Example data. Fieldnote is a fictional company used to fill all 61 boards, so the examples tell one consistent story.
Pre-PMF Fieldnote, before building offline mode: Karim, Tom and Ana map the assumptions behind the first version.
What the team took away
'Customer present' was the riskiest unproven assumption: it became the first experiment, with 5 technicians and a stopwatch.
How to run it
Preparation
Invite the trio: product, tech, design. Add the sponsor only if the initiative is big enough that their beliefs matter.
Send the problem statement or the vision board the day before. Ask everyone to arrive with 10 assumptions written down.
On the board: the 2x2 (importance × evidence) and three colored sticky sets: desirability, feasibility, viability.
Prepare the list of what you already know: analytics, past research, competitor facts. It stops people from testing what is already proven.
Agenda
10 min
Explain the three risks Desirability: do they want it. Feasibility: can we build it. Viability: does it work for the business. One example of each from a past failure of the company.
25 min
Dump assumptions Silent writing: 'we believe that…', one per sticky, colored by risk type. Aim for 40 or more. Quantity first, no filtering.
20 min
Deduplicate and sharpen Group duplicates, split vague ones. 'Users will like it' becomes 'admins will finish setup in under 10 minutes without help'.
30 min
Place on the grid One at a time, the group places each sticky: how much does the project die if this is false, how much evidence do we have. Argue for 30 seconds max per sticky.
20 min
Pick the top right The high-importance, low-evidence corner is the test list. Rank it. Take the top 3 to 5.
15 min
Design the tests For each: the cheapest test that could prove it false in under two weeks, the metric, the owner, the deadline.
Pitfalls
Assumptions written as facts ('the market is growing') do not get tested. Force the 'we believe that' prefix.
Do not let engineers only write feasibility and PMs only desirability. Everyone writes all three colors.
A test that takes six weeks is not a test, it is the project. Shrink it until it fits in two weeks.
Afterwards
The top assumptions become the rows of the experiment board and the questions of the research plan. Bring the grid back after each test and move the stickies: evidence up, or importance down.