Start with the Posts

There's a communication framework called The Minto Pyramid Principle that business schools and consulting firms love to teach. Lead with the conclusion, place the supporting arguments beneath it, and arrange the evidence so that the reader can move downward from the answer into the reasoning without having to extract the recommendation from twenty pages of analysis.

The idea is sound. A surprising amount of business communication improves once someone is forced to say what they think before explaining how they came to think it.

But the metaphor stops too early.

A pyramid describes the argument once it's been assembled. It tells you what the audience should encounter first and how the supporting pieces should sit beneath it. It says less about how to determine whether those pieces are strong enough, whether the logic between them has gone slack, or whether one questionable claim is quietly pulling the rest of the structure out of shape.

For that, I think an argument looks more like a fence.

A fence works because the posts are sound and set deeply enough to hold. The wire, boards, or rails between them matter, but they can only carry what the posts support. You can replace a board, restring a section of wire, or reinforce a connection, but if a post begins to lean, the weakness travels outward until an entire span starts to sag. The posts establish the argument. Everything else has to span the distance between them.

Arguments behave similarly. The posts are the claims someone is willing to defend. The material between them is the evidence and connective logic that explains why one claim leads to the next. When the posts are firm, the argument can survive an imperfect chart or a question nobody anticipated. When they're weak, adding more material usually makes the problem harder to see.

Most communication failures that look like missing detail are actually missing posts. You've seen the presentation: twenty slides, clean charts, carefully managed transitions, and enough analysis to suggest that an answer must exist somewhere inside it. Then someone reaches the end and asks, "So what are we actually saying?"

The problem is rarely that the team has no information. More often, nobody has converted the information into a small number of claims clear enough to be challenged. There are findings, implications, considerations, and observations, all connected by language smooth enough to carry the audience from one slide to the next, but nothing has been set firmly enough that the room can push against it.^[1]

When the argument begins to sag, the instinct is to tighten the wire. Add another chart or introduce more context. Explain the methodology. Include a slide answering the objection someone raised in the pre-read. Each addition may be reasonable, but wire strung between weak posts doesn't strengthen them. It increases the tension on claims that were never set deeply enough to carry the weight, until the conclusion arrives so qualified that nobody is sure what's left.

Fences have to be walked. You move along the line one post at a time, testing where each claim sits and what it's expected to support. If this statement is true, does the next one follow? If this assumption changes, which section moves with it? Is the distance between two claims small enough for the evidence to span, or has the argument quietly asked the audience to cross a gap it never acknowledged?

Finance people use similar language when they talk about walking a model build. You step through the revenue assumption, the hiring plan, the margin logic, and the sensitivities in sequence, not because the finished model can't be summarized from above but because its weaknesses are easier to find up close. A good walk invites someone to press on each assumption and see what moves. A bad one becomes a tour of the finished output, with the guide moving quickly past anything that might not survive inspection.^[2]

The spacing matters, too. An argument with too many posts feels anxious, as though every sentence needs its own supporting claim and every objection must be resolved before the audience is allowed to continue. The line becomes crowded with qualifications, sub-arguments, and distinctions that may all be true but leave no room for the central idea to carry.

Too few posts create the opposite problem. The argument asks a broad conclusion to rest on claims spaced way too far apart, with the evidence drooping between them. The recommendation may sound clear from a distance, but anyone walking the line can find the place where the reasoning stops and the speaker's confidence takes over.

You can often tell how someone thinks by where they choose to set the posts. One person begins with market demand, another with internal capability, another with the cost of doing nothing. Those choices reveal which facts they consider structural and which they treat as supporting detail. Two people can agree on the conclusion and still build arguments that behave very differently under pressure because they've chosen different things to anchor them.

Arguments age. A claim that felt solid when it was first made may settle differently after time passes, the arrival of a new information, or a clear change in the assumption beneath it. The deck can remain exactly as it was, still polished and logically arranged, while one of its central posts has begun to lean. Walking the argument again is how you notice that the evidence no longer hangs the way it once did.

The pyramid still has its place. It tells you to respect the audience's time, state the conclusion, and organize what follows so that the reasoning can be understood. But the clean shape of the finished argument creates its own illusion. Once every point has found its place beneath the recommendation, the structure begins to look more inevitable than the thinking that produced it.

A fence makes a different demand. It asks you to commit to a line across uneven ground, deciding where the argument needs support, how much distance the evidence can carry, and which sections will move if one of the central claims gives way. When it fails, it fails in a specific place that can be found, tested, and repaired.

You can tell when an argument will hold by how it behaves when you walk it.