Validation SOPs › Customer Discovery
SOP 15 of 100 · Customer Discovery
Customer Co-Creation SOP
- Difficulty
- Advanced
- Time
- 1 week to recruit, 2-3 sessions over 2-4 weeks
- Cost
- $200-1,000
- Best for
- Startups with a clear problem and lead customers who already hack their own fixes
- Tools
- Miro, FigJam, Figma, Paper and markers, Zoom, Google Forms
- Measure
- Session return rate, Customer-originated elements, Commitment signals, Workarounds shared
A customer co-creation SOP brings 5 to 8 of your most motivated customers into the design of the solution, over two or three sessions spread across two to four weeks. Interviews tell you the problem and focus groups tell you which concept people prefer. Co-creation goes further: customers sketch, build and argue about the product with you, and the ones who helped build it become your first buyers, references and loudest supporters. It is the most demanding SOP in Customer Discovery, and the one with the strongest signal at the end.
What is the customer co-creation SOP and when should you use it?
Customer co-creation means designing the solution together with a small group of customers. The startup version builds on Eric von Hippel's lead user research: lead users face a need earlier than the rest of the market, and many build their own fixes. C.K. Prahalad and Venkat Ramaswamy later popularised "co-creation of value" in business strategy. A familiar public case is LEGO Ideas, where fans submit set designs and a design that gathers 10,000 supporters goes to LEGO for review.
Use this SOP:
- After interviews confirm a strong problem and you have met people who already hack their own fixes.
- When the right solution is unclear and you have two or three directions.
- When early trust matters, such as B2B sales or tight professional communities, where buyers back what they helped shape.
If you only need people to compare finished concepts, a focus group is cheaper. If you want ongoing advice from customers after launch, set up a customer advisory board. For a sketching session with your own team, use the solution sketching workshop.
What do you need before you start?
- A validated problem from 15 or more customer interviews.
- 5 to 8 lead customers: people with the problem at its sharpest who have already built a spreadsheet, script, template or process to cope.
- A simple agreement that says ideas shared in the sessions can be used by your company, what participants get in return and how you handle confidential information. Ask a lawyer to check it.
- Materials: a Miro or FigJam board for remote sessions, paper, markers and sticky notes for in-person ones, and a Figma file for the final version.
- A thank-you: a fee, free access for a year, or credit as a founding customer.
How do you run it, step by step?
1. Recruit lead customers with a workaround filter
From your interview notes and communities, pick people who already built something to cope with the problem. Ask them to bring it to the first session. A person who made a 12-tab spreadsheet thinks about the solution far more than someone who only complains.
2. Session 1: show and tell, then define success (90 min)
Each participant shows their workaround and explains what it does well and where it breaks. Together, write a shared list of "a good solution must…" statements. End with each person ranking the top five.
3. Between sessions: you build two or three rough concepts
Turn the ranked list and the best parts of their workarounds into two or three low-fidelity concepts: paper screens, a slide flow or a spreadsheet mock. Label which customer each element came from.
4. Session 2: build and break (2 hours)
Give participants the concepts and ask them to change, combine or redraw them. Work in pairs, then present to the group. Your role is to ask "why?" and to note every change. Hold back from defending your own ideas; the moment you argue, people stop building.
5. Between sessions: merge into one clickable concept
Build a single clickable prototype in Figma from the session 2 results. Keep a list of which elements came from customers and which from you.
6. Session 3: test and commit (60-90 min)
Each participant runs through a real task with the prototype while others watch. Collect final changes. Then ask for a commitment that costs something: a signed letter of intent, a paid pre-order or an agreement to run a pilot with their team.
7. Write the co-creation brief
Document the final concept, the elements each customer contributed, the commitments and the open questions. Thank every participant publicly if they agree, and keep your promise on the thank-you.
Template you can copy
CO-CREATION PLAN
Problem (validated): __________________________
Lead customer filter: has built a workaround for [problem] in the last 12 months
PARTICIPANTS
Name/ID | Role | Workaround they built | Agreement signed Y/N | Thank-you
SESSION 1: Show and tell (90 min)
- Each person shows their workaround (8 min each)
- Group list: "A good solution must..."
