The pyramid principle

One answer, three reasons.

Lead with the decision. Hold it up with three reasons that do not overlap. Put your numbers under each one.

A triangle in three tiers with the narrow top tier filled solid
Scenario

Thursday, 10:00. Fifteen minutes at the engineering review: why the team is shipping late, and the ask to fix it.

Sixty seconds, tier by tier

Five sentences, in the order they leave your mouth. Everything else you brought answers a question nobody asked.

Why these three

01 Why change?
02 Why this?
03 Why believe you?

The answer Level one

Governing thought

Put two engineers on the test suite for six weeks and quarantine any test that fails twice on main. Median merge time is back under eight hours by the end of Q3.

Then say “there are three reasons” and stop.


Finding or answer

“Merge time went from six hours to three days” is a finding: the room nods, nothing happens. An answer says who does what, by when, for how much.


Steal the skeleton

The same five lines, waiting for your numbers.

Setup

We fact nobody argues with. Since when, what changed: what it cost. So the question is the one your answer answers.

Answer

Who does what for how long, and the metric hits number by date. Three reasons.

01

Why change: where the cost actually goes, and the number nobody has seen.

02

Why this: the cause, narrow enough to name, and the number that kills the obvious alternative.

03

Why believe you: where it already worked here, and the number from your own org.

Sort the numbers you already have into three piles. The answer is whatever they add up to, price included.

Where it fails

When nobody has agreed there is a problem. Lead with the answer and you defend a fix to a complication the room never accepted; agree the complication first.

Minto calls the setup’s three sentences the SCQ Framework: barbaraminto.com.

If your first sentence could be true and change nothing, it is not an answer.