Quote to Cash · step 7 of 11
Planning and making it
Turning an accepted order into a scheduled job, materials, machine time and finished parts — and reporting honestly when the date is going to move.
Also called: Production planning · Scheduling · Manufacturing · Execution
- Owned by
- Production Planner
- Also involved
- Inside Sales / Sales Coordinator, Despatch Coordinator
- Documents
- Certificate of Conformity (CoC)
- Measured by
- On-Time Delivery (OTD), Gross Margin
What it is
This is the step where the promise is either kept or lost. The order arrives as a date and a quantity; production has to turn that into a sequence of operations on specific machines by specific people, with the right material arriving before the first of them starts.
It contains two distinct jobs that get confused. Planning decides what should be made and when, based on capacity, material and priorities. Execution makes it, and reports what actually happened. Companies that plan well and report badly are just as late as companies that never planned — they simply find out later.
For the quote-to-cash chain, what matters is not how the parts are made but what production owes the rest of the chain: a realistic date, early warning when it moves, and an honest signal that the goods are ready to ship.
Why companies do it
Capacity is finite and shared. Every order competes with every other order for the same machines and the same people, and the sequence they run in decides who is on time. Left unplanned, the sequence is set by whoever shouts loudest, which reliably means the quiet customers are late and the loud ones are only slightly late. Planning exists to make that decision deliberately, once, rather than fifty times a day at the machine.
Inputs and outputs
What has to be there first
- The released sales order with its promise date — Order entry
- Routing and standard times — Engineering
- Material availability and purchase lead times — Purchasing
- Available capacity by work centre — Operations
- Drawing at the correct revision — Document control
The revision named on the order, not the newest one on file.
- Outside process capacity and queue times — Subcontractors
What it produces
- A works order with a scheduled start and finish — Shop floor
- Material demand and purchase requisitions — Purchasing
- Finished goods, inspected and released — Despatch and shipment
- Actual hours and quantities booked against the job — Costing the job
The only honest feedback the estimate ever gets.
- Early warning when a date will move — Inside Sales / Sales Coordinator
How it is done
01Convert the order into works orders
One works order per batch, carrying the part, revision, quantity, routing and the date it is needed by. Where the sales order has multiple call-offs, each becomes its own works order — otherwise the whole quantity is made at once and the rest sits as finished stock at your expense.
02Check material and capacity before scheduling, not after
A schedule that assumes material arrives on time and every machine is available is a wish. Confirm the material lead time, the outside process queue and the capacity at the constraint before committing a date. The constraint is the only work centre whose capacity actually sets the output.
03Sequence the work
Decide the order jobs run in, on each work centre, and make that decision visible. The sequence should follow the promise dates, adjusted for setup families where changing between similar jobs saves hours. What it must not follow is the phone.
04Release only what can actually start
Releasing every order to the floor as soon as it is entered creates a queue on the shop floor instead of in the system. Work in progress rises, lead times get longer, and nobody can see what is genuinely next. Release when material is available and capacity is close.
05Run first-off inspection before the batch
Check the first part against the drawing at the stated revision, and record it. The cost of finding a dimensional error on part one is one part. On part 400 it is 400 parts, plus the delivery, plus the customer's opinion.
06Book time and quantity as it happens
Actual hours, actual good quantity, actual scrap, against the job. This is what tells you whether the estimate was right and where the losses are. Booked at the end of the week from memory, it tells you nothing.
07Manage the exceptions daily
A short daily review of what is late, what is at risk and what is blocked, with a named owner for each. The purpose is not to report the problem but to decide the recovery while there is still time for one.
08Signal the slip early
The moment a promise date is known to be at risk, sales is told — not when it is missed. A customer given three weeks' notice can re-plan; a customer told on the due date has already committed to their own downstream promise.
09Complete, inspect and release to finished goods
Final inspection against the specification, certification produced if required, quantity confirmed, and the job formally closed so despatch can see it is available. An order that is finished but not booked in is invisible to everyone waiting for it.
