How to run participatory design workshops: a practical guide with AI prompts

What is participatory design?

Participatory design is a collaborative approach where end users, stakeholders, and designers work together as co-creators in structured workshop sessions to generate ideas, shape concepts, and define solutions. Rooted in the Scandinavian workplace democracy movement of the 1960s-70s, the method shifts users from passive research subjects to active design partners. Workshop activities use low-fidelity tools — sketches, collages, storyboards, physical prototypes — that require no design skills, ensuring that all participants can contribute on equal footing.

What question does it answer?

What solutions would users design for themselves if they had the tools and opportunity?
Which aspects of a problem matter most to the people who live with it daily?
What tacit knowledge do users hold that they cannot easily articulate in an interview but can express through making?
How do different stakeholder groups envision the same problem differently, and where do their visions converge?
What design directions would have genuine adoption because users helped shape them?

When to use

When designing for communities, contexts, or experiences where the design team lacks lived experience — participatory design brings in the expertise that no amount of desk research can replace.
When a product or service affects multiple stakeholder groups with potentially competing needs, and the design process itself must surface and negotiate those tensions.
Early in the design process (discovery or ideation), when the solution space is still open and participant input can genuinely shape direction rather than polish details.
When previous design efforts failed due to low adoption, and the team suspects the root cause is that users were not involved in defining the problem or the solution.
When organizational buy-in depends on stakeholders feeling ownership of the design — co-creating with people builds commitment that presenting research findings does not.
When working on public services, healthcare, education, or community tools where the ethical case for involving affected populations in design is strong.

Not the right method when the design space is tightly constrained (e.g., a regulatory form with fixed fields), when participants cannot be recruited or compensated for the time investment (workshops typically run 2-4 hours), when the team needs evaluative data about an existing design (use usability testing instead), or when speed is the priority and a single designer can iterate faster with user feedback loops than a workshop cycle allows. Participatory design also requires skilled facilitation — without it, workshops default to the loudest voices or become unfocused brainstorms that produce nothing actionable.

What you get (deliverables)

Design concepts and sketches created by participants, often capturing needs that the design team would not have identified alone
Prioritized problem statements reflecting what matters most to the people affected
Low-fidelity prototypes (paper, cardboard, digital collages) that express user mental models
Storyboards or journey narratives showing how participants envision interacting with a future solution
Stakeholder alignment: shared understanding of the problem and agreement on direction, built through the process itself
Design principles or requirements derived from workshop outputs
Photo and video documentation of the sessions for the broader team

Participants and duration

Participants: 6-12 per workshop. Fewer than 6 limits the diversity of perspectives. More than 12 makes facilitation difficult and reduces each person’s opportunity to contribute. For complex projects, run multiple workshops with different stakeholder groups (e.g., end users, frontline staff, administrators) rather than mixing everyone into one session.
Participant selection: Recruit for diversity of experience, not demographics. Include people with different levels of product familiarity, different use contexts, and different attitudes toward the problem. At least 50% should be end users; the remainder can include stakeholders, subject-matter experts, or frontline staff.
Session length: 2-4 hours for a single workshop. Shorter sessions rush participants past the warm-up phase where trust is built. Full-day sessions (6-8 hours) work for intensive design sprints but require breaks and varied activities to maintain energy.
Number of sessions: 1-3 workshops per project phase. One workshop generates ideas; a second can refine and prioritize; a third can evaluate low-fidelity prototypes with the same or new participants.
Total timeline: 2-5 weeks (1-2 weeks for planning and recruitment, 1 week for workshops, 1-2 weeks for synthesis and translation into design requirements).

How to run a participatory design workshop (step-by-step)

1. Define the design challenge

Frame the challenge as an open question that leaves room for participants to shape the direction. “How might we make the appointment booking experience work for patients who manage multiple chronic conditions?” is a good frame. “Should we add a calendar widget to the booking page?” is too narrow — it pre-decides the solution and reduces participants to validators. Write the challenge on a single page and test it with a colleague: if someone outside the project understands what they are being asked to help with, the framing works.

