Executive Paper
Executive Decision Quality
A model for assessing process quality without confusing it with the subsequent outcome.
Executive summary
A sound decision is not merely one that ends well. It is taken against a defined question, genuine alternatives, evidence proportionate to risk, visible assumptions and an agreed way to review or stop the action. Outcomes also reflect chance and information unavailable at the time; process quality and subsequent results should therefore be assessed separately.
What must be defined
Before options are compared, state what is being decided, who has authority, when the decision is due and what happens if no action is taken. A useful formulation sets the objective, constraints and maximum acceptable exposure. A proposal containing only one solution is not yet a comparison.
- Decision: the specific choice to be made.
- Options: at least one viable alternative and the no-action case.
- Criteria: expected value, risk, reversibility, resources, dependencies and time.
- Evidence: confirmed facts kept separate from estimates, views and assumptions.
- Governance: owner, approvers, execution conditions and review date.
Evidence proportionate to risk
The Green Book calls for proportionate comparison of options, costs, benefits and risks, with uncertainty made explicit and evaluation planned. This does not mean collecting every available fact. It means identifying the evidence capable of changing the choice and testing that first. Reversible decisions can proceed at a reasonable evidence threshold and learn; difficult-to-recover capital, legal commitments or potential harm to others require a higher standard.
Speed without improvisation
Eisenhardt’s study of high-velocity firms found that fast decisions did not dispense with information: they used real-time data, multiple alternatives and active conflict resolution. It is a bounded study rather than a universal rule, but it helps distinguish speed from haste. Sustainable speed comes from defining decision rights, reliable sources and escalation thresholds in advance.
Minimum decision record
A concise record should preserve the question, options, sources, critical assumptions, discrepancies, exposure, decision, owners and review triggers. It should also specify what evidence would invalidate the thesis. This prevents retrospective reconstruction and supports learning without judging a decision solely by its outcome.
Limits
This framework structures reasoning but does not replace audit, legal advice, independent valuation or technical verification. Nor does it turn an estimate into a fact. Where evidence is insufficient, the defensible outcome may be to seek more information, limit commitment, wait or stop.
Sources and scope
- HM Treasury, The Green Book 2026. Primary source; publication or version: ; accessed: .
- Kathleen M. Eisenhardt, Making Fast Strategic Decisions in High-Velocity Environments. Primary source; publication or version: ; accessed: .
- James G. March, Exploration and Exploitation in Organizational Learning. Primary source; publication or version: ; accessed: .
- OECD, G20/OECD Principles of Corporate Governance 2023. Primary source; publication or version: ; accessed: .
These sources support the framework and its general claims. Applying it to a specific decision requires current data, transaction documents and specialist advice where appropriate.