Master production schedule, arithmetic.
Also called: MPS · Build plan · Production plan · Run schedule
Deciding what to make, how many, and when, by balancing demand against supplier lead times and capacity so you neither stock out nor over-build.
The MPS is arithmetic, not ambition. Lay demand (pre-orders plus forecast) against the longest lead time and your capacity, then back-date the order dates. Get it wrong one way and you stock out; wrong the other way and cash sits dead in unsold stock.
What a master production schedule is
A master production schedule is the dated answer to three numbers: how many units, by which date, ordered when. It sits on top of three inputs that rarely agree with each other, and the whole job is reconciling them.
Demand is what the market will take. For a first run that is firm pre-orders plus a forecast you should hold loosely. Lead time is how long each part takes from order to your door, and the longest one (the long pole) sets the earliest date the finished unit can exist. Capacity is how much you can actually build per week once parts arrive, set by assembly hours, test rigs, and how many hands you have. The schedule is where those three meet on a calendar.
The discipline is back-dating. You start from the date stock must be on the shelf, subtract the long-pole lead time, and that gives you the date the purchase order has to go out. Miss that date and no amount of effort downstream recovers it, because a kiln does not fire faster because you are anxious. In my experience the long pole is almost never the part founders watch; it is usually the one they assumed was easy.
The two failure directions are symmetrical. Build to a number bigger than demand and you have converted working capital into stock that depreciates in a unit. Build to a number smaller than demand and you stock out in the one window that mattered, which for a seasonal product can be the whole year. The MPS exists to keep you off both rocks.
Here is what that schedule looked like for the proofing box we ran through it, so you can see the shape of a real answer rather than a generic template.
Notice the schedule names dates, not hopes. The Year-1 number of roughly 3,000 is a target, but every run inside it is pulled by real sell-through and pushed by the firing lead time.
- “Order a few thousand, it’ll sell.”
- Ignore the firing lead time until autumn.
- One big run, no repeat trigger.
- Capacity assumed, never counted.
- Run size = pre-orders + a modest forecast buffer.
- Back-date the ceramic order from the shelf date.
- Repeat runs triggered by sell-through.
- Weekly capacity counted, not assumed.
How it fits the bigger picture
Master production schedule is activity 09.20.01, the planning spine of Stage 09 Manufacture. It sets the dates and quantities that the next activity, routing instructions (09.20.02), then turns into the step-by-step build sequence on the floor.
What it can do
It turns demand, lead times, and capacity into a dated plan you can put a purchase order against. It tells you the exact day the long-pole order has to leave, and it keeps cash from drowning in stock you guessed at.
What it can’t do
It can’t fix demand you haven’t measured; a schedule built on a forecast you invented is just confident guessing. And it can’t shorten a real lead time. If the Stoke-on-Trent firing takes what it takes, the MPS plans around it rather than wishing it away.
See the full 10-stage process →
Try it yourself
Take your firm pre-orders, add a forecast buffer you’d defend out loud, and that is your first run size. List every bought-in part with its lead time, circle the longest, and call it the long pole. Fix the date stock must be on the shelf, subtract the long-pole lead time, and that is the day your purchase order has to go out. Put it in the calendar today.
Want a structured first pass at the demand side? Start the Free Sprint → and the GPT will help you frame the numbers.
Your MPS checklist
Project notes: the long pole was the kiln
▸ From the notebook · optional reading
Scheduling the proofing box’s first run, and the Stoke-on-Trent firing lead time that quietly set every other date.
3 min read · click to open
Dan and Anna Hartley wanted stock landed for Christmas. Sensible: a £149 counter object is a gift, and the gifting window is most of the first year’s volume. The question was a simple-sounding one. “When do we press the button on the first run?”
Working it backwards
We laid the parts out with their lead times. The Manchester PCB had a known turnaround. The wood cladding was fast. The ceramic shell, fired in Stoke-on-Trent, was weeks longer than anything else, and the kiln batched firings, so we did not even own the start date outright.
So the ceramic was the long pole, and I worked the whole schedule off it. Shelf date for Christmas, minus assembly and test, minus the firing window, minus a buffer for the batch slot. That landed the ceramic purchase order at roughly the end of summer. Everything else we ordered to arrive just before it, because holding PCBs while waiting on ceramic is just cash sitting still.
Sizing the run
Dan’s instinct was to order a few thousand and “be done with it.” We pushed back. The firm pre-orders plus a forecast we’d actually defend came to a first run of 500 to 1,000, not 3,000. The Year-1 target of around 3,000 stayed on the plan, but as repeat runs triggered by sell-through, each one re-back-dated from the same firing lead time. One big speculative run would have parked most of the year’s cash in a unit, in stock, betting on a forecast nobody had earned yet.
The discipline was dull and it worked. We hit the shelf date, the kiln slot held, and the second run went in the moment sell-through justified it. The only thing that nearly bit us was assuming the buffer on the firing slot was generous; it wasn’t, and a week’s slippage there would have cost the whole Christmas window.
— Manufacture stage, project notes, 2026
— Next in Manufacture → Routing instructions