Start from the demand signal, not the machine
Every run should be traceable to a reason: a confirmed customer order, a finished-goods level below its minimum, or a forecast someone is willing to put their name to. Runs that exist because a machine was free and someone thought stock looked low are how a workshop ends up with three pallets of an item nobody has ordered since spring.
Write the reason on the work order. A run tagged to a sales order can be reprioritised sensibly when the customer moves their date; a run tagged to nothing gets treated as filler and shuffled until it becomes urgent for no reason anyone can reconstruct.
Where the demand is a stock target rather than an order, check the target is still right before building to it. Minimum levels set eighteen months ago against a different sales pattern are one of the quieter sources of overproduction.
Decide how much to make
The run size sits between two costs. Every changeover, setup and clean-down costs time that produces nothing, which pushes towards long runs. Everything you build beyond near-term demand ties up cash, occupies space and risks becoming obsolete, which pushes towards short ones.
The practical approach for most small manufacturers is to cover known orders plus a defined period of expected demand, not to build to a theoretical economic batch quantity. Two weeks of cover on a fast-moving item and build-to-order on a slow one is a policy the shop floor can actually follow.
Constraints usually decide it anyway. Minimum order quantities on components, the capacity of a curing oven, the shelf life of a mixed batch, or a pallet that holds forty-eight units will all round your ideal number to a practical one, and pretending otherwise just adds a rounding argument to every plan.
- Cover confirmed orders first, then a defined period of expected demand
- Round to the constraint: batch capacity, pallet quantity, component MOQ
- Respect shelf life on anything mixed, coated or perishable
- Do not build a year of a slow mover to avoid one changeover
Explode the bill of materials against real availability
For each component, multiply the quantity per unit by the run size, add the scrap allowance, and compare the result against stock that is not already committed to another released run. That last clause is the one that gets skipped, and it is the reason two runs both look feasible on Monday and both stall on Wednesday.
Do the same for subassemblies. If the finished item consumes a frame that you also build, the frame is either in stock, in a released run of its own, or it is a second production run that has to be planned and finished first. Treating a made component as if it were a bought one is the most common planning error in a small shop.
Where a component is short, the decision is deliberate: reduce the run to what the material supports, split it into a first pass now and a second when the delivery lands, or hold the whole run. All three are fine. Releasing the full quantity and discovering the shortage at the bench is not.
Add the scrap allowance at this stage rather than after. A five percent loss on steel tube is not an accounting adjustment, it is two extra lengths that have to physically be in the rack before the run can finish.
Sequence against capacity, then release
Order the runs that pass the material check by due date first, then look for changeovers you can avoid. Three runs needing the same tooling or the same colour are cheaper together than spread across a week, and moving one of them forward two days often costs nothing at all.
Check the constraint, not every station. Most workshops have one process that determines throughput, and a plan that keeps that one busy without starving it is good enough. Balancing every work centre is an optimisation problem you do not have and cannot maintain.
Release only what can actually start. A released work order should mean material is available, the work centre is free within the planned window, and someone owns it. Everything else stays planned, which keeps the released list meaningful rather than aspirational.
In StockFlow, the Production module turns bills of materials into work orders on a planning board and reports parts demand across runs, so the material check and the sequencing happen against the same live stock figures rather than a spreadsheet copied on Monday. It is included on Professional and above, or available as a $4/month add-on.
- Sort by due date, then group shared setups
- Protect the constraint process rather than balancing all of them
- Release only runs with material, capacity and an owner
- Keep blocked runs visibly on hold with the reason recorded
Close the run honestly
Record completed and scrapped quantities as separate numbers. A run planned for twenty that produced nineteen good units and one reject is a ninety-five percent yield, and that is a fact worth knowing; recording it as nineteen complete leaves one unit permanently in limbo and hides the loss.
Consume the components that were actually used. Where a build was completed with a substitution or an extra length cut wrong, the deduction should reflect that rather than the theoretical bill of materials, otherwise stock records drift by exactly the amount of the variance.
Compare planned against actual and use it on the next run. Yield below expectation on the same product three times running is a bill of materials problem, a process problem or a supplier problem, and the fix is cheaper than continuing to plan runs that consistently deliver less than they promise.
Update the BOM when reality disagrees with it repeatedly. A scrap allowance that has been wrong for six months is not a rounding error, it is a costing error on every quote produced since.
Release runs you know can finish
Turn bills of materials into work orders, check component availability across every released run, and record builds with the stock they actually consumed. Free to start.
Trusted by small businesses
What our customers say
“Super Kind! Quick replies from their support and very easy fixes, changed the dashboard a bit and customized it. Also gave me 450 items extra on the free plan just for me. Highly recommend and again great service!”
“Best customer service! Stockflow's customer support is fast and extremely helpful. They assisted me with customization of the software to improve my experience as a user.”