What needs to be finished first
A concrete initiative, MVP concept, or strategic hypothesis is written down and clearly depends on assumptions for success.
The team and stakeholders are willing to name assumptions explicitly instead of treating them as facts.
What needs to be ready before start
2x2 matrix as a whiteboard or digital board (axes: importance and uncertainty); sticky notes for assumptions; markers; example assumptions from comparable projects; initiative description visible.
One facilitator with discovery experience; Product Owner or strategy owner; 3-7 participants from product, engineering, research, and sales for a mix of perspectives; optionally a Devil's Advocate.
Initiative description; known hypotheses; research status (validated, open, disproved); existing test backlog items; time horizon for validation.
45-60 min
Put the matrix on the wall: X-axis uncertainty (low on the left, high on the right), Y-axis importance (low at the bottom, critical at the top). Label all four quadrants (critical to test in the top right, known in the bottom left). Put the initiative statement at the top. Set the timer to 45 min.
The one question this method answers
Which assumptions do we need to validate first because they are both critical to the initiative and uncertain?
Marker: Phase
| Step | Duration | Action | Hint |
|---|---|---|---|
1Phase 1: Collect assumptions | 10-15 min | Everyone writes down the assumptions underlying the initiative on their own first. Trigger categories: desirability, viability, feasibility, usability. | Phrase assumptions explicitly as 'We assume that ...'. Anything stated as a fact is suspicious and usually also an assumption. |
2Phase 2: Rate importance | 10 min | For each assumption: if it is wrong, does the initiative break? Y-axis: high (showstopper) to low (tolerable). Move the sticky notes to the right height. | Importance is often underestimated for supposedly obvious things. Ask for each assumption: what happens if the opposite is true? |
3Phase 3: Rate uncertainty | 10 min | For each assumption: how much evidence do we have? X-axis: left (validated several times) to right (pure guess). Include research or past tests. | Personal experience is not evidence for a target group. If the answer is just 'we believe so', it belongs on the uncertain side. |
4Phase 4: Top risks in the upper right quadrant | 5-10 min | The upper-right quadrant (critical + uncertain) becomes the test backlog. Prioritize the top 3-5 assumptions there for the next validation round. | If there are 10+ assumptions in the red quadrant, there are too many critical unknowns. The initiative is too early or the scope is too broad. Split it up. |
5Phase 5: Tests per top assumption | 10-15 min | For each top assumption, choose a validation method: interview, survey, prototype test, fake door, competitor analysis, data spike. Note success criteria and owner. Estimate effort. | The method must fit the assumption. You do not validate desirability with an engineering spike. You do not validate feasibility with a survey. |
What comes out at the end
Assumption Map (2x2 matrix as image or export) with placed assumptions, top-assumption list with rationale, plus a test backlog with validation method, success criterion, owner, date, and status tracking.
One map per initiative with date. After every assumption test, update the assumption as validated (move left), disproved (pivot the initiative), or refined. Refresh the map every 2-4 weeks.
Checklist for critical assumptions by risk and knowledge level.
Concrete filled scenario, fictional example
Concrete filled scenario, fictional example
Initiative: B2B mobility bundle for industrial employers with 300-2000 employees. Pricing EUR 8/employee/month. Pilot with 3 customers.
Assumption clusters by position:
Upper right (critical + uncertain) - top 5 test backlog
Upper left (critical + confirmed)
Lower right (unimportant + uncertain)
Lower left (unimportant + confirmed)
Test backlog
| # | Assumption | Method | Success criterion | Owner | Date |
|---|---|---|---|---|---|
| 1 | HR budget authority | 5 structured interviews with HR leads | >=3/5 confirm authority up to EUR 10k/month | @marcus | week 22 |
| 2 | Tax clarification | Tax advisor consultation + example model | Clear answer on tax treatment is possible | @anna | week 23 |
| 3 | Personio integration | Technical spike with Personio API | Sample data flow works in 3 days | @ben | week 22 |
| 4 | Provider margin | Negotiation with DB plus tier for pilot | Written margin commitment | @marcus | week 24 |
| 5 | Employee adoption | Letter-of-intent survey with 30 employees per pilot customer | >=50% LOI rate | @lisa | week 26 |
Recognize symptoms and steer against them
Stakeholders phrase statements as facts ('users want this'), and no assumptions are marked.
The facilitator asks for evidence: 'How do we know? Which data or tests support this?' If there is no evidence, it remains an assumption.
The list contains only 5-8 assumptions, and many important ones are missed.
Work with trigger categories (desirability, viability, feasibility, usability). Aim for at least 3 assumptions per category. Have a Devil's Advocate actively ask for more.
Assumptions are moved into the lower half because they seem 'not that important'.
Test: if the assumption is false, what happens? If the initiative could fail, it belongs at the top. If nobody is affected, it belongs at the bottom.
Assumptions are packed into the right half even though evidence exists.
Ask for evidence per assumption. Anyone with evidence (research, data, tests) may move left. Otherwise it stays on the right.
The validation plan exists, but nobody owns the tests.
An owner is mandatory for each test. Without an owner, the test is wishful thinking. Convert tests into sprint items or spikes.
Done signals checkable in under a minute
Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.