What Would Change the Ranking?

Most uncertainty in an analysis is irrelevant, and treating all of it as equally worth reducing is how a six-week project becomes a six-month project.

The uncertainty that matters is the uncertainty that could change which alternative you pick. Everything else is scenery, however large the error bars.

Ranges where a range changes someone’s stomach

Give a range when the range would change what someone does.

We recommend a class-based replenishment rule. That ranking survives if missed-job cost on the protected class is anywhere from roughly half to twice our base case. It breaks if dealers can bypass the system on the protected class, or if finance forbids item-level exceptions.

Notice what is not in that paragraph: a fabricated 87 percent confidence. Confidence belongs to claims, and even then as language rather than as jewelry.

“We are willing to run this for a quarter” is a real sentence about operational confidence, and it completes the thought. A p-value completes a different thought, about a different object, for a different audience. It is almost never the reason a vice president should change a rule.

The one-way table

A one-way table is usually enough. Rows are the inputs that actually argue. Mark the value at which the recommendation changes.

If this is true	Then pick
Missed-job cost roughly uniform across items	The status quo, or a least-damaging cut if the cash event is mandatory. Class-based stocking is theater.
Bypass on the protected class cannot be closed	Not a class-based rule, because you cannot enforce it.
Missed-job cost concentrated and bypass controllable	The class-based rule.

The table does something a paragraph cannot. It lets people argue about the inputs, and it lets them see, sometimes for the first time, that they are not arguing about your personality.

Value of information is a delay tax

“Let us get more data first” is a proposal with a price, and the price is rarely stated.

The price is the delay: the freeze that passes, the quarter that runs on the current rule, the cash event that happens anyway. Against that, the value of the information is bounded by the probability that it flips the ranking multiplied by the cost of being wrong.

Most proposed studies, priced this way, are not worth their delay. This is a calculation you can do in five minutes on a whiteboard, and doing it out loud is the most effective response to a request for more certainty.

Robust beats optimal on a point

An alternative that is second-best across the plausible range usually beats one that is best at the base case and poor elsewhere.

This is not risk aversion. It is a recognition that the base case is a guess, and that the cost of being wrong is asymmetric in most operational settings. A rule that degrades gracefully when an input is missing is worth more than a rule that is optimal when every input is present, because inputs go missing.

What the choice costs to undo

Ask this separately, because the alternatives frequently differ on it more than they differ on expected value.

A rule you can switch off next month is a different proposition from a contract you cannot exit, even if the spreadsheet ranks them within a rounding error. When reversibility differs, say so explicitly. It is often the argument that actually decides the room, and leaving it implicit means it gets made badly in the corridor.