Charts That Sell · Field notes from the pitch room

Answer first: the pyramid principle, translated for people who pitch

Field note #7 · 7 min read

Most pitch decks are written in the order the analysis happened. "First we looked at the market, then we examined your data, here's a chart of that, and another chart, then we did interviews…" Twenty minutes in, the presenter reaches slide 22 and the recommendation finally appears — to a room that stopped listening at slide 8.

Barbara Minto gave the fix in 1973 while running recruiting at McKinsey: the Pyramid Principle. It's the single most valuable writing rule ever handed to a consultant, and almost nobody outside the big firms applies it to sales decks. Here's the whole thing translated for pitch work.

The shape: one answer, three arguments, data at the base

A pyramid has three levels. Build yours explicitly:

• The apex — the governing answer. One sentence: what should the client do, and why. "You should adopt our solution in the South region this quarter, because the gap there is large, growing, and specifically addressable by what we do."
• The middle — the supporting arguments, usually three, MECE: mutually exclusive, collectively exhaustive. The gap is large. The gap is growing. We can close it. Each one is a single sentence a skeptical CFO could attack independently.
• The base — the evidence. Numbers, charts, quotes, case studies. This is where every chart in your deck lives: directly underneath the argument it proves, and nowhere else.

The executive reads only the apex. The skeptic reads an argument and demands proof. The auditor reads the base. The pyramid serves all three in the same deck.

Answer first — literally first

The hardest habit for an analyst: the recommendation is slide 2, not slide 22. After the title and one page of situation, state the answer. Not "we recommend a strategic initiative" — the actual answer, with the three arguments previewed in a sentence each. Executives don't resent the spoiler; they paid for the answer. Giving it first grants them something rare in their calendar — a meeting where they know within three minutes whether they agree, and can spend the remaining twenty-seven testing why.

The instinct to withhold comes from essay-writing school: build suspense, earn the conclusion. A pitch is not an essay. Your reader is a decision-maker who has read two hundred analyses this quarter; suspense reads as disorganization.

Why three, and why in that order

Three arguments isn't superstition. Two looks thin and coincidental; four nobody remembers. Three forms a stable structure — and it matches how claims get challenged. Order them strongest-first when trust is low (lead with the argument most likely to be accepted); order them logically when the room is analytical (the conclusion of argument one feeds the premise of argument two).

The MECE test is where sloppy pitches die. "The gap is large / the gap is growing / the gap is important" is not three arguments — it's one argument wearing three hats, and a sharp client will say so. Three genuinely distinct angles: magnitude, trajectory, solvability. Each gets its own chart or chart pair.

Where the charts go, and where they never do

This is the part the previous notes have been circling: choose the chart by the conclusion, then make each slide prove one claim — because in a pyramid deck a chart is not information, it is cross-examination material. You put a claim on the table; the room's skeptic opens the chart to see if it holds.

• No chart at the apex. The opening states the answer in words; an opening chart forces the client to do the synthesis you were hired for.
• One chart per argument as the default. Two only when the argument genuinely needs two views (the trend and the attribution — which is the waterfall behind field note #1's example).
• Every chart must map upward. Point at any slide in the deck and ask: "Which of the three arguments does this prove?" If the answer is "context" or "interesting" — cut it, move it to the appendix, or the appendix is where your sale goes to die anyway so really: cut it.

Building it backwards, presenting it forwards

Minto's trick for actually writing one: build the pyramid from the base up, present it from the top down. You already have the data — that's all the Friday-night cleanup from note #5. Group the evidence into themes; three themes emerge; look at them and write the one sentence they force. That sentence is your apex. If the data doesn't force a clean sentence, you don't have a weak deck — you have a weak recommendation, and finding that out on Friday night is the whole point.

Then present apex-first: answer, three arguments, and walk down only as far as the room demands. The base exists so you can answer questions, not so you must read it aloud.

— Written between pitches. Field note #7 of Charts That Sell. Last note: the four beats your opening should land — SCQA.