Risk vs uncertainty

Risk and uncertainty are not the same condition. Risk can be framed with known variables, plausible outcomes and at least approximate probabilities. Uncertainty appears when relevant variables, outcomes or causal links are not fully knowable before acting.

In brief

Many SME decisions are treated as risk problems when they are actually uncertainty problems. This matters because the two conditions require different decision logic.

Risk can be managed with estimates, scenarios, sensitivity analysis, insurance, buffers and controls. Uncertainty requires learning, reversibility, optionality, staged commitments and stop rules.

Confusing the two creates false precision. The company may produce numbers that look rational while the real problem is that the future state has not yet revealed itself.

Operational definition

A decision is under risk when the main variables are known and can be assigned reasonable ranges or probabilities. The company may not know what will happen, but it knows what could happen well enough to compare options.

A decision is under uncertainty when the relevant variables, outcomes or interactions are partly unknown. More analysis can help, but it cannot remove the condition entirely.

The practical distinction is not philosophical. It changes the action. Under risk, the company tries to choose the best expected option. Under uncertainty, it tries to choose a path that preserves learning and avoids irreversible commitments too early.

Why it matters for SMEs

SMEs often face decisions with limited data: entering a new market, hiring a first manager, buying an ERP, acquiring a competitor, changing the operating model, delegating a decision previously held by the founder.

These decisions are rarely pure risk. They contain uncertainty because the organisation will learn only after acting. The new manager’s fit, the real adoption of a system, the reaction of customers, the effect on informal routines: these cannot be fully known in advance.

If the company treats uncertainty as risk, it overinvests in plans and underinvests in learning design.

Observable signals

Look for plans that depend on assumptions nobody can test yet.

Look for forecasts with precise numbers but weak causal evidence.

Look for decisions framed as “yes or no” when a staged option exists.

Look for discomfort with the phrase “we do not know yet”.

Look for irreversible commitments made before the company has learned enough.

Common mistakes

The first mistake is adding decimals to ignorance. A spreadsheet can make uncertainty look measurable without making it better understood.

The second mistake is waiting for certainty that cannot arrive. Some information exists only after action.

The third mistake is making a large irreversible commitment when a smaller learning step is possible.

The fourth mistake is confusing caution with inaction. Under uncertainty, good action is often small, reversible and informative.

Operational example

An SME wants to enter a new customer segment. The risk framing asks for a five-year revenue forecast. The uncertainty framing asks a different set of questions: what must be true for this segment to work, what can be tested in 60 days, what commitment would close future options, and what signal would make us stop?

The company runs a limited pilot with three target accounts, one offer variant and a defined stop rule. It learns that the segment is promising but requires a different onboarding process. Because the commitment was staged, the company can adapt before scaling.

Diagnostic questions

Which variables are known enough to estimate?

Which variables will only become visible after action?

Which decision would be hard to reverse?

What is the smallest action that would generate useful evidence?

Which options should remain open while the company learns?

What stop rule protects the company from continuing a wrong path?

Practical implications

Classify decisions before planning them. If the decision is mostly risk, use analysis, scenarios and controls. If it contains deep uncertainty, design staged commitments.

Prefer reversible experiments when possible. Preserve optionality. Name assumptions. Define learning milestones. Decide in advance what evidence would expand, revise or stop the initiative.

MARTRO reading

In MARTRO’s reading, the distinction between risk and uncertainty protects organisations from false precision. It prevents the company from treating structural unknowns as if they were only numbers missing from a model.

The method therefore links uncertainty to optionality, stop rules and door-closing decisions. The question is not only “what is the expected outcome?” but “what does this decision make impossible later?”

Frequently asked questions

Is uncertainty the same as lack of data? Not exactly. Lack of data can sometimes be fixed. Uncertainty remains when the relevant future state cannot be fully known before acting.

Can risk and uncertainty exist together? Yes. Many decisions have measurable risk components and uncertain structural components.

What is the best response to uncertainty? Staged action, learning loops, reversible commitments, optionality and explicit stop rules.

Is waiting safer under uncertainty? Not always. Waiting may preserve options, but it may also close them. The point is to act in a way that produces evidence without overcommitting.

Why does this matter in organisational diagnosis? Because many operating problems cannot be solved by more forecasting alone. The organisation must learn through controlled action.