What needs approval
Approvals only work when they happen before the commitment. Retrospective approval is a manager being told what has already been promised.
| Trigger | Approver |
|---|---|
| Promise date will be missedOperations decides the recovery; sales decides how it is communicated. | Sales, before the customer is told |
| Overtime or weekend working to recover a date | Operations manager |
| Concession or deviation from the specificationShipping non-conforming goods without written agreement is how a delivery becomes a claim. | Quality, and usually the customer |
| Scrap above the allowance, requiring a rerun | Operations and, where it affects the date, sales |
| Subcontracting an operation normally done in house | Operations and quality |
Why the bracket order was five days late
A worked example with real numbers.
The 400 brackets were promised in 28 days. The routing said 6 minutes per part of milling, 3.5 hours of setup, then deburr, inspect and anodise. Forty hours of actual work, in a 28-day window. It went out five days late.
None of the loss was on the machine. The material was ordered on day 2 and arrived on day 9 — the bar supplier's standard lead time is 5 days but the order was placed on a Friday afternoon. Milling was scheduled to start on day 10 but the mill was still running a job that had been expedited on day 8 for another customer, so it started on day 13 and finished on day 15.
Anodising was sent on day 16. The subcontractor's stated queue was 10 days, which was in the plan. What was not in the plan was that they returned 388 good parts and 12 rejects for surface marking, discovered at goods-in on day 26. Twelve parts had to be remade — one more setup, one more anodising cycle, one more queue.
The reruns went out on day 33. The customer was told on day 27, when the anodising rejects were found. Had the expedite on day 8 been declined, or had the material been ordered on day 1, or had 405 parts been started instead of 400, the order would have shipped on time.
| Days | Activity | Value-adding? |
|---|---|---|
| 1–2 | Order entry, works order raised, material ordered | No |
| 3–9 | Waiting for bar stock | No |
| 10–12 | Queued behind an expedited job | No |
| 13–15 | Setup and milling | Yes |
| 16 | Deburr, inspect, despatch to anodiser | Partly |
| 17–26 | Anodising queue and return | Partly |
| 27–33 | 12 rejects remade and re-anodised | No |
The point
Forty hours of work took 33 days, and roughly four of those days were spent adding value. The delay was not caused by production being slow. It was caused by three ordinary decisions — a Friday material order, an unpriced expedite, and starting exactly the quantity ordered on a process with a known yield loss.
Common mistakes
MistakePlanning to standard times nobody has ever achieved
Why it happens: The routing was set when the part was new and has never been revisited. Every schedule built on it is optimistic by the same margin, and the plant is permanently and inexplicably late.
What to do instead: Compare booked hours to standard on completed jobs and update the routings. A standard time that is consistently beaten or consistently missed is not a target, it is a measurement error.
MistakeReleasing everything to the floor at once
Why it happens: It feels like progress and it keeps operators busy. It also fills the shop with work in progress, so lead time stretches, priorities become invisible and every job is simultaneously urgent.
What to do instead: Release against the constraint, not against the order book. Less work on the floor moves through it faster — which is counter-intuitive enough that it needs to be measured to be believed.
MistakeRescheduling by phone call
Why it happens: A customer chases, someone promises to expedite, and a job is pushed to the front. The four jobs it displaced are now late, and nobody has told those four customers.
What to do instead: Expediting is a decision with a cost — make the cost visible. If a job jumps the queue, name what it displaced and who is telling that customer.
MistakeHiding the slip
Why it happens: Nobody wants to deliver bad news that might still be recovered. So the date holds in the system until the day it fails, and sales finds out from the customer.
What to do instead: Report at-risk before late. A date flagged three weeks out and then recovered costs nothing; a date that fails without warning costs the relationship.
MistakeMaking the whole blanket quantity in one go
Why it happens: It is more efficient per part — one setup instead of four — and the efficiency figures reward it. The stock then sits for nine months, financed by you, at the risk of a revision change that makes it scrap.
