Running small experiments
A method for letting an idea meet reality while the cost of being wrong remains bounded and reversible.
A small experiment is not a miniature performance designed to confirm the idea. It is a bounded encounter that could produce an inconvenient answer.
Define:
- the assumption being tested;
- the smallest real behaviour that could challenge it;
- the maximum acceptable cost;
- the observation window;
- the conditions for continuing, changing, or stopping.
An experiment may be inconclusive. That is different from failure. It may show that the question was too broad, the signal too weak, or the situation too unstable.
Use Mapping assumptions first. Then reflect through Decision journaling.
// RESEARCH BASIS
Why we hold this for now.
Evidence strengthens a position without making it universal. This records the current basis, limits, possibilities, and conditions for revision.
- Current position
- Supported, moderate confidence
- Basis
- research synthesis
- Scope
- Decisions where a meaningful assumption can be tested through a bounded, informative, and reasonably reversible intervention.
- Last reviewed
- Aug 08, 2026
- Review by
- Aug 08, 2027
- Research links
- 1 evidence link · 0 counterpoints
Evidence & sources
Counterpoints
No explicit counterpoint is attached yet. Absence of a counterpoint is not evidence of consensus.
BOUNDARY CONDITIONS
- Some choices are irreversible, ethically unsuitable for experimentation, too rare, or too slow to generate useful feedback.
- Small tests can produce misleading signals when samples are weak, outcomes are noisy, or the experiment changes the system being measured.
- The strongest source attached here is entrepreneurial and should not be generalized automatically to every domain.
POSSIBILITIES
- Explicit predictions and stopping conditions may reduce escalation of commitment by making disconfirming evidence harder to reinterpret after the fact.
- Evidence that hypothesis-driven bounded testing consistently worsens decision quality relative to comparable unstructured iteration.
- Evidence that the benefits observed in entrepreneurial settings do not transfer to the kinds of product and organizational decisions we apply this method to.