Skip to content
Better Decision Making
Module 1 of 5

Separate Musts, Wants and Adverse Consequences

11 minutes · Advanced

Learning objective

By the end of this lesson, you can prevent an attractive option from compensating for a failed mandatory requirement and can separate benefit scoring from risk judgement.

Why it matters

A weighted spreadsheet can make any choice look rigorous. If mandatory conditions, desired outcomes and risks share one score, a legally or operationally unacceptable option may still rank first.

Core concept: eligibility before preference

Use three gates:

  1. Musts establish eligibility with measurable pass/fail evidence.
  2. Wants compare eligible alternatives using weights defined before scoring.
  3. Adverse consequences test whether the leading benefit is worth its residual downside.

A failed Must cannot be offset by a high Want score. A weighted total informs judgement; it does not own the decision.

Visual explanation

Text alternative: alternatives first pass mandatory requirements. Eligible choices are compared on desirable outcomes, then checked for adverse consequences before an accountable decision.

Worked example

An autonomous support-routing option has the best expected time saving but fails the required urgent-case sensitivity. It is ineligible. The team may change the requirement only through an explicit governance decision—not by raising other weights.

A supervised option wins on weighted Wants. Its main adverse consequence is reviewer overload, so the pilot includes an exception-age limit and return to manual routing.

Common mistake

Mistake: changing weights after the team sees which alternative wins.

This converts importance into retrospective justification. Define and record weights first; if the context changes, version the decision model openly.

Quick check

An option fails a mandatory data-location requirement. What should happen?

A. Add more desirable criteria.
B. Remove it from the eligible set unless the requirement is explicitly changed.
C. Give the Must a high weight.
D. Score the unknown as average.

Answer: B. Musts are pass/fail conditions, not weighted preferences.

Practical prompt

Choose a current technology decision. Write three testable Musts, four Wants, one evidence note per Want and two adverse consequences for the leading alternative.

Summary

  • Define eligibility before preference.
  • Test Musts with evidence.
  • Weight Wants before scoring.
  • Review adverse consequences separately.
  • Keep the decision owner and review trigger visible.

Next lesson

Continue with Compare Two at a Time, where a portfolio of AI opportunities must be ranked without changing the comparison criterion.

Next step

Compare Two at a Time Without Moving the Criterion

Continue