What to do instead: Make the call-off, not the contract. If batching up genuinely saves money, calculate the holding cost and the obsolescence risk against the setup saving before deciding.
MistakeNot booking scrap honestly
Why it happens: Scrap looks bad, so a rerun is quietly absorbed into the next job. The yield figure stays healthy, the estimate keeps using the wrong yield, and the same job loses money every time it runs.
What to do instead: Separate the reporting of scrap from the judgement of it. You cannot price yield you do not measure.
Where the work happens
- ERP or MRP
- Finite capacity scheduling
- Shop-floor data collection
- Daily production board
- Quality management system
Terminology
- Works order
- The instruction to make a specific quantity of a specific part, carrying the routing and the due date.
- Routing
- The sequence of operations, the work centre each runs on, and the standard time for each.
- WIP
- Work in progress — material that has been started but not finished. Cash sitting on the shop floor.
- Constraint
- The work centre whose capacity limits total output. Time lost there is lost for the whole plant; time saved anywhere else is an illusion.
- First-off
- The first part from a setup, inspected and approved before the batch runs.
- Concession
- Written agreement, usually from the customer, to accept something that does not fully meet the specification.
- Expedite
- Moving a job up the queue. Always at the expense of another job, whether or not that cost is acknowledged.
- Yield
- Good parts out as a proportion of parts started. If it is below 100%, you must start more than you intend to ship.
Common questions
›What is the difference between planning and scheduling?
Planning decides what should be made and roughly when, over weeks or months, against capacity and material. Scheduling decides the exact sequence on each machine over the next hours or days. Companies usually have one and believe they have both.
›Why do jobs take so much longer than the hours on the routing?
Because almost all of the elapsed time is queuing, not working. A job with 40 hours of operations routinely takes weeks, and the difference is waiting — for material, for a machine, for inspection, for a subcontractor. Reducing lead time is mostly about reducing waiting, which is why adding machine capacity often changes nothing.
›When should the customer be told a date will slip?
As soon as it is known to be at risk, not when it is confirmed as late. The value of the warning falls to nearly zero on the due date, because by then the customer has already made their own commitments against it.
›Should you make more than the order quantity?
On any process with a known yield loss, yes — starting exactly the ordered quantity guarantees a shortfall. The overrun should be calculated from the actual yield and included in the cost, not decided by the operator on the day.
›Why does expediting make everything later?
Because capacity is conserved. Moving one job forward pushes every job behind it back, and each of those may then need expediting in turn. A plant that expedites routinely is one where the sequence is set by chasing rather than by dates, and its average lateness rises even though each individual expedite looked justified.
›What should production report back to sales?
Two things, reliably: the current expected completion date for every open order, and any change to it as soon as it is known. Everything else — efficiency, utilisation, scrap — matters internally but is not what the rest of the chain is waiting for.
Tools for this
Downloadable, self-contained, and yours to keep — they run in a browser with no account and no data leaving your machine.
- Production ScheduleProSequence jobs on your work centres with planned dates, hours and priorities, catch late-running work before the promise breaks, and measure schedule adherence and estimating accuracy on every completed job. Nothing is uploaded.
- Production Capacity PlannerFreePlan weekly production capacity: define machines and shift patterns, load them with product demand and routings, find bottlenecks and test what-if scenarios. Runs entirely in your browser.
- Daily Production BoardProRun an hour-by-hour production board — plan against actual for every period, the loss reason and countermeasure written in the hour it happened, attainment, first-pass quality and a downtime Pareto, printed for the shift handover. Nothing is uploaded.
- WIP Aging TrackerFreeTrack work-in-process jobs, see how long each job has sat in its current stage, spot late work and stages hoarding value. Runs entirely in your browser. No installation, no account, no upload.
- Bottleneck AnalyserProFind the real constraint in a process chain by effective capacity — rate, uptime and yield combined — then size the utilisation, the near-constraints and what closing the demand gap is worth. Nothing is uploaded.
Reviewed 2026-08-15. Nothing here is legal, tax or accounting advice — contractual and tax practice varies by jurisdiction.