Check units and scales for consistency

Catch order-of-magnitude errors by ensuring your units are consistent across the estimate.

Why it works

Most large estimation errors are unit or scale errors: confusing thousands with millions, minutes with hours, or individuals with households. Units act as an error-detection system — when the units of the product do not match the units of the answer, something is wrong. Tracking units explicitly throughout the decomposition catches these errors before they produce a wildly wrong answer.

How to do it

1. Label every sub-estimate with its units (people/household, minutes/day, dollars/unit).
2. Verify that the units of each product combine to give the units of the quantity you want.
3. If units do not work out, find the step where they went wrong.
4. As a sanity check, compute whether the order of magnitude seems consistent with an everyday anchor.

Evidence

Dimensional analysis is a foundational technique in physics and engineering for catching errors in calculations. Its application to estimation is mechanistically straightforward and widely taught in quantitative reasoning curricula. (mechanistic)

Unit-checking catches scale errors but not errors in the values themselves; correct units with wrong numbers still produces a wrong answer.

Common mistake

Carrying units in your head rather than writing them down — unit errors are almost universally caught by writing them and almost universally missed when they are implicit.