not to interrupt someone mid-sentence. Announce the approaching boundary.

Every comment becomes an action. Capture questions and learning separately. Unowned tasks create misleading minutes.

For remote sessions, send access instructions beforehand and allow cards through chat or a document. For a sensitive change discussion, offer an advance private channel to the facilitator, explaining its confidentiality limits. For a large community, run several small tables and share only a few insights at the end.

Evidence and next steps

Lean Coffee is a practitioner-derived format. Its origin page supports the attribution and basic mechanics. The research and institutional sources here inform participation choices and limitations; they are not trials demonstrating that Lean Coffee improves project performance. The timing, prompts, example, and closing routine are this guide’s practical adaptation.

After a session, check whether people found the conversation useful and whether accepted follow-ups happened. If the next need is to explore solutions to one defined problem, use the 90-minute idea workshop. If a question concerns who is missing from a change discussion, start with stakeholder mapping.

Sources: