Evidence before scale
The smallest useful test before a major commitment
The smallest useful test is the least irreversible action that can produce evidence strong enough to change the decision. It tests one risky claim with relevant participants, observable behaviour, a predefined threshold, a time and cash limit, and a stop condition. A test is not useful merely because it is cheap or easy. A survey that cannot change the decision, a prototype shown only to friends, or a landing page with no meaningful commitment may create activity without reducing the decisive uncertainty.
Start with the commitment you are trying to avoid making too early. Name the one assumption that could make it premature. Then design a safe test that resembles the real decision closely enough to reveal friction without pretending to prove everything that comes later.
Evidence boundary
Separate what you know from what you hope
| Known | Assumed | Still unknown |
|---|---|---|
| The proposed commitment, downside, available time/cash and safety constraints. | The chosen participant and behaviour are close enough to the real buying or operating decision. | Whether the observed behaviour will clear the threshold under realistic friction. |
| The current alternative and the decision that follows each possible result. | A positive result would justify the next bounded step rather than a different explanation. | Whether delivery cost, repeatability and later constraints will remain workable. |
A useful test specifies what evidence would move the decision before data is collected.
Decision focus
The decisive unknown: what evidence would change the decision?
If no plausible result would make you continue, revise or stop, the test is theatre. Write the three decision paths first. Select a behaviour that is difficult to fake and proportionate to the risk: payment where appropriate, access to a real workflow, completion of a manual trial, repeated use, or another scarce commitment. Do not manufacture pressure or collect sensitive information to make the signal stronger.
Illustrative scenario — not a customer case or outcome claim
A manual workflow tests a software assumption
A consultant is considering an automated client-intake tool. The decisive unknown is whether clients will complete a shorter structured intake without repeated chasing. A polished interface will not answer that.
For two weeks, five eligible new clients receive the same manual intake, deadline and one reminder. The consultant records completion, missing information, time to a usable brief and follow-up count. The threshold is four usable briefs with no more than one follow-up each. The test costs no new software and stops if sensitive data would be required. It cannot prove that software will integrate safely, sell or reduce long-term cost.
Bounded action
Run the smallest useful test
- Write one hypothesis. Name the participant, situation, observable behaviour and reason it matters.
- Choose the least irreversible simulation. Deliver the outcome manually or through an existing safe path before building infrastructure.
- Set a threshold. Define the count, completion, payment, time or quality evidence that maps to each decision.
- Set time and cash limits. Use a short calendar window and a fixed maximum loss you can accept.
- Add safety and stop conditions. Stop if the test needs sensitive data, unclear consent, unqualified advice or a promise you cannot fulfil.
- Record friction, not only success. Separate behaviour, participant comments and your interpretation.
Conditional judgment
Continue, revise, or stop
Continue
The relevant behaviour clears the predefined threshold and the next commitment remains bounded and safe.
Revise
The test reveals real interest or need, but the participant, promise, workflow, channel or threshold was wrong.
Stop
The right participant will not make the commitment, the test is unsafe, or no result can justify the proposed next step.
What this evidence can and cannot tell you
One small test can reduce one material uncertainty. It cannot prove scalable demand, legal compliance, profitability, repeatability or future results. Match the test to the commitment: development, commercial lease readiness, or hire versus outsource.
If you want one defined commercial opportunity judged before choosing its next test, explore the existing Kairos Idea Preflight. Regulated, legal, tax, employment and financial questions still require the appropriate qualified professional.