Decision-making series · Part 6 of 6
Decision quality is not outcome quality: five records that make decisions governable
21 October 2026 · 6 min read · DecixionIQ
A well-reasoned decision can produce a poor outcome because of events nobody could foresee. A poorly reasoned decision can produce a good one through luck. If a board judges decisions only by their results, it rewards luck and punishes sound judgement — and learns the wrong lessons.
Measure four things separately
The most important principle in governing decisions is to keep four kinds of quality apart, because each can fail on its own.
| Quality | The question | Where the evidence lives |
|---|---|---|
| Decision quality | Was the reasoning sound with what was known at the time? | Framing, evidence, assumptions, alternatives, challenge |
| Decision authority | Did the right person or forum decide, within their powers? | Approval record, conflicts and recusals, conditions |
| Execution quality | Was the decision carried out as approved? | Owned actions, milestones, controls |
| Outcome quality | Did it produce the intended result? | Outcome review against the approved case |
When an outcome disappoints, this separation tells you what to fix. Weak reasoning calls for better challenge. A decision taken outside authority calls for clearer delegations. Poor execution calls for better ownership. And a sound, properly authorised, well-executed decision that still fell short may simply have met bad luck — which is a lesson about uncertainty, not about the people involved.
Five records that make the separation possible
You can only judge these four things later if the right records were kept at the time. Five connected instruments do most of the work:
- Decision Brief. The structured case: the problem, alternatives, evidence, financial implications, risk assessment and the proposed resolution. This is the evidence for decision quality.
- Decision Rights Matrix. A formal map of who recommends, who challenges, who approves, who executes and who reviews — the reference for decision authority.
- Decision Register. The authoritative record of decisions, owners, approvals, assumptions, deadlines, execution status and outcomes.
- Decision Map. A visual view of how decisions relate to the authorities that approved them, the meetings where they were taken, their dependencies and their implementation progress.
- Decision Assurance and Learning System. Independent challenge, periodic reviews, outcome measurement, and feedback into future decisions.
Together they let a board answer the question that matters most when a decision is questioned: who decided, on what evidence, and why — and what happened next.
Freeze the case at approval
One practical detail makes a large difference. When the outcome is reviewed, it must be compared with the case as it stood when the decision was approved, not with a version edited afterwards. Otherwise hindsight quietly rewrites the reasoning, and the organisation loses the ability to tell good judgement from good luck.
Where DecixionIQ fits
DecixionIQ keeps these records in one place: the Decision Brief, approval authorities, the decision register, a decision map, independent assurance reviews and outcome reviews. Approval history cannot be changed afterwards, and each outcome review is linked to a frozen copy of the case at approval. AI can help draft and challenge, but it never approves — only an authorised person can.
This is the last part of the series. Start again from Part 1, or see the governance guides for South African legal checklists.
Sources and further reading
- The IIA — The Three Lines Model (2020)
- Rogers & Blenko — Who Has the D?, Harvard Business Review (2006)
- Companies Act 2008, section 76 — Standards of directors' conduct
- IoDSA — King V Disclosure Framework
General information, not legal or professional advice.