Before you begin
- Before this: Chapter 7

Lesson 01 · Section 1 of 10 · Reading
Chapters 2 to 7 taught you to work with numbers. This one is about not being worked by them.
Chapters 2 to 7 taught you to work with numbers. This one is about not being worked by them.
Every figure you will meet in the forty modules — and in any real meeting — was produced by somebody, using a method, for a purpose. None of those is visible in the number. A learner who can compute an NPV perfectly and cannot ask where the cash flows came from has half a skill.
And the chapter ends by turning the reader into a producer. 8.12 builds an estimate for a question nobody has data for, in four steps, reproducibly.
| Subchapters | ||
|---|---|---|
| 1 | Who made it — the producer, the purpose, and what kind of object this is | 8.1 – 8.3 |
| 2 | How it was allowed to be made — accounting choice, earnings quality, adjustments, red flags | 8.4 – 8.7 |
| 3 | What to compare it against — benchmarks, base rates, your own record, false precision | 8.8 – 8.11 |
| 4 | How to make one yourself — build an estimate, four questions, when to distrust a model | 8.12 – 8.14 |
| 8.1 | Every number was made by someone | Numbers are produced, not found |
| 8.2 | Who made this, and what do they want? | The first question, and the direction of the lean |
| 8.3 | Facts, estimates, and forecasts | Three objects, one font |
| 8.4 | Where accounting leaves room for choice | Same reality, different reported number, both legal |
| 8.5 | Earnings quality in plain words | Is the profit real — i.e. did the cash arrive? |
| 8.6 | Adjusted, pro-forma and "underlying" | What gets added back, and by whom |
| 8.7 | Red flags you can spot without training | Eight of them, each with what it might mean |
| 8.8 | Benchmarks: compared to what? | A ratio with no comparison is decoration |
| 8.9 | Base rates: what usually happens | The outside view. What happened the last 22 times |
| 8.10 | Your own track record is evidence | Your last three plans predict your next one |
| 8.11 | Precision that isn't accuracy | Eight digits from four judgements |
| 8.12 | Building an estimate from nothing | Decompose · bound · check. Reader becomes producer |
| 8.13 | Four questions to ask of any number | The pocket checklist |
| 8.14 | When to distrust a model — including your own | Six situations, and the two kinds of failure |
8.2 — who made this, and what do they want. The producer's purpose shapes every choice inside the number, and the direction of the lean is predictable before you know anything else. Not cynicism about people — realism about incentives, and it applies hardest to numbers you produced.
8.5 — earnings quality. One ratio answers is the profit real: operating cash divided by net profit, over three years. The worked case shows profit up 55% while operating cash fell 46%, with $1.9M of reported profit sitting in receivables — visible from published accounts in two minutes, and the whole prerequisite for fsa-03.
8.9 — base rates. NorthRange's launches hit their first-year target 7 times out of 22 — 32% — and nobody in the room knew that until somebody spent an afternoon in the sales system. It is the most predictive number available and it is never asked for.
8.12 — building an estimate from nothing. Four steps, twenty minutes, and it works on almost anything. The worked case ends with the sanity check disagreeing with the estimate, which is the useful outcome — a check that agrees confirms nothing you did not already believe.
Every figure is checkable on paper.
| Earnings quality, 3 years | conversion 120% → 81% → 42%, DSO 55 → 96 days |
| Cumulative profit vs cash | $7.7M against $5.8M — a $1.9M gap |
| Base rate | 7 of 22 launches — 32% |
| Track record | 87%, 83%, 89% — mean 86.3%; a $16M plan reads as $13.8M |
| The estimate | central $3.54M, bounded $1.11M–$10.75M, a factor of 9.7 |
| The sanity check | implies 11.9% share against an assumed 4–8% — it disagrees |
Seven terms, and four of them are the words a whole fpr module rests on:
plan (335 uses) · forecast (253) · budget (72) · actual (59) · peer (53) · accrual (50) · comparable (31)
plan, forecast, budget and actual are four different objects — 719 appearances in the modules between them, and half the confusion in fpr-03 comes from using the first three interchangeably.
About 1 hour 45 for all fourteen. There is no gate; the chapter's exercise is 8.12, which is meant to be done on a question from your own work.
The short route is 8.2 → 8.5 → 8.9 → 8.12, about 35 minutes, and those four carry most of the chapter.
Chapter 9 — Deciding in a Room Full of People. Chapter 8 is about the numbers; Chapter 9 is about everybody else in the room, and the two are independently readable in either order.