# The Complexity Ladder

> From *The Decision Product* (Chapter 8). A one-page check before you build.
>
> Work from the bottom. Climb only when a lower rung cannot settle a
> load-bearing claim.

---

| Rung | What it answers | Climb past it only when |
|---|---|---|
| **A count, a slice, a query** | Is the damage concentrated? Is the current rule even being followed? | A concentration question is settled and a *mechanism* claim is still open. |
| **A back-of-envelope economic model** | What does a miss cost? What does a unit of inventory cost? Are those numbers in the same universe? | The two numbers are close enough that the ranking actually turns on precision. |
| **A simple statistical relationship** | Is there something stable enough to use, and does it change an alternative? | The relationship is real but you cannot act on it without intervening. |
| **An experiment** | Does the intervention work when you actually do it? | You cannot intervene, or the dynamics rather than the effect are what matter. |
| **A forecast** | What is the input to the rule? | Never as the product. A forecast that does not change an order is a diagnostic with a press release. |
| **A simulation** | What happens when queues, bullwhips, and constraints bite in combination? | Averages are adequate, or you wanted something impressive for slide nineteen. |
| **An optimization model** | How should scarce resources be allocated under real constraints? | A heuristic would leave value on the table *in a way you can estimate*. |

---

## The rent

Each rung costs build time, review time, and, most expensively, **legibility**.

A planner who does not understand the rule will route around it. A model nobody
can follow has a lower effective value than a worse model that gets used.

**The test for climbing:** name the claim the current rung cannot settle.

If you cannot name one, you are climbing because the higher rung is more
interesting to build. That is an honest motive and a bad reason.

---

## Heuristics are first-class

A heuristic that a planner will actually follow, that degrades gracefully when
an input is missing, and that can be explained in one sentence is frequently
the better decision even where a solver would produce a tighter number.

The comparison that matters is not model against model. It is the realized
outcome of a rule people run against the realized outcome of a rule people work
around.

---

## Get the answer a second way

Whatever rung you land on, reach the answer once more by a different route,
and cheaply.

If the optimization says class-based stocking saves $1.2M, check whether a
back-of-envelope on the protected class lands in the same order of magnitude.
When the two disagree, one of them is wrong, and you have found it before the
meeting rather than during it.

This is the highest-return twenty minutes in most projects. It is skipped
almost universally, because the first answer is already in hand and checking it
feels like doubting yourself.

---

The Decision Product  ·  thedecisionproduct.com