← Research
Research

The decision quality chain

A decision is only as good as its weakest link, and there are six of them.

Last reviewed: 26 September 2026 · Next review due: 26 September 2027

The practitioner framework from decision analysis: frame, alternatives, information, values, reasoning, commitment. Where machines help, where they cannot, and what it has never been tested on.

Question this page answersAll 996 questions this research covers

The decision quality chain says a decision is only as good as its weakest link, and it names six links so an organisation can find which one is failing. It comes out of decision analysis, it is a practitioner framework rather than a tested theory, and the most useful thing on this page for anybody who has to run a decision next week.

The answer, in one line

Six requirements that together make a decision sound, set out by Spetzler, Winter and Meyer in Decision Quality (2016) on Ronald Howard's decision analysis: an appropriate frame, creative alternatives, relevant and reliable information, clear values and trade-offs, sound reasoning, and commitment to action.

Share as a card

Definition#

The decision quality chain: six requirements that together make a decision sound, set out by Spetzler, Winter and Meyer in Decision Quality (2016) and built on Ronald Howard's decision analysis. An appropriate frame, creative alternatives, relevant and reliable information, clear values and trade-offs, sound reasoning, and commitment to action. Quality is set by the weakest of the six, hence the chain.

Share this definition as a card

Frame. What are we deciding, and what is out of scope. A model will answer the question it is given and will never tell you it is the wrong question, so the frame is now the least assisted and most consequential link in the chain.

Alternatives. What else could we do. Generation is the link machines genuinely help with, and the risk is the opposite of scarcity: a list of twelve plausible options that all sit inside the frame the model inferred.

Information. What do we know, how reliable is it, and what would we need to know to change our minds. Retrieval is cheap and provenance is not, so this estate keeps a decision provenance page at all.

Values and trade-offs. What are we willing to give up. This is the link nobody can delegate, because it is a statement about the organisation rather than about the world, and a model asked to supply it will produce a fluent average of what other organisations have said.

Reasoning. Does the conclusion follow. Machine-written reasoning reads better than most human reasoning and is not therefore sounder, which is the specific failure the research on verification keeps finding.

Commitment. Will the organisation actually act. A decision nobody executes was not a decision, and no tool changes that.

Why it is worth having despite the evidence gap#

There is no controlled trial showing that organisations using the chain make better decisions. It is a framework from practice, taught in executive education and used in oil, pharmaceutical and capital-project decisions where a single call is worth hundreds of millions. Treat it as a checklist that makes a failure visible rather than as a finding.

What earns its place here is the weakest-link property. It stops the common move of improving the link that is easiest to improve. Most AI investment lands on information and alternatives, which are already the two strongest links in most organisations, and leaves frame and values untouched, which are usually the two weakest.

How it relates to the protocol on this site#

The Decision Quality Protocol is narrower and answers a different question. The chain asks whether a decision was well made. The protocol asks who owns it, how the machine takes part, and what quality control looks like, which is a governance question rather than an analytical one. A board can adopt the protocol in an afternoon; the chain takes a facilitator and a real decision.

Key sources

Essay · SS-2026-313

Cite this page

Hirji, R. (2026). The decision quality chain. The SuperSkills evidence base, SS-2026-313. https://thesuperskills.com/research/the-decision-quality-chain. Last reviewed 26 September 2026.

An evidence review by Rahim Hirji, not peer-reviewed research. For a material claim, cite the underlying study as well; every study here carries its own permanent link.

How citations and IDs work
Questions answered on this page

What is the decision quality chain?

Six requirements that together make a decision sound, set out by Spetzler, Winter and Meyer in Decision Quality (2016) on Ronald Howard's decision analysis: an appropriate frame, creative alternatives, relevant and reliable information, clear values and trade-offs, sound reasoning, and commitment to action. Quality is set by the weakest of the six, hence the chain.

Which link does AI help with?

Alternatives and information, mainly. Generation and retrieval are the two things machines do cheaply. The risk is not scarcity but volume: twelve plausible options that all sit inside a frame the model inferred rather than one the organisation chose.

Which link can a machine not supply?

Values and trade-offs. What an organisation is willing to give up is a statement about the organisation, not about the world, and a model asked to supply it produces a fluent average of what other organisations have said.

Is the decision quality chain evidence-based?

It is a practitioner framework rather than a tested theory. There is no controlled trial showing that organisations using it make better decisions. It is taught and used where a single decision is worth hundreds of millions, and it earns its place through the weakest-link property rather than through trial evidence.

How does it differ from the Decision Quality Protocol?

The chain asks whether a decision was well made, which is an analytical question. The protocol asks how each class of decision is made with AI in the loop, who owns it by name and what quality control looks like, which is a governance question. A board can adopt the protocol in an afternoon; the chain needs a facilitator and a live decision.

In this hub

Definitions

The terms this field uses, defined against their primary sources.

Ask the evidence
What does the evidence actually show?What should our board be asking about this?Where does Rahim disagree with the consensus?
Bring this into your organisation

If this describes something happening in your teams, say so.

Keynotes, board sessions and advisory work, drawing on research across more than 200 organisations in 30 countries. Tell me the room, the date and the shift you need. A reply within 24 hours.

Start a conversation

Topics and audiences  ·  All research

Box of Amazing

Rahim’s free weekly letter on AI and human capability

If this was useful, the weekly letter is where the thinking happens first. Most of what ends up on this site starts there. Weekly essays on AI, capability and the future of work. Read by 25,000 people, every week since 2017. Free, and one click to stop.

Opens Substack to confirm. No pitch in it, unsubscribe in one click, and nobody follows up because you read something.

Running an event, or responsible for how AI arrives in your organisation? Keynotes  ·  Advisory  ·  Boards  ·  Enquire