Master Production Schedule: The Promise a Factory Makes to Itself

Sales comes to the planning meeting with a number: 800 units next month. Production answers with another: the line can make 650. Somebody has to write one number into the plan, and it will drive purchasing, staffing and every delivery date for the next four weeks.

That number lives in the master production schedule (MPS): the plan that says which finished products, or key assemblies, the factory will build, how many, and in which week. It sits between the forecast and the detailed machine schedule. Think of it as the promise a factory makes to itself.

The master production schedule is a reservation book, not a wish list

A restaurant with 60 seats does not book 80 guests for 8 p.m. because 80 people called. It books 60 and offers the rest a later table. The reservation book is not a record of demand. It is a record of what the kitchen has agreed to cook.

The MPS works the same way. The forecast and the order book are inputs; the MPS is the decision. Everything downstream (MRP, purchasing, the shop schedule) treats it as true.

Worked example: 800 requested, 650 possible

Step 1: Split the demand. Of the 800 units, 520 are firm customer orders and 280 are forecast.

Step 2: Subtract usable stock. There are 60 finished units in the warehouse, not reserved for anyone. Demand to produce: 800 − 60 = 740.

Step 3: Compare with capacity. 740 needed, 650 possible. Gap: 90 units.

Step 4: Protect firm orders first. All 520 go in. That leaves 650 − 520 = 130 slots for forecast demand, against 280 − 60 = 220 forecast units still uncovered. 220 − 130 = 90 units don't fit.

Step 5: Find them a home. Next month's MPS holds 550 units against the same 650 capacity, so there are 100 free slots. Sales agrees to promise those 90 forecast units for next month.

Final MPS: 650 this month, 550 + 90 = 640 next month. Every unit has capacity behind it, and sales knows today, not on the 28th, which 90 units slip.

A Saturday shift or a subcontractor were options too. What matters is that someone chooses, on purpose, before MRP runs. Whether forecast units should be built ahead at all is its own question, answered in make-to-order vs make-to-stock.

Same 650 units, different month

"650 per month" hides an assumption: the usual product mix. The real limit here is the coating line, with 400 hours a month. Standard units need 0.4 coating hours each, premium units 1.2 hours.

• Usual month: 500 standard + 150 premium. Coating load: 500 × 0.4 + 150 × 1.2 = 200 + 180 = 380 hours. Fits.
• A premium-heavy month: 350 standard + 300 premium. Coating load: 350 × 0.4 + 300 × 1.2 = 140 + 360 = 500 hours. That is 100 hours over, 25% more than the line has.

Same 650 units; one month fits and one does not. That is why a master production schedule is checked in hours on key resources, not only in units. Rough-cut capacity planning (RCCP) catches this in minutes.

Time fences: frozen, slushy, liquid

An MPS that changes every day is not a promise; one that never changes ignores customers. Time fences split the horizon so both sides know the rules. Planners describe them like water:

• Frozen (roughly the next 1–2 weeks). Material is issued and jobs are started. Changes need a senior sign-off, because each one costs setups, expediting or scrap.
• Slushy (roughly weeks 3–6). Materials are ordered but not all have arrived. Swaps within a product family are fine if material and hours exist.
• Liquid (beyond that). Anything can move.

A useful rule: the frozen zone is at least as long as it takes from releasing a job to finishing it, and the slushy zone covers your longest purchased-material lead time. At suppliers with tight customer call-offs, like the plants described in automotive production planning, the frozen zone is often set by contract.

Myth: "Load full demand and let the system sort it out"

It feels safe, because nobody has to hear "no". But MRP assumes infinite capacity, so it plans materials for all 800 units as if the line could make 800. Purchasing buys for 800. The 150 units that can't be made turn into shortages, late orders and WIP, decided by accident on the shop floor instead of by choice in the planning meeting.

An overloaded MPS does not avoid the "no". It only delays it until it costs more.

Monthly production planning: where the MPS fits

In most plants the MPS is reviewed monthly, right after the S&OP meeting. S&OP agrees volumes by product family; the MPS breaks them into products and weeks. From there, MRP works out components and purchase orders.

What the planner sees in factory.online

In factory.online, the 520 firm units arrive as sales orders that create manufacturing demand, and the planner adds the forecast units as planned manufacturing orders. RCCP shows whether the mix fits the coating line's 400 hours, or flags 500 of 400 in a premium-heavy month. In a planning scenario the planner moves the 90 forecast units to next month and sees the load on both months before anything is released. MRP then runs from the agreed plan, so purchasing buys for 650, not 800.

If your plan is really the order book with dates on it, we can walk through it together in a demo.