The first unlock is not a feature, a market map, or a vision deck. It is a completed exchange that a defined participant accepts and chooses to repeat. Until that happens, potential is a hypothesis. The work is to make the hypothesis small enough that reality can answer it.
A door is a constraint with an owner
"Find the door" is useful only when it points to a precise constraint: someone cannot complete an important job, uses an expensive workaround, and has authority to change the current path. A broad market pain is not yet a door. A founder needs the participant, the moment of friction, the current substitute, the consequence of delay, and the person who can accept a different result.
The accepted-exchange test
The smallest unit of venture truth is a result that crosses a real boundary.
Core idea: An exchange is accepted when a participant uses the outcome in their work, takes the next step it enables, and would notice its removal.
- One participant has a bounded need now, not an abstract preference
- The venture produces an observable result, not only insight or activity
- The participant has a way to accept, reject, correct, or escalate the result
- The result changes a decision, handoff, commitment, or completed task
- The next exchange can happen with less founder orchestration
Do not confuse attention with acceptance
Interested users, positive interviews, and demo applause can be useful signals. None proves that a new exchange belongs in a workflow. Acceptance carries a cost: the participant gives time, data, permission, reputation, money, or a changed habit. That cost is exactly why it is more valuable than a stated intention. A practical discovery engagement should seek this evidence before it promises scale.
Design the first exchange by hand
Manual work is not a failure of ambition at this stage. It is how a team learns what must be standardised, where judgment belongs, and what the participant actually values. If the team cannot deliver the result with a clear human owner, it does not yet know enough to automate it. The manual version should be narrow, time-bound, and instrumented - not a disguised consulting project that hides an undefined product.
What would disconfirm it
Stop or narrow the hypothesis if the participant will not grant the necessary access, cannot explain why the result matters, returns only after repeated founder prompting, or corrects the work at a cost greater than the current workaround. These are not minor onboarding issues. They are information about the chain you are trying to unlock.