Most research deliverables are written in the order the work was done. Methodology, then market background, then findings, then, somewhere around page nine, the point. The analyst experiences this as thoroughness. The reader experiences it as a document that hides its conclusion, and readers with authority do not go looking; they skim, form their own version of the point, and act on that instead of yours.
Barbara Minto solved this problem at McKinsey in the 1970s, and her solution, the Pyramid Principle with its SCQA opening, became the writing operating system of the entire consulting industry. This piece applies both tools to research and CI deliverables specifically, and it starts with a distinction that saves a lot of confusion.
The CI Report Pyramid and Minto's pyramid are different tools that happen to share a shape. The CI Report Pyramid governs the thinking: whether your work actually climbs from data through insight and implication to a recommended action. Minto governs the telling: the order in which a finished argument is delivered to a reader. You climb your pyramid to build the finding. The reader descends Minto's: conclusion first, then the support, layer by layer, only as deep as they need. Most bad deliverables are not badly thought; they are well-thought work delivered in excavation order rather than reader order.
The worked thread for this piece is one deliverable, rebuilt. The setup: you are the research lead at a European maker of industrial valves. A Chinese competitor, call it Jianhe Flow Control, has spent 18 months quietly winning European certifications, and your team has produced a 22-page assessment. The before version opens like this:
"This report examines recent developments in the European industrial valve market. Section 1 reviews market structure and sizing methodology. Section 2 profiles Jianhe Flow Control. Section 3 assesses certification activity. Findings are presented in Section 4."
Accurate, professional, and dead. Twenty-two pages of table stakes before a stake. Rebuild it with SCQA as the spine.
S: Situation. Start where the reader already nods
The Situation states what the reader knows and agrees with, in one or two sentences. Its job is not to inform; it is to establish shared ground so the reader's first reaction is a nod, not a question. The test: if a reasonable reader could dispute your Situation, it is not a Situation yet.
Research deliverables usually fail here by overloading. Fifteen paragraphs of market context is not a Situation; it is the analyst proving they did the reading. The reader who commissioned a competitive assessment does not need the market explained. They need to be met where they stand.
The rebuild: "We hold roughly 30 percent of the European industrial valve market, and our pricing has assumed that Asian entrants remain locked out of certified, safety-critical segments."
One sentence. The reader nods, and notice what the sentence quietly does: it states the assumption that is about to break.
C: Complication. The thing that changed
The Complication is why this document exists now. Something moved: a signal fired, a trend crossed a threshold, an assumption stopped holding. In CI terms, the Complication is your signal cluster, and if you triaged properly, you already know what it is; it is whatever earned this deliverable its place on the calendar.
The discipline here is honesty of scale. A Complication oversold ("the market is being disrupted") burns credibility; one undersold gets the deliverable filed. State what happened, dated and labeled, at the size the evidence supports.
The rebuild: "Over the past 18 months, Jianhe Flow Control has obtained the three certifications that gate our most profitable segments, hired a former certification-body engineer to lead its Rotterdam office, and quoted 30 to 40 percent below us in two tenders we can verify."
Three facts, checkable, no adjectives doing the work.
Q: Question. The pivot that disciplines the whole document
The Complication forces a question in the reader's mind, and the Question makes it explicit. This is the least glamorous element of SCQA and the most structurally important, because of one rule: the Question must be the question your deliverable actually answers. If your document answers "how exposed are we and what should we do," but your opening implies "who is Jianhe," the reader finishes the wrong document.
This is the telling-side twin of a thinking-side rule from market sizing: the definition is the sizing. On the writing side, the Question is the deliverable. Getting it wrong misaligns every page that follows. In practice, writing the Q is a test of the work itself: if you cannot state the single question the document answers, the document answers several, poorly, and the problem is not the writing.
The rebuild: "How much of our certified-segment revenue is exposed, how fast, and what do we do before the next tender cycle?"
A: Answer. The top of your pyramid, stated first
Now the move that research culture resists: give the answer, immediately, before the evidence. The Answer is Minto's "governing thought," and for a CI deliverable it is simply the top of your CI Report Pyramid, the implication and action you climbed to, restated as the document's opening claim, with its Fact vs. Inference Ladder label attached.
The rebuild: "We assess that roughly a fifth of our European revenue becomes contestable within 24 months, [CI] corroborated inference from the certification, hiring, and tender evidence. We recommend defending the top 15 certified accounts with multi-year service contracts this quarter, and re-testing this assessment in 90 days against Jianhe's tender activity."
Analysts resist answer-first for an honest reason: it feels unearned, a verdict before the trial. Two replies. First, executives read openings and skim bodies; answer-first is not a style preference, it is meeting the actual reading behavior of the people who decide. Second, the Ladder label is what makes answer-first honest. A conclusion stated up front with its confidence level declared is more truthful than a conclusion smuggled in on page nine after the reader's attention has left. Answer-first without labels is arrogance. Answer-first with labels is a briefing.
After the A: the body is Minto's pyramid, descended
SCQA gets the reader to the governing thought in half a page. The Pyramid Principle then structures everything below it, and three of Minto's rules carry directly into research work.
Vertical logic: every statement provokes the question the level below answers. The Answer above makes the reader ask "why contestable, and why a fifth?" The next layer answers with the three key-line claims: certifications remove the formal barrier; the tender evidence shows an executable price gap; the exposed accounts can be named and sized. Each of those provokes its own "how do you know," which the layer below answers with the labeled evidence. If a section does not answer the question raised by the statement above it, the section is misplaced or padding. This dialogue test, statement, reader's question, answer, is the fastest structural edit that exists.
Horizontal logic: groupings must be MECE and ordered by logic, not chronology. Ideas sitting side by side must be the same kind of thing, not overlap, and together cover the claim above them. Research drafts violate this constantly by grouping findings in discovery order. The reader does not care what you found first; they care what supports what.
Headings are claims. Every heading in the body is a full assertion someone could dispute, never a topic label. This is the action-title rule from slide work, applied to prose: a reader who reads only your headings should receive the entire argument. "Certification activity" is a filing label. "Certification is complete for our three most profitable segments" is a sentence doing its job.
Notice the symmetry that makes the two pyramids complementary rather than redundant. Building the work, you moved up: data, insight, implication, action. Writing the work, the reader moves down: action and implication first, insights as the key line, data at the base, available on demand. Same pyramid, opposite directions of travel, and the Ladder labels ride along at every level in both.
Variants and the one-page test
Two practical notes. For standing deliverables, weekly signal briefs, escalation notes, invert the order: lead with the Complication, one line of Situation only if the reader needs re-anchoring, then the Answer. Recurring readers do not need the ground re-established. And for any deliverable of any length, the full SCQA belongs on one page or one slide. If the opening needs two pages, the Question has not been found yet.
The complete system, in one line each: the CI Report Pyramid tells you whether you have an argument. The Ladder tells you how honestly it is labeled. SCQA and Minto tell you the order in which a busy, skeptical, powerful reader should meet it. The thinking earns the conclusion. The telling starts with it.