Workshop Design Canvas: Plan Who Hears What, and When

Nine to five, and nothing on the wall at four

Picture the agenda. 9:00, welcome and objectives. 9:30, market context. 10:30, breakouts on growth opportunities. 12:00, lunch. 1:00, report-backs. 2:30, prioritization. 4:30, next steps. It looks like a workshop. It is a list of topics with times attached.

At four o'clock the walls are covered in sticky notes, the report-backs ran long, prioritization became a show of hands, and “next steps” is a promise to circulate the photos. Nobody did anything wrong. The day was never designed to produce anything in particular.

Workshop design is deciding, before the day, what the session must produce and how each part of it will run: who works alone, who works in pairs, when the whole room talks, and how it converges. The canvas on this page puts that on one sheet in ten boxes. The idea behind it: an agenda lists topics; a design decides who hears what, and when. That second question, more than the content, is where workshops are won or lost.

An agenda lists topics. A design decides who hears what, and when.

Why a canvas and not an agenda

An agenda answers one question: when does each topic happen? The canvas answers the four that decide the outcome. What will exist at the end? How will the group converge, and who decides? Who is in the room, and what does each person know that others do not? And in what order will people speak?

The first question is the one most often skipped. In Michael Mankins's survey of top teams at 187 large companies, more than 65% of leadership meetings were held for information sharing, group input or group discussion rather than to make a decision, and the meetings that did focus on strategy were most commonly off-site brainstorming sessions that produced few tangible outputs ¹. A workshop with no named output tends to become that.

The fourth question is the one most often left to chance. Cass Sunstein and Reid Hastie describe how groups follow whoever speaks first and neglect information only one or two people hold ⁵. Order is a design choice, so the canvas treats it as one. Facilitation handbooks treat agenda design as a skill in its own right; the 2014 third edition of Sam Kaner's Facilitator's Guide to Participatory Decision-Making expanded its material on designing realistic agendas to three chapters ⁶.

Running the day is a separate craft, covered in the strategy workshop facilitation guide. If the session is a leadership offsite that must end in decisions, the offsite agenda builder assembles the agenda once this canvas is done.

The canvas

Copy the ten boxes onto one page, or draw them on a whiteboard with the people who will co-design the session. Each prompt is the question the box must answer. A box you cannot fill is the first thing to fix.

Boxes 1, 4 and 7 carry most of the weight. Box 1 tells you whether you need a workshop at all. Box 4 tells you whether it can end. Box 7 is where the design actually happens.

How to fill it in

1. Start with the output, not the agendaWrite box 1 first and show it to the sponsor. If you cannot describe the artifact, you are not ready to design the day. Often the honest answer is a smaller output than the sponsor hoped: three real options, not a strategy.
2. Choose one kind of sessionBox 2 forces a choice. Do not plan to generate options and choose among them in one sitting when the stakes are high. Textron split each strategic issue into two half-day sessions, the first to agree facts and a limited set of viable alternatives, the second to evaluate them and choose ².
3. Write the rule before the peopleFill box 4 with the sponsor before you invite anyone. If the sponsor will decide, say so on the invitation. If the room will vote, say how ties are broken.
4. Invite for information, not statusFor box 5, list each person next to what they bring. Anyone with nothing in the second column is an observer; ask whether they need to be there. Anyone missing who holds a key piece of information is a gap.
5. Design the flow lastBox 7 comes only after boxes 1 to 6, because the blocks exist to produce the output from the people you have. Work backward from the final block.
6. Name the risks out loudBox 8 is for you and the sponsor, not the wall. Agree how you will handle each one, including when the sponsor will speak.

Designing the flow: modes, not just topics

Every block in box 7 gets a mode as well as a topic. The modes below are the building blocks. Each one either keeps views independent or brings them together; a good flow alternates, and never lets the room converge on a question before people have committed their own view to paper.

Three rules we use, which are practitioner judgment rather than research: put a silent or small-group block before every whole-room discussion that matters; budget at least as much time for converging as for generating, because convergence always runs over; and end every block with an output written where everyone can see it, so the next block starts from something rather than from memory.

