Name the Choice | The Decision Product

“Inventory and service” is a heading. “By the April freeze, Priya will choose how we replenish dealer-facing parts for the next cycle” is a thought.

The first cannot be accepted or rejected. The second can. That difference is the whole of this entry, and the reason it matters is practical rather than literary.

Naming the incumbent tells you what a challenger has to beat. It does not tell you which decision the challenger is competing in, and the two are easy to confuse because the incumbent is visible and the decision usually is not.

A heading does not constrain work

“Inventory and service” licenses any analysis that can be said to touch it: a dashboard, a forecast, a segmentation, a tour of every distribution center. Nothing is out of scope, so the work expands to fill the calendar. “By the April freeze, Priya will choose among three replenishment strategies” licenses only work that could change that ranking, in time. Everything else becomes visibly optional, which is usually the first honest thing that has happened to the project.

The shape

By this date, this person will choose among these alternatives, in order to this, knowing that this will still be uncertain.

Three questions, answered in plain sentences and without a notebook, supply most of it. What is actually being scored: the consequence the company would still care about if its reports were redesigned tomorrow? Who is actually deciding: the person who can commit and could later undo it, rather than the person who asked? What will still be unknown when they commit?

Write it even if the first draft is ugly; an unwritable sentence is information. It can also fail while looking complete, because completeness is a property of the blanks, not of the grammar.

The four things that go wrong

The owner is a requester. The person who sent the email is often not the person who can commit. “The S&OP meeting” is not an owner; a meeting cannot own a parameter. The test is who can change the action without convening another group.

The date is the meeting. The meeting is when people want to feel current. The freeze, the contract window, the quarter-end cash event: those are when the choice goes stale. A team that treats the meeting as its deadline delivers a dashboard, because that is what a meeting can consume. A team that treats the freeze as its deadline delivers a rule, or an honest statement that the freeze will pass under the old one. Put reversibility beside the date: a liquidation that cannot be undone and a class assignment reversible at the next freeze are different objects.

The alternatives are not the same kind of thing. “Change the rule, build a dashboard, or hire two analysts” cannot be ranked, because its items do not compete for the same commitment. Every alternative must be something the owner could commit to in place of the others. The dashboard may still get built; mixing it into the menu is how a study postpones a commitment while looking reasonable.

The incumbent is missing. The status quo is always an alternative, because it is the rule already running and what the process will do if the meeting produces nothing. Omitting it treats it as costless and frames the decision so that something gets built. Unattractive alternatives belong too: finance’s cut is on the menu because finance put it there, and leaving it out only ensures the analysis is ambushed by it later.

Rank complete strategies over a common horizon

When a one-time action and a repeating rule both move the same inventory, they are different kinds of commitment, and evidence gathered for one will not rank the other. If they are genuinely separate objects, write two sentences. Combining them into “the inventory decision” guarantees that one is decided by default.

When they must be ranked against each other, rank complete strategies rather than events. Each alternative states what happens to stock at the freeze and which rule replenishes afterward, and all are scored over the same horizon. Finance’s cut, written as an event, is not an alternative at all, because with the old rule left running behind it the next order run buys the stock back. Written as a strategy it is: every item’s stock down 20 percent by the freeze, then the current rule at targets 20 percent lower. That can be ranked against a class rule, and it may turn out to be the least damaging way to meet a mandatory cash requirement, or indefensible. Either finding is useful.

Constraints are not dislikes

What is illegal, contractual, or physically impossible belongs in the choice as a constraint. What is merely disliked does not.

Lead time, dock capacity, and what a dealer can physically hold are constraints. “We need 98 percent fill” is usually a dislike wearing a number, often the shape a remembered stockout takes, because a loss weighs more than an equal gain. It deserves respect as evidence of an interest. Encoded as a hard constraint, it turns the ranking into a search for a point that satisfies two slogans, and often no such point exists.

The objective belongs in the sentence in units of consequence. “Improve fill and turns” names two proxies in tension; “reduce working capital and the cost of stranded jobs” names two consequences that can be priced and argued about.

Uncertainty you will still have

The last clause is the one people delete first, and it is load-bearing. Name the two or three uncertainties that would make a different alternative look better: demand noise, whether dealers will work around a new rule, whether finance will allow class-level parameters. No distribution is needed yet. A choice with no remaining uncertainty claims the ranking is a fact, which is how point estimates reach Friday’s slides.

When there is not one owner

Ask for a single owner anyway, because a decision four people must jointly endorse is one the process is unlikely to run. When authority really is divided, record the division as it is. Three patterns cover most cases.

The first is a split that can be named. One person owns the rule; others hold constraints or adoption risks. Priya owns replenishment, finance owns a cash event this cycle cannot avoid, and dealer operations can defeat the rule in the field. The sentence carries the constraint (“Priya will choose the replenishment rule, subject to finance’s cash requirement for this cycle”), and the ranking carries the adoption risk.

The second is genuine dual commitment: two people on opposing scorecards who must both say yes, with nobody above them to break the tie this cycle. The sentence carries both names, and the alternative that survives is usually one that serves both interests better than either position does. If none exists, say so and name who absorbs the loss.

The third is a committee that will not name anyone. Do not invent an owner to complete the template. Send the sentence back with the owner line blank: until someone answers for the outcome after the freeze, the work is a study rather than a decision.

The person who will run the installed rule is often different from the person who commits to it, and both names belong in the write-up. A commitment with no operating owner leaves the process unchanged after a good meeting.

Name the role of your own work at the head of the write-up, because the same facts support three different documents. Recommend: you name and rank the alternatives, and someone else commits; most technical work is this. Decide: you commit; rarer than technical people think, and more common than they admit when they quietly change a threshold in production. Inform: you support a pick that has already been made; noticing when you have been handed this under the name of a recommendation is its own skill. If you cannot name the role, you do not yet know what the week is for.

The filled choice for Meridian

Priya Shah, vice president of Parts, owns it; not analytics, not the S&OP facilitator. The deadline is the April planning freeze, reversible at the following freeze for the class assignments but not for units already ordered. The alternatives are three complete strategies scored over the twelve months from the freeze: keep the current rule with no action on stock; finance’s cut as written above; or reset targets by class at the freeze, draining the cheap class to its cash cap and rebuilding the protected class, then run the three-class rule. Whether finance’s cash event needs a one-time cut beyond what the chosen strategy releases is a separate choice, owned by finance and due at quarter close. The objective is working capital and the cost of jobs that stop vehicles. The uncertainties are demand noise, dealer bypass, and whether finance will allow class-level parameters. It is a repeating rule, and the one-time action each strategy takes at the freeze is part of the strategy, not a second decision hidden inside it.

The test

Fill the sentence for the brief on your desk in five minutes, then check each blank. Can the named owner change the action without convening another group? Is the date when the choice goes stale, and does it say whether the commitment can be undone? Could the owner commit to any one alternative in place of the others, is the incumbent among them, and does each state what happens at the start and which rule runs after? Is the objective in units of consequence? Are the uncertainties ones that could change the pick?

Any blank you had to invent is the first thing to take back to the person who sent the brief.