2. Recruit participants with diverse lived experience

Identify who is most affected by the design challenge and recruit from those groups. If designing a school communication app, recruit parents, teachers, and school administrators — not just one group. Use behavioral criteria: parents who have missed school notifications, teachers who spend more than 30 minutes a day on parent communication. Over-recruit by 25-30%. Offer incentives appropriate to the time commitment ($100-200 for a 3-hour workshop; higher for specialized professionals). Be transparent about what participants are being asked to do and how their input will be used.

3. Design the workshop activities

Build a sequence of activities that moves from divergent (many ideas) to convergent (shared direction):

Warm-up (15-20 min): An icebreaker activity related to the topic — instead of “tell us your name and role,” use a low-stakes making exercise — draw your morning routine, build a Lego model of your biggest frustration. This normalizes the idea that everyone will be making things, not just talking.

Context sharing (20-30 min): Participants share their experiences with the problem space. Use structured prompts: “Draw or describe the last time you tried to [task]. What went well? What went wrong?” This surfaces the raw material that later activities will build on.

Generative activities (60-90 min): The core of the workshop. Choose 1-2 activities based on the design challenge:

Collaging or mood boards: Participants select and arrange images, words, and textures to express what a future experience should feel like. Good for exploring emotional and aspirational dimensions.
Paper prototyping: Participants sketch or build screens, physical products, or service touchpoints using paper, markers, sticky notes, and found materials. Works for tangible design challenges.
Storyboarding: Participants draw or narrate a sequence of events showing how they would interact with a future solution across time. Good for services and multi-step experiences.
Design-the-box: Participants design the packaging for a product or service that does not yet exist, forcing them to articulate what makes it valuable. Good for value proposition exploration.

Sharing and discussion (20-30 min): Each participant or small group presents what they created. The facilitator asks clarifying questions but does not critique. Other participants build on ideas (“I would add…” or “That reminds me of…”).

Prioritization (15-20 min): Dot voting, forced ranking, or a group discussion to identify which concepts or themes resonate most strongly across the group.

4. Prepare the physical or digital space

For in-person workshops: large tables, whiteboards, markers in multiple colors, sticky notes, printed images for collaging, scissors, glue, blank paper in various sizes, Lego bricks or modeling clay if relevant. Arrange the space so that every participant can see others and reach materials without asking.

For remote workshops: Miro or FigJam board pre-loaded with templates for each activity, a shared image library for collaging exercises, breakout rooms configured in Zoom, and a clear visual guide for how each activity works digitally.

5. Assign roles and brief the team

A participatory design workshop needs at minimum:

Lead facilitator: Guides the session, manages time, adapts activities based on group energy, ensures every voice is heard.
Note-taker/documentarian: Photographs all physical artifacts, captures key quotes and decisions, records the session (with consent).
Activity assistant(s): For groups of 10+, assign one assistant per 4-5 participants to help with materials, answer process questions, and keep small-group work on track.

If stakeholders or clients observe, establish ground rules: they watch and listen but do not steer or evaluate during the session. Their role shifts to active participation or to debriefing afterward.

6. Run the workshop

Open by explaining the purpose, setting ground rules (every idea has value, build on each other’s work, no hierarchy in this room), and getting consent for recording and photography. Move through activities at a steady pace — participants lose energy if one exercise drags, and they lose focus if transitions are rushed. Watch for power dynamics: if a senior stakeholder is in the room with junior staff, create structures that equalize contributions (written-before-verbal, small-group work before full-group sharing). Name the dynamic if needed: “In this room, a nurse’s experience with this problem is just as important as a director’s.”

7. Debrief with participants and close the loop

Before participants leave, spend 10 minutes on reflection: “What surprised you today? What idea are you most excited about? What did we miss?” Promise a follow-up (email summary, invitation to review prototypes). Closing the loop is an ethical obligation — participants gave their time and creativity, and they deserve to see what happened with their contributions.

8. Synthesize and translate into design requirements

Within 2-3 days, the design team reviews all workshop outputs: sketches, prototypes, storyboards, photos, notes. Cluster outputs by theme. Identify patterns: which needs appeared across multiple participants and multiple activities? Which design directions generated the most energy and alignment? Write design requirements in the form of “As [persona from the workshop], I need [capability], so that [outcome participants articulated].” Flag areas of disagreement between participant groups — these are design tensions to resolve, not to average.

