Evidence-first methodology
Turn each assumption into something the market can answer.
The O1SF method connects every strategic claim to an observable interaction. It helps a founder distinguish a confident story from a tested market signal.
The loop
Six steps from hypothesis to action.
A learning cycle is only complete when the evidence changes - or deliberately confirms - a decision. Conversation volume alone is not the goal.
- Name the assumption.Write a falsifiable claim about the customer, problem, buyer, channel, proof, or timing. Add the decision that would change if the claim is wrong.
- Create a testable asset.Turn the claim into positioning, an interview prompt, an outreach message, a target list, a product narrative, or another concrete artifact.
- Choose a relevant interaction.Match the asset to a person and context capable of producing useful evidence - not merely encouragement.
- Record observation separately from interpretation.Capture what was said or done, then state what the team believes it means. Keep contradictions and unknowns visible.
- Revise one decision.Change the target, message, proof, channel, offer, or next test. Avoid broad rewrites that make it impossible to know what created the difference.
- Commit to the next measurable action.Assign an owner, time horizon, threshold, and stopping rule for the next 3-, 6-, or 12-month decision.
Evidence quality
Separate signal from enthusiasm.
Signals are not equal. The method ranks evidence by relevance and behavior so that polite interest does not silently become a forecast.
Stronger signals
- A target customer describes the problem in their own language.
- A buyer explains the current workflow, budget, authority, and timing.
- A relevant person takes a costly next step: shares data, invites a stakeholder, tests an asset, or schedules a defined follow-up.
- Patterns repeat across independent conversations in the same segment.
Weaker signals
- General praise without a concrete use case or next action.
- Feedback from people outside the target buying context.
- Introductions that do not progress to a relevant conversation.
- A single memorable opinion treated as a market pattern.
Evidence log
Make every conclusion traceable.
A lightweight record makes the reasoning reviewable after the energy of a meeting or trip has faded.
For each interaction, record the segment and role, the assumption tested, the prompt or asset used, the observation, the team’s interpretation, confidence level, contradiction, and next decision. Do not store unnecessary personal data.
Review the log by segment and assumption - not by meeting count. A useful weekly synthesis names what became more likely, what became less likely, what remains unknown, and what the team will stop doing.
Use the customer-discovery guide to plan interviews and the go-to-market guide to translate learning into time-bound decisions.
Continue exploring
From assumptions to 💲 paying customers.
Review the evidence-first methodology, program structure, pricing, and prepared answers before drawing conclusions about fit. Paying customers and revenue are not guaranteed.