- Private ranking of top 5 must-haves
SESSION 2: Build and break (2 h)
- Present 2-3 rough concepts (10 min)
- Pairs change, combine or redraw (45 min)
- Pairs present, group discusses (45 min)
- Note: element | from whom | why
SESSION 3: Test and commit (60-90 min)
- Each person completes a real task in the prototype
- Final changes
- Commitment ask: letter of intent / pre-order / pilot
SCORECARD
Returned for all sessions: __ of __
Final concept elements from customers: __ of __
Commitments: LOI __ Pre-orders __ Pilots __
Worked example: shift swaps for hospital nurses
This is an illustrative case with invented numbers. A founder has interviewed 18 hospital nurses about the stress of swapping shifts. She recruits 7 who built their own fixes: three WhatsApp groups with strict rules, two shared spreadsheets, a paper board in a break room and one nurse's colour-coded calendar system.
In session 1, the group agrees on five must-haves, led by "the ward manager sees swaps without being asked" and "nobody breaks staffing rules by accident". In session 2, two nurses merge the spreadsheet and paper board ideas into a swap request that checks qualifications automatically. Of 14 elements in the final concept, 9 come from participants.
Six of 7 return for all three sessions (86%). In session 3, four nurses agree to introduce the founder to their ward managers, and two managers sign letters of intent for an unpaid four-week pilot. The founder now has a concept, a pilot and seven people who will tell colleagues about it.
What should you measure, and when do you stop?
| Metric | Pass signal | Warning signal |
|---|---|---|
| Session return rate | 70%+ attend every session | Half drop out after session 1 |
| Customer-originated elements | 50%+ of final concept elements came from participants | Participants only approved your ideas |
| Commitment signals | 3+ letters of intent, pre-orders or pilots | Praise and no commitment |
| Workarounds shared | Every participant brings a real workaround | Participants have never tried to fix the problem |
Stop after three sessions. If the group cannot agree on one concept by session 3, split it by customer type and run a shorter round with each half. If nobody commits at the end, the problem is real and the urgency is low, so you go back to problem validation.
Which mistakes should you avoid?
- Recruiting fans instead of lead users. Friends and enthusiasts approve; people with workarounds build.
- Arriving with a finished design. Customers then only polish your idea. Bring rough concepts they can tear apart.
- Skipping the agreement. Unclear ownership of ideas causes trouble later. Sort it out before session 1.
- Letting one participant dominate. Use pair work and private rankings so quiet experts are heard.
- Ending without an ask. Co-creation that ends with "thanks, we'll be in touch" wastes the strongest moment of trust you will have.
How can AI speed this up?
Between sessions, paste session notes and photos of sketches into an AI assistant and ask it to list every proposed element, who suggested it and how many people in the group agreed. Ask for two or three ways to merge the elements into one flow. AI tools also turn a paper sketch into a first Figma wireframe description quickly. Keep customers' own words and sketches visible in the next session, because people commit to what they recognise as theirs, and an AI-polished version can erase that.
Where does this SOP fit in your validation plan?
Co-creation closes Customer Discovery and opens Solution Validation.
- Confirm the problem with customer interviews.
- Plan the product's first release with user story mapping.
- Design it with lead customers using this SOP.
- Turn letters of intent into payments with a pre-order campaign.
See all 15 Customer Discovery SOPs. The Fe/male Switch startup game has founders build with early customers under real deadlines, and our sister site of 100 startup methods covers related workshop formats.
FAQ
What is customer co-creation?
A process where a small group of customers designs the solution with you over several sessions: they share their workarounds, sketch and change concepts, test a prototype and then commit to it.
How many people should take part in co-creation sessions?
Five to eight. Fewer gives too little variety; more makes building together slow and loud.
Who owns ideas that come out of co-creation sessions?
Whoever your agreement says. Have participants sign a short agreement before session 1 that assigns ideas to your company and states what they get in return, and ask a lawyer to check it.
What is a lead user?
Eric von Hippel's term for a customer who faces a need before most of the market and builds their own fix. Lead users are the best co-creation participants.