Judgment Playbook

Decision Quality: six links

Treat decision quality as a chain whose weakest link limits the whole recommendation.

5 minute readtaught in course chapter 1
FramingEvidenceAlternativesUncertaintyCommitment

What it is

Decision Quality treats a recommendation as a chain of six links: an appropriate frame, clear values and trade-offs, creative alternatives, reliable information, sound reasoning, and commitment to action. The quality of the decision is capped by the weakest of the six, the way a chain is capped by its weakest link. So the method is a diagnostic rather than a template. You find the link that limits the decision and spend your remaining effort there instead of polishing the part you have already done well.

Where it comes from

The discipline behind it is decision analysis, named by Ronald A. Howard at Stanford in the 1960s and set out in his 1966 paper "Decision Analysis: Applied Decision Theory."

The six-link chain itself has no single author. It was worked out in consulting practice at Strategic Decisions Group, the firm Carl Spetzler and Howard founded in 1981, and SDG describes the method today as the product of more than 35 years of application inside large organisations. The fullest written form is the 2016 Wiley book Decision Quality: Value Creation from Better Business Decisions, by Carl Spetzler, Hannah Winter and Jennifer Meyer. The Decision Education Foundation teaches the same six elements to school students under the name "the decision chain." Wording differs between these sources: you will see "framing" for frame, and "useful information" for reliable information. The six links do not differ.

What it corrects

The situation is a review of work that is good. An analyst brings a financial model, a market estimate with ranges, and a clear preferred option. The reviewer judges quality by the depth of what is on the page: the model is checked, the estimate is stress-tested, the writing is tightened.

What the competent reviewer does wrong is to audit only the links that are present. The two links that fail most often leave no page to inspect. A missing alternative shows up as a confident comparison between one plan and doing nothing. An unnamed objective shows up as the word "best," which quietly assumes everyone wants the same thing. Ordinary care does not fix this, because care is applied per link. Reading the model twice makes the strongest link stronger and never reveals the option nobody drafted.

How it works

  1. State the frame: who decides, what choice is open, what is out of bounds, over what horizon.
  2. Write the objective and the trade-off, in the form "more of X while protecting Y," rather than the word "best."
  3. List alternatives that are genuinely different in kind, not three versions of the same plan.
  4. Separate what is known from what is assumed, and attach a range to each assumption that matters.
  5. Connect facts to the recommendation with reasoning someone could contradict.
  6. Name the action, the owner, the first commitment of money or people, and the date you review it.
  7. Score the six links quickly, then improve the lowest one. Stop when further work on it would cost more than the decision stands to gain.

Worked example

On the evening of 27 January 1986, engineers at Morton Thiokol recommended against launching the Space Shuttle Challenger. Their reason was temperature: the booster joint seals had never been flown below 53°F, and the next morning at Kennedy Space Center was forecast near freezing. NASA's George Hardy said he was "appalled" at the recommendation. Thiokol's managers caucused for about thirty minutes and reversed their own engineers. Challenger was launched on 28 January and broke up 73 seconds later.

Read as a chain, the information link held: Thiokol had flight data on seal behaviour in cold, and its engineers put it on the table. The link that failed was the frame, specifically the question being asked. Roger Boisjoly told the Presidential Commission that June: "This was a meeting where the determination was to launch, and it was up to us to prove beyond a shadow of a doubt that it was not safe to do so. This is in total reverse to what the position usually is." Flight readiness normally requires proof that the vehicle is ready. That night it required proof of failure, which no one could supply. Every other link in the chain was working, and none of them could lift the decision above the one that had broken.

In a Business Case Weekly case

In the Booking.com case, Geert-Jan Bruinsma faces a fork in mid-2002: sell most of his shares to a Dutch investor group, hold and hire a chief executive, or hold and run the company himself. Growth data is the easy link here, and it is not the binding one. The six links push you to say whose decision this is, since the founder's objective and the company's are not the same thing, and to write the trade-off between a settled family and an uncertain claim on European travel. Only then does the list of alternatives look short or complete.

In your answer

  • "The decision-maker here is …, deciding … before …; out of scope is …"
  • "The objective is … while protecting …, and I am willing to give up … to get it."
  • "The weakest link in this recommendation is …, because …, so more work belongs there rather than on the forecast."
  • "I commit to …, funded by …, and I will reverse it if … by …"

Common misuse

The common misuse is the six links as six headings. A paragraph appears under each label, every link is described as adequate, and the recommendation ends exactly where it started. That version does none of the method's work, because the point is not coverage but a ranking. One-line test: if your answer never names one link as the weakest, and nothing in the recommendation changes as a result, you have written a table of contents rather than run the diagnostic.

References

Falsifiability

Next method

Turn a confident claim into one that names the evidence capable of proving it wrong.

Open next tool