How AI changes this method

AI compatibility: partial — AI can support workshop preparation (designing activities, generating stimulus materials) and post-workshop synthesis (clustering outputs, drafting design requirements), but the live workshop itself depends entirely on human facilitation. The value of participatory design lies in the social process of co-creation — people reacting to each other’s ideas, building on shared experience, negotiating priorities in real time — and this cannot be replicated by AI.

What AI can do

Workshop activity design: Given a design challenge and participant profile, an LLM can suggest a sequence of participatory activities with timing, materials lists, and facilitator instructions — a solid starting point that the facilitator then adapts.
Stimulus material preparation: AI can generate discussion prompts, scenario cards, “what if” questions, and persona cards to use as workshop props, saving hours of preparation time.
Post-workshop synthesis: An LLM can process facilitator notes, transcripts, and photographed sticky notes (via OCR) to identify recurring themes, cluster ideas, and draft a summary of key findings and design directions.
Prototype translation: AI image generation or wireframing tools can turn participants’ rough sketches and descriptions into higher-fidelity mockups that the design team can then refine — bridging the gap between workshop output and design artifact.
Reporting and stakeholder communication: An LLM can draft a workshop report with structured findings, participant quotes, and recommended next steps, reducing the time from workshop to action.

What requires a human researcher

Running the workshop: Reading participant energy, adjusting activities on the fly, managing group dynamics and power imbalances, and creating an environment where non-designers feel genuinely empowered to create — these are irreducibly human skills. A poorly run participatory design session produces outputs that reflect the facilitator’s assumptions, not participants’ ideas.
Navigating power dynamics: When a manager and their direct reports are in the same room, or when a designer’s expertise subtly overrides a user’s lived experience, only a human facilitator can recognize and correct these dynamics in real time.
Making ethical judgments about representation: Deciding who to recruit, whose voice to amplify, and how to handle disagreements between participant groups requires moral reasoning about equity and representation that AI cannot perform.
Interpreting ambiguity in creative outputs: A participant’s collage or sketch may contain meaning that is not literal — a dark color scheme may represent frustration, a recurring arrow may signal a desire for speed. The facilitator who was in the room understands this context; an AI analyzing a photo does not.

AI-enhanced workflow

The most time-consuming phase of participatory design has always been preparation: designing activities, creating materials, and logistics. An LLM can generate a complete workshop plan — activity sequence, timing, materials list, facilitator script — in 20 minutes, compared to the 4-8 hours a facilitator typically spends. The facilitator then reviews and customizes based on their knowledge of the specific participants and context. Stimulus materials (scenario cards, question prompts, persona summaries for participants) can similarly be drafted by AI and refined rather than created from scratch.

Post-workshop synthesis is the other major time sink. Traditionally, a facilitator spends 1-2 days reviewing photos of sticky notes, transcribing key quotes, and clustering insights into themes. With OCR tools that extract text from whiteboard photos and an LLM that clusters and summarizes, the first pass of synthesis can happen within hours. The facilitator then validates, corrects, and adds the contextual interpretation that only someone who was present can provide.

The workshop itself remains entirely human. A 3-hour participatory design session is a live collaborative event where the quality of the output depends on the facilitator’s ability to build trust, manage conflict, and help non-designers express their ideas through making. There is no AI shortcut for sitting next to a participant, handing them markers, and saying “Show me what this would look like.”

