Prevention creates value. It also creates a measurement problem.

How do you value a loss that did not happen?

A technical failure is corrected before rankings decline. A Product explanation is clarified before a campaign sends more customers into the journey. A competitor's movement is identified early enough for the organisation to protect the segment.

What is that worth? The honest answer is rarely a single certain number.

Explore Growth and Value Review Results and Proof
Start with the condition, not the claim

Record what was exposed, the affected Product, customer or market, the loss mechanism, baseline evidence, likelihood and potential consequence, the approved intervention, the changed condition and residual exposure.

Use a controlled estimate

Value exposed × likelihood of loss × reasonable influence factor × period = modelled avoided value

The inputs and assumptions must be declared. Sensitivity ranges are usually more honest than a single number.

Keep financial authority clear

SI can establish the evidence, loss mechanism and contribution logic. Finance should validate the financial method and the use of the resulting number. Until then, the claim remains Modelled—not Validated.

Value can be non-financial and still material

Protected customer confidence, preserved first-page presence, reduced paid dependency, avoided rework, reduced technical exposure, faster decision-making, retained category position. Not every useful outcome needs to be forced into revenue.

A disciplined avoided-loss example

A high-value application journey fails intermittently for a defined mobile segment. The organisation can observe the failure rate, affected sessions, Product value, historical completion behaviour and the period of exposure — it may model avoided value using a range of assumptions.

Several uncertainties remain: not every failed session would have completed, customers may return through another channel, and seasonality may change demand. A responsible record uses a range, labels the assumptions and preserves alternative explanations.

Three levels of value language
Descriptive

"A commercially important journey failure was removed."

Supported contribution

"Completion improved after the failure was corrected, while other material conditions were reviewed."

Modelled financial value

"Based on the approved assumptions, the intervention may have protected value within a defined range."

The language should advance only as the evidence advances. Avoided-loss models are often more credible when they show a conservative, central and upper range — a single precise figure can imply a level of certainty the evidence does not support. The range makes uncertainty visible and allows Finance to challenge the assumptions. The organisation should also preserve the non-financial result: removing a material failure, protecting Product truth or retaining a strong category position can be valuable even when the financial amount cannot yet be validated.

Avoided cost is broader than revenue

The organisation may also avoid emergency remediation, repeated investigation, additional paid acquisition, customer-support demand, reputation response, duplicated supplier effort, delayed Product launch, and future migration complexity.

Governance questions for Finance
Which baseline is approved?
What counterfactual is being assumed?
Which period is reasonable?
What other factors could explain the result?
What confidence range is appropriate?
Who may use the estimate and for what purpose?
When can the claim be treated as Validated?

The goal is not to undervalue prevention. It is to value it in a way that survives challenge.

Value prevention with discipline.

Explore Growth and Value, or return to the cornerstone framework.

Explore Growth and Value Return to the cornerstone
Related guidance and implementation pathways
Search Intelligence for CFOs

Help CFOs inspect value created, protected, recovered or avoided across demand, paid…

Explore →
Decision Confidence

Understand customer, evidence and organisational Decision Confidence—and why more data does…

Explore →
Evidence Trail

See how SI connects the business question, baseline, evidence, owner, implementation,…

Explore →