# The Structure > From *The Decision Product* (Chapter 13, Writing the Decision). Free to use, > adapt, and redistribute inside your organization. > > The Write-up worksheet tells you **what sections a decision note contains**. > This one tells you **what order the argument goes in and how to tell whether > it holds**. Use them together. Use this one first if the sections are already > full and the note still does not land. > > Work through it with the document closed. Everything here is done on one > sheet of paper before you write prose. --- **Document:** **Reader who can commit:** **Date:** --- ## 1. The governing thought One sentence at the top of the document. The pick, not the topic. > We recommend ______________________________________________________, > > for ______________________________________ (scope), > > starting ______________________________ (when). Two checks before you go on. ☐ It names an action, not a finding. "Demand is more variable in the 8 percent" is a finding. "Adopt a three-class stocking rule at the April freeze" is a pick. ☐ It provokes an immediate and predictable *why*. If a reader can read it and have no question, it is either obvious or empty. Either way it is not worth a document. --- ## 2. The opening that makes the pick land Not a preamble. This is the governing thought written out backwards, so that the reader arrives at your question already wanting the answer. **The situation they already accept** (uncontroversial, theirs not yours): _____________________________________________________________________ **What changed, and what it has cost** (the disturbance, in consequence): _____________________________________________________________________ **The question that creates** (write it as a literal question): _____________________________________________________________________ **The answer** — which must be, word for word, your governing thought above. ☐ The question in the opening and the recommendation in section 1 are the same question and its answer. If they are not, one of them is wrong. ☐ Nothing in the opening is contested. The opening is where the reader nods. Disagreement belongs downstream, where you have evidence for it. *Familiar name: this is what Situation–Complication–Question–Answer (SCQA) and Situation–Complication–Resolution (SCR) describe. You do not need the acronym and the reader must never see it. Fill the four lines.* --- ## 3. The reasons Two to four. Each must be able to change the ranking if it turns out to be false. If a reason cannot do that, it is context, and context goes below. 1. ______________________________________________________________ 2. ______________________________________________________________ 3. ______________________________________________________________ 4. ______________________________________________________________ ### The removal test — ten seconds, and most people skip it Read the reasons and **delete the second one**. Then answer honestly: ☐ **The argument still stands.** You have a *parallel group*. Each reason supports the recommendation on its own. The case survives losing one, in weakened form. This is what most decision write-ups should be, because the ordinary fate of evidence is that one of your three reasons turns out to be shakier than you thought. ☐ **The argument collapses.** You have a *chain*: each step depends on the one before, and the recommendation is the last link. A chain is more persuasive than a group when it holds, because it reads as a proof rather than a pile. It is also brittle in a way people underestimate. An objection to any link takes down everything after it, and a reader who rejects step two has no reason to read three and four. If you have a chain, do one of two things and write down which: ☐ Say so plainly in the document: "the case rests on ____________________." ☐ Go find independent support so that it does not. The failure worth guarding against is writing a chain while believing you wrote a group, and then being surprised when one challenged premise takes the whole recommendation down. ### Are these the same kind of thing? Reasons in a group have to be summarizable, which means they have to be the same kind of idea. A group containing "margin is higher", "the plant in Toledo disagrees", and "we already have the data" is three different kinds of claim wearing one label. ☐ I can write one sentence that is a genuine summary of all of these, not just a heading that sits above the pile. Write it: _____________________________________________________________________ ☐ No two reasons are the same reason in different words. ☐ Nothing that could change the ranking has been left out. *Familiar name: the second and third boxes are what MECE — mutually exclusive, collectively exhaustive — is asking for. The first box is the harder test and the one the acronym does not capture. A set can be perfectly MECE and still be a pile, if no single thought summarizes it.* --- ## 4. The order Parallel reasons are arranged deliberately or they are arranged by accident. Choose one and write down why. ☐ **Importance.** The reader is deciding whether to keep reading. This is most readers, most of the time, and it is the default for a decision note. ☐ **Time.** The argument is a sequence of events and the order is the point. ☐ **Structure.** By region, item class, team, or stage. Suits a reader who is auditing coverage rather than weighing a case. Chosen because: _____________________________________________________ ☐ The order is **not** the order the reasons occurred to me. That is the one arrangement guaranteed to carry no information for the reader. --- ## 5. Headings that carry the argument Write these before the prose. Full sentences. Each one an assertion. | # | Heading as a claim | |---|---| | 1 | | | 2 | | | 3 | | | 4 | | A filing label names a topic. A claim states something that could be false. > "Benchmark results" is a filing label. > "The misses are concentrated in 8 percent of items, and those items strand > repair jobs" is a claim. The first transfers the analytical work to the reader. The second does the work for them, which is what they are paying for. ### The heading-only read Read the governing thought, then the headings in sequence, ignoring the body entirely. ☐ They make the case on their own. The structure holds. ☐ They read as a table of contents. The structure is decorative. Rearrange until they make the case, and only then write underneath them. Senior readers do this read whether or not you designed for it. Designing for it means they get your argument rather than a guess at it. --- ## 6. The layers A one-page version and a full note are the same argument at two depths, not two audiences receiving different amounts of truth. | Layer | Contains | |---|---| | One page | Choice, alternatives, recommendation, reasons, flip conditions, rule | | Full note | The above, plus economics and the evidence trail | | Behind both | Methods, code, formulations, reproducibility | ☐ **If the ranking in the appendix disagrees with the recommendation on page one, the write-up is wrong.** Layers must be consistent, not merely shorter. --- ## 7. What to cut Everything that was scaffolding for you rather than support for the reader. ☐ The exploratory cuts that went nowhere ☐ The method you tried first ☐ The chronology of the project ☐ Any chart that is present because it was made rather than because it argues for something Sunk production cost is not an argument for inclusion, and it is the most common one actually operating. --- ## Before you send ☐ The pick is in the first paragraph, not the last ☐ The opening's question is answered by the governing thought ☐ Each reason could change the ranking if false ☐ I ran the removal test and know whether I have a group or a chain ☐ One sentence genuinely summarizes the reasons ☐ The order was chosen, and I can say why ☐ Every heading is an assertion, and the headings alone make the case ☐ The layers agree with each other ☐ No framework name appears anywhere in the document --- Do not put a framework name on the page. Fill the page. --- The Decision Product · thedecisionproduct.com