A filled-in canvas

The company is invented. The session is a half-day diverge workshop, the kind most often mistaken for a decide workshop.

Look at box 5 and box 8 together. The person most likely to change the outcome, the support lead with data on customers outside the core, is also the person most likely to stay quiet. The flow puts silent writing and small groups ahead of every whole-room discussion for exactly that reason.

Designing for the room you actually have

Three second-order problems show up once the canvas is in regular use.

• A room of people who already agree. Sunstein and Hastie report that deliberating groups tend to end up more extreme than their members' starting point, on matters of fact as well as values ⁵. If box 5 lists only people from one function or one camp, the session will amplify its starting view. Add someone who disagrees, or a red-team block with real time.
• The sponsor who has already decided. If box 4 says “the CEO decides” and the CEO has decided, a diverge workshop is theater, and participants usually sense it. Ask the sponsor directly what answer would surprise them. If none would, change box 2 to commit and run a planning session honestly.
• Remote and hybrid rooms. In our experience, silent writing and private votes carry over to a screen more easily than whole-room talk does. Liberating Structures suggests a chat variant for sharing in 1-2-4-All, in which each person types one idea and holds it until prompted so that all appear at once, and a 1-3-All shortcut when moving pairs into fours is impractical ⁴.

And one discipline for the designer: keep box 9 honest. Every off-the-agenda item needs a destination, a meeting or an owner. A parking lot with no exit teaches people that raising a concern is how you lose it.

The case for less design

The fair objection: over-designed workshops feel like a training course. Every minute is scripted, energy drains, and the best conversation of the day, the one that wandered somewhere unexpected, gets cut off because the timer went.

There is truth in it. The Liberating Structures guidance for 1-2-4-All itself pairs precise timing with a warning against over-helping or overcontrolling the group's conversations ⁴. Structure and freedom are not opposites; the question is what you hold fixed.

Hold boxes 1 and 4 fixed: the output and the rule. Let box 7 flex. If a conversation is producing exactly what box 1 needs, extend it and cut a later block. If it is producing something else, however interesting, park it in box 9 with a destination. The canvas is not a script. It is the agreement that lets you improvise without losing the point.

Before you send the invitation

• Box 1 describes an artifact, not an activity.
• Box 2 names one kind of session.
• Box 4 names a person or a rule, agreed with the sponsor.
• Every name in box 5 has something in the second column.
• Every whole-room discussion in box 7 follows a silent or small-group block.
• Converging gets at least as much time as generating.
• Box 10 has a name and a date.

The agenda at the top of this page had a time for everything and a purpose for nothing. Fill in the canvas first and the agenda almost writes itself. Skip it, and the day will be decided by whoever happens to speak first.

Questions

What is a workshop design canvas?

A workshop design canvas is a one-page plan for a session, filled in before the agenda. It records the output the workshop must produce, the kind of session, the questions, who decides and how the room converges, who attends and what each brings, pre-work, the flow of blocks with their modes, risks to the conversation, off-agenda topics and the follow-up.

How do you design a workshop?

Work backward. Write what must exist at the end, decide whether the session widens options, chooses, or commits, agree the decision rule with the sponsor, then invite people for the information they hold. Design the flow last, giving every block a mode as well as a topic, and put silent or small-group work before each whole-room discussion that matters.

What is the difference between a workshop design and an agenda?

An agenda says when each topic happens. A design says what the session will produce, who decides, who brings which information and in what order people will speak. The agenda is one output of the design. Two workshops with identical agendas can produce very different results depending on whether people write their views before the most senior person speaks.

Should a workshop both generate options and choose among them?

For high-stakes choices, separate them. Textron, as Mankins and Steele describe it, handled each strategic issue in two half-day sessions: one to agree the facts and a limited set of viable alternatives, and a second to evaluate the alternatives and choose. For small, reversible choices one session is fine.

Where to go next

• Strategy Workshop Facilitation: Who Speaks First Decides
• Strategy Offsite Agenda Builder
• Decision Meeting Checklist: Most of It Happens the Day Before
• Pre-Mortem Worksheet
• Design Thinking