Inversion is solving a problem backward. Instead of asking how to reach an outcome, ask what would reliably prevent it, then avoid those things.

Why it works

The routes to success are many, uncertain and specific to circumstances. The routes to failure are fewer, more predictable, and largely the same across situations. A list of ways to ruin a business is short and reliable; a list of ways to build one is long and contested.

Avoiding known failures is also higher confidence work. You can be reasonably sure that running out of cash ends the business. You cannot be equally sure that any particular growth tactic will work.

Using it

On a plan. Before committing, ask what would guarantee this fails. Write the list. Most of it will be within your control, and the plan improves more from removing those items than from adding tactics.

On a decision. Ask what the worst version of this looks like and whether you could survive it. Decisions that cannot be recovered from deserve different treatment from decisions that can. See decision under uncertainty.

On a goal. "How do I get promoted" produces vague answers. "What would guarantee I am not promoted" produces a specific and uncomfortable list, and the items are usually things you already know.

On a system. Ask where it is fragile before asking where it could be optimized. Removing a failure mode is more dependable than adding an improvement, because it survives being wrong about your model.

The limit

Inversion protects and does not create. A business that has eliminated every failure mode and has no customers has avoided its way to nothing.

It is a filter applied to plans rather than a substitute for having one. The useful sequence is to decide what you are trying to do, then invert to find what would stop it, then proceed with those removed. See Poor Charlie's Almanack and second-order effects.