Tools

Workshop facilitation (in-person): Sticky notes, markers, blank paper, printed image libraries for collaging, Lego Serious Play kits, scissors, glue, whiteboards
Workshop facilitation (remote): Miro (templates for participatory design activities, voting, clustering), FigJam (collaborative whiteboarding), MURAL, Zoom (breakout rooms)
Documentation: High-resolution camera or phone for capturing artifacts, Otter.ai (session transcription), video recording with consent
Synthesis: Dovetail (tagging and clustering workshop outputs), Miro (digital affinity mapping), Google Slides (visual summary for stakeholders)
AI-assisted: ChatGPT/Claude (activity design, stimulus generation, output synthesis), OCR tools (extracting text from whiteboard photos), Midjourney/DALL-E (translating rough sketches into concept visuals)
Recruitment: User Interviews, Respondent, community organizations (for specialized populations)

Works well with

In-depth interview (Di): Run interviews first to understand the problem space deeply, then use participatory design workshops to generate solutions with users who represent the patterns found in interviews.
Journey mapping (Jm): Use workshop outputs (storyboards, scenarios) as raw material for building journey maps that reflect how users actually envision the experience, not just how the team imagines it.
Concept testing (Ct): After the workshop produces design concepts, test them with a broader audience through concept testing to validate whether the co-designed direction resonates beyond the workshop participants.
Persona building (Ps): Personas created from prior research help the facilitator design workshop activities and recruit participants who represent each persona type, ensuring the workshop covers the full range of user needs.
Card sorting (Cs): A card sorting exercise embedded in a participatory design workshop reveals how users categorize information and name things — directly informing information architecture decisions.

Example from practice

A city government was redesigning its online permit application system. The previous redesign, done by an external agency with no user involvement, had produced a technically polished system that residents found confusing and time-consuming. Completion rates were below 40%, and the call center handled 200+ calls per week from people who could not finish their applications online.

The design team ran three participatory design workshops: one with residents who had recently applied for permits, one with city staff who processed applications, and one with small business owners who applied frequently. In the resident workshop, participants built paper prototypes of their ideal application flow. A recurring pattern emerged: residents wanted to see all required documents upfront before starting, rather than discovering missing requirements mid-process (the existing system revealed requirements one step at a time). Staff participants, in a separate session, prototyped a dashboard that showed application status from the reviewer’s perspective — revealing that they spent 30% of their time on back-and-forth emails requesting missing documents from residents.

The team combined both insights into a redesigned flow: a pre-application checklist that showed all requirements before the applicant started, plus a status tracker visible to both applicant and reviewer. Six months after launch, online completion rates rose from 38% to 71%, and call center volume dropped by 45%. The city attributed the improvement directly to the co-design process, noting that the document checklist — the single most impactful feature — had come from a participant’s paper prototype, not from the design team.

Beginner mistakes

1. Treating participants as feedback machines rather than co-creators

The most common mistake is inviting users to a “participatory design” workshop and then presenting pre-made concepts for them to react to. This is concept testing, not co-creation. In genuine participatory design, participants generate the ideas — they sketch, build, and prototype. The facilitator provides tools and structure, not solutions. If participants leave feeling like they were consulted rather than that they created something, the method was misapplied.

2. Ignoring power dynamics in the room

When a department head sits at the same table as a junior employee, the junior employee self-censors. When a designer presents alongside a non-designer, the designer’s sketches carry implicit authority. Participatory design requires deliberate structures to equalize power: small-group work before large-group sharing, written-before-verbal responses, explicit statements that lived experience is as valuable as professional expertise, and seating arrangements that do not reinforce organizational hierarchy.

3. Designing too many activities for the time available

A 3-hour workshop with five generative activities, two sharing rounds, and a prioritization exercise will rush through everything without depth. Participants need time to think, make, discuss, and build on each other’s work. Two well-paced generative activities produce richer output than four rushed ones. Leave 20% buffer time for activities that run long and for the unplanned conversations that often produce the best insights.

4. Not closing the loop with participants

Participants who invest 3 hours co-creating solutions and then never hear what happened feel used. This damages trust and makes future recruitment harder, especially in community contexts. Send a follow-up within two weeks: what themes emerged, which ideas are being taken forward, and how participants can stay involved. If a participant’s idea becomes a real feature, tell them.

5. Running the workshop without prior research

Participatory design works best when it builds on existing understanding. A workshop with no prior research produces a wish list that may or may not reflect real needs — participants design for their most recent frustration rather than the underlying pattern. Conduct interviews or field research first to understand the problem space, then use the workshop to generate solutions grounded in that understanding.