structured-thinking · part 1 of 1
The Answer Goes First
Part 1 of a series on structured problem-solving and senior-level communication.
Reading time: about 5 minutes.
Notice how a senior person's status update reads sharper than yours. Same facts, same week, same amount of work done. The difference isn't better writing. It's that their conclusion is the first sentence and yours is somewhere in paragraph three.
You can hand a draft to ChatGPT or Claude and ask it to tighten the prose, fix the grammar, make the sentences shorter. That's real, and worth doing. But it won't fix the order. An AI can polish a paragraph. It won't tell you that your actual point is buried under three paragraphs of setup, because the setup isn't wrong, it's just first. Fixing that is a structural edit, not a language edit, and it's a different skill from prompting well.
What This Article Answers
- What is the Pyramid Principle, in one sentence? (State the conclusion first, then group your support underneath it.)
- Does this only apply to formal reports and slide decks?
- How is this different from just "being concise"?
- What does a pyramid-shaped email or slide actually look like?
- Why do so many status updates end up chronological instead?
- What's the fastest way to fix a document that's already written the wrong way round?
TL;DR
The Pyramid Principle says every piece of business writing should open with its conclusion, then supply the reasoning underneath it, each layer supporting the one above. Barbara Minto developed it at McKinsey in the 1970s; it survives because readers scan rather than read, so anything placed after the answer competes for attention that usually isn't there.
- State the conclusion first, always. Not a teaser. The reader should be able to stop after sentence one and still know what you think.
- Everything below the answer exists to support it, not build toward it. If a paragraph doesn't defend the top line, cut it.
- Group the supporting points, don't just list them. Three reasons that form one argument beat three unrelated facts.
- Chronological order is the default trap. "First we did X, then Y, then Z" is how you experienced the work, not how the reader should receive the conclusion.
- This is a rewriting discipline, not a talent. You're reorganizing a finished argument, not writing brilliantly on the first pass.
The shape that gives the technique its name: one answer at the top, a small number of groups supporting it directly, evidence underneath each group. Read top to bottom and every line still makes sense on its own.
The Shape, in One Rule
Barbara Minto spent two decades editing McKinsey reports before they went to a client, and published the technique in 1985 as The Pyramid Principle. The firm still teaches it to every new hire. That pedigree matters less than the mechanism itself: the top of the pyramid has to be a summary of what's below it, because it's derived from what's below it, not asserted on its own. One governing idea at the top, a small number of reasons directly underneath it, evidence under each reason. If you can delete the top line and the paragraphs underneath still make the same argument without it, you built a list, not a pyramid.
The Email
Written the way it usually gets written: "This week we finished the ingestion pipeline, ran it against the test dataset, found three edge cases in date parsing, fixed two of them, and are blocked on the third pending a decision about how to handle partial records. Testing continues next week." Five facts, in order. Accurate, and hard to act on: the reader has to hold all five in memory and do the synthesis themselves before they know what's being asked of them.
Rebuilt as a pyramid: "We need a decision on how to handle partial records by Wednesday, or testing stalls next week. Ingestion is otherwise on track: built, tested against real data, two of three edge cases already fixed. The remaining one is the partial-records question above." Same five facts. Only the order moved, and the effect isn't cosmetic: the first version makes the reader extract the ask, the second one hands it to them.
Same five facts, same word count, two different structures. Only the second one tells the reader what to do without them doing the synthesis themselves.
The Slide Deck
A deck that walks findings in the order they were discovered is a chronological pyramid built upside down: slide 1 is background, slide 2 is methodology, and the actual recommendation shows up on slide 11, if the reader is still there. Rebuilt, slide 1 states the recommendation and the decision it requires. Slides 2 through 4 are the reasons, each one a distinct argument, not a restatement. Slide 5 onward is the backup data, there for whoever wants to dig, not required reading to reach the point.
The same failure shows up live, in the room. A stakeholder meeting defaults to whatever order the discussion happens to take, and the risk that actually matters gets raised last, if at all, the same pattern covered in architecture review meetings.
The recommendation moves from slide 11 to slide 1. Nothing else about the deck has to change.
Why "Just Be Concise" Isn't the Same Advice
A document can be short and still bury the lede, if every sentence is brief but the conclusion is the last one. Conciseness is word count. The Pyramid Principle is sequence, and sequence is the part that gets skipped, because reordering a finished argument feels like extra work after the thinking is done. Nielsen Norman Group's eye-tracking research found readers absorb on average only 20 to 28 percent of the words on a page, scanning in an F-shaped pattern that thins out fast after the first few lines. Anything placed at paragraph four is competing for attention that mostly isn't there anymore.
Rewrite the Next Thing You Send
Pick whatever you're about to send today, a status update, a Slack message, the top of a deck. Write the actual conclusion as its own sentence, separate from everything else, and check if it's currently the first sentence in the document. If it isn't, move it. Then check whether each remaining paragraph is a reason to believe that conclusion, or just something that happened to be true this week. Cut or demote anything that isn't doing the first job.
That's the whole technique. It costs two minutes on something you already wrote, and those two minutes buy the reader the entire document without them hunting for the part that mattered. Sometimes the rewrite exposes something sharper: a document with no clear top line often means the decision behind it hasn't actually been made yet, not just poorly written.
A conclusion buried in paragraph four isn't wrong, it's invisible. The reader can only act on what they actually reach.
The next time you write an email, remember the reader isn't going to read all of it. Put the point first, or it may as well not be there.
Coming Up in This Series
Putting the answer at the top only works if the groups underneath it are actually sound: each supporting point distinct, and together covering the whole argument with nothing missing. That's a separate skill from sequencing, and it's where most pyramids that look right on the surface fall apart under a second read. Next: MECE thinking, and the test for whether your buckets actually hold up.
This is part of a series on structured problem-solving and senior-level communication.