Inversion
Most planning asks how to make something succeed. Inversion turns the question around: what would make it fail for certain? The second list is usually easier to write, more concrete and more honest — and every item on it can be turned into a check.
The idea
The technique is old. The 19th-century mathematician Carl Jacobi advised his students to “invert, always invert”: when a problem is hard to attack directly, restate it in reverse. In planning, the reverse question works for a simple reason. Success is vague; it can take many shapes. Failure is specific. “Ship a good product” gives you nothing to check. “Nobody tested the upgrade path” does.
People are also reliably optimistic when they plan forward. Asking directly for failure gives everyone permission to say the uncomfortable things that a forward plan politely leaves out.
Failure Mode and Effects Analysis
Reliability engineers turned inversion into a formal method in the late 1940s. Failure Mode and Effects Analysis (FMEA) is now routine in aerospace, automotive and medical-device design. A team goes through a product part by part and, for every way each part could fail, asks three questions: how severe would the effect be, how often is it likely to happen, and how likely is it to be caught before it does harm? Each answer is scored from 1 to 10, and the product of the three — the risk priority number — sorts the list.
A simplified FMEA of an ordinary electric kettle might look like this:
| Part | Failure mode | Effect | Sev. | Occ. | Det. | RPN |
|---|---|---|---|---|---|---|
| Thermostat | Sticks closed | Boils dry, overheats | 9 | 3 | 4 | 108 |
| Lid latch | Opens while pouring | Hot water spills | 8 | 4 | 3 | 96 |
| Power cord | Strain at the base | Power cuts in and out | 6 | 2 | 6 | 72 |
| Element | Limescale build-up | Heats slowly | 3 | 8 | 2 | 48 |
Two things stand out. The most frequent problem, limescale, ranks lowest, because it is mild and easy to notice. And the top item leads straight to a design rule: a thermal fuse that cuts the power if the kettle boils dry. Most kettles have one precisely because someone asked how the product could fail. Many teams also treat any severity of 9 or 10 as must-fix whatever its total, because a multiplied score can hide a rare but serious failure.
How to use it
- Write the goal in one sentence. A vague goal produces vague failures.
- Spend ten minutes listing ways it could certainly fail. Aim for at least ten; the useful ones often come after the obvious first five.
- For each item, ask the three FMEA questions: how bad, how likely, and would we notice in time?
- Turn the top three into something concrete: a rule, a test, a checklist item, a time buffer, or a second person who understands the system.
- Re-read the list at the next milestone. Some risks will be gone; new ones will have appeared.
Where it misleads
Avoiding failure is not the same as achieving anything. Inversion is a filter, not a strategy: it tells you what not to do and leaves the direction to you.
The list of possible failures is also endless, and an unranked list can stall a project, so rank before you act. Finally, people mostly imagine failures they have already seen. Genuinely new failure modes are the hardest to list, which is why inversion works best alongside testing and real feedback, not instead of them.
In one line: write down how it fails, then make each of those paths hard to reach.