Your Schedule Is a Materials Commitment, Not a Demand Plan

The rule that made the week hold

A planner at a food manufacturer changed one rule, and the schedule started holding. She stopped placing a job on the week when its material was not yet in the building. Before that change, a job could sit on Wednesday while the film was still in transit on Monday. The schedule looked complete. It was a list of intentions.

The second half of her rule carried as much weight. When a job missed its week, she stopped re-fitting it into the current week. The recovery went into a future week that had open hours. Her reasoning was blunt. Move last week's miss into this week, and this week's work moves into next week, and then you never finish.

Both halves describe one decision. A schedule is a commitment of material, time and equipment. Treat it as a statement of what the customer wants, and every supplier delay becomes a plant problem.

A schedule is a sequence of steps, not a statement of demand

Most products pass through more than one step, and some of those steps hold time that cannot be compressed without changing the process. Blending and portioning can fit inside a week. A packing step that waits on an intermediate hold cannot. The earliest honest date for an item is therefore a derived number: last material arrival, plus the hold time between steps, plus run time for each step, placed in hours the plant actually has open.

A demand date is a different number. It arrives from a customer or a forecast, and it carries no information about arrivals or holds. When those two numbers are treated as interchangeable, the plant becomes the place where the difference gets absorbed.

Absorption has a shape, and it is not a late start. The job cannot run, so the schedule moves. Moving one job on a loaded week means resequencing the jobs around it. That adds changeovers nobody planned, splits runs that were sized to be economic, and pulls crew hours into setup instead of production. The late order is visible to everyone. The extra changeovers and the short runs land in scheduling variance, where they read as planning noise rather than as the cost of a receipt that slipped.

Then it compounds. Push the recovery into the current week and the week absorbs work it was never sized for, which displaces jobs into the following week. Each week inherits a deficit larger than the one it created. The plant reads this as a capacity problem, because that is what a chronically full week looks like from the inside. The binding constraint is upstream of the line, in the gap between a promised arrival date and a physical receipt.

This is why line rate is the wrong place to look first. A faster line shortens run time, which is the one term in the derived date that was never the problem. The terms that move attainment are the arrival and the hold.

Where the decision sits, and who owns it

The gate is the release step, and it belongs to one named person. The rule is testable: a job is not released until every input has been received and the required hold steps fit inside the scheduled window. Everything else about the schedule can stay as it is.

Then count what the gate catches. Record every release that fails it, with the cause, and keep the count weekly. That list is not a planning complaint. It is a sourcing input. Items whose material routinely lands inside the week it is needed have a lead time or terms problem, and the plant has been paying for it in changeovers. Where a customer or a supplier agreed to deliver material inside the production week, that agreement has a price, and the price is the reshuffle it buys. Name it in the next commercial conversation, with the count behind it.

The rule I apply when I review a plant's schedule is to read the release list before the attainment chart. Attainment tells you the week missed. The release list tells you whether the week was ever feasible.

Verify the result where it lands, not on the schedule. A working release gate shows up as fewer unplanned changeovers at the same volume and mix, and as fewer jobs moved after release. Those are operating results. They become financial results only when the recovered hours are converted: reduced paid overtime, a shift or a hire avoided as volume grows, or volume run that the plant previously could not promise. Fewer reshuffles is not margin until someone names which of those three happened.

What a well-run planning week reads like

Every released job has all inputs received and its hold steps inside the window. The count of jobs released without material is zero, and it is counted rather than assumed. Last week's miss appears in a future week with open hours, never in the current one. The planner can state the earliest feasible date for any item from its step sequence and material lead time without calling a supplier. Unplanned changeovers are counted separately from planned ones, and at a steady mix that count is flat week over week.

Read your own release list

Pull last week's released jobs. Mark each one where every input was physically on site at the moment of release. Then list every reschedule the week produced and trace it to a release. If most of those reschedules trace back to jobs released without material, the next capacity you gain will not come from the line. It will come from refusing to schedule what has not arrived.