Production planning: turning a demand picture into a buildable plan

Setting the horizon so the plan can still buy material

The planning horizon has to reach past the longest cumulative lead time in the bill of material, otherwise the plan is issuing build instructions for parts that cannot arrive in time. Most plants run nested buckets: daily detail for the coming weeks, weekly beyond that, and monthly at the far end where only aggregate volume matters. The mistake is treating the far horizon as decoration. It is the horizon that triggers long-lead purchase commitments, tooling orders and seasonal hiring, so error there surfaces later as a shortage on the line. Planners should revisit which components drive the horizon whenever a new part or a new supplier enters the picture.

The feasibility test a plan must pass before it is published

Before a plan leaves the planning office it should be checked against the things the floor cannot conjure: machine hours on the key resources, trained labour by shift, material arriving at the required dates, and tooling or fixtures that may be shared between products. Publishing a plan that fails any of those teaches supervisors to ignore plans, which is expensive to undo. Where a check fails, the planner chooses — move the work, add capacity, or renegotiate the date — and that choice belongs on the record with a name against it. Silent infeasibility, where nobody admits the plan cannot be built, is the most common failure in planning offices.

Freezing the near term without freezing the business

Most plants operate zones: a frozen period where only a named manager may authorise a change, a negotiable middle where changes cost setup time and expediting, and an open far end where commercial teams can move anything. The frozen period should be set by the physical reality of setup, material call-off and staffing notice, not by planner preference. It also needs a named breaker, someone senior enough to accept the cost of an emergency insertion and to see the knock-on for other customers. Plants that leave the frozen zone undefined end up with an informal one anyway, enforced by supervisors quietly refusing to change over.

Counting plan churn instead of complaining about it

Churn is measurable. Record every change made inside the frozen and negotiable zones, tag it with a cause — customer pull-in, quality hold, material shortage, forecast error, engineering change — and report the tally to the same meeting that approves the plan. Once churn has an owner and a cause code, the conversation shifts from blaming planners to fixing the source. Commercial teams that can see their pull-ins consuming changeovers and expedited freight moderate the requests. A supplier whose late deliveries force resequencing gets a specific conversation rather than a general complaint. The record also settles arguments later about why a particular customer date slipped.

Judging the plan on adherence rather than on volume

A plant can hit its monthly volume and still have missed almost every individual commitment, because it built the easy products and pushed the awkward ones to the end. Adherence measured at the item-and-day level exposes that; aggregate output hides it. Track whether the item planned for a day was the item built that day, in the quantity planned, and treat overbuild as a miss too, because it consumes capacity and material that belonged to something else. The planner owns the number jointly with the production manager, since roughly half the causes sit in the office and half on the floor.

Frequently asked questions

Who should own the production plan, planning or operations?

Planning owns the plan; operations owns delivering it. The split only works if operations holds a formal veto during the feasibility review and uses it before publication rather than afterwards on the floor. Where the two functions report to different directors, disputes escalate instead of resolving, so many plants put both under the plant manager and hold a short daily meeting where yesterday's misses are attributed before the next plan is confirmed.

How long should the frozen period be?

Long enough to cover the longest thing you cannot undo cheaply: the call-off notice you owe suppliers, changeover and setup time on the constraint resource, and the notice period agency or shift labour requires. If those differ widely, set the fence by the longest and allow named exceptions rather than shortening it across the board. Revisit it whenever a supplier changes or a product with different setup behaviour joins the mix, because an inherited fence quietly stops matching reality.

What is the difference between the production plan and the schedule?

The plan says what will be built and roughly when; the schedule says which machine, which operator and in which order. Plans are stated in weeks or days at item level and are agreed with commercial and purchasing colleagues. Schedules are stated in shifts and sequences and belong to the supervisor. A plan can be feasible in aggregate yet produce an unrunnable schedule if the sequence forces excessive changeovers, which is why the two are reviewed together.