Most missed due dates are quoting failures, not shop-floor failures

When a job shop misses a due date, the post-mortem almost always points at the floor. A machine went down. An operator called in sick. Material showed up late. Those things are real and they happen. But walk the timeline back far enough on most late jobs and the actual failure happened weeks earlier, at a desk, the moment someone promised a date the shop was never going to keep.

Quoting is where on-time delivery is won or lost. The shop floor mostly just absorbs the consequences of a number that was wrong before the first setup.

This is an uncomfortable place to look, because the quote desk and the floor are usually run by different people with different incentives. The quoter wants to win the job. The floor has to deliver it. When those two are disconnected, the gap shows up as missed due dates — and gets blamed on execution.

Here's how to close that gap: how to tie quoting discipline back to your shop's real, current capacity, so the dates you promise are dates you can actually keep.

A late job is a contract the floor couldn't honor

There are two ways to read a late job. Either the floor underperformed against a fair date, or the floor performed fine against an unfair date. Working closely with SMB manufacturing operations, we see the second far more often than shops admit.

The reason is structural. Quoting usually runs on standard lead times — "we say four weeks for that kind of work" — or on gut feel calibrated to a slow week. Neither reflects what is actually on the machines the day the quote goes out. A four-week lead time means nothing if the shop already has six weeks of committed work queued on the constraint cell.

The tell is shops that miss dates in clusters rather than at random. The hidden cost of manual scheduling sits at 5–10% of revenue for a typical job shop (Qlector 2025) — for a $2M shop, that's $128,000–$276,000 a year once the downstream costs are added in. A meaningful slice of that is rework on the schedule itself: expediting, resequencing, and the overtime you burn to recover a date you should never have promised.

When the date is the problem, no amount of floor heroics fixes it permanently. You can save this week's late job with overtime. You'll be saving next week's the same way.

Why the quote desk can't see the shop's real capacity

The core problem is visibility. Most quote desks are flying blind on current load.

A quoter is typically working from one of three things: a standard lead-time table, an ERP module that reports infinite capacity, or memory. All three share the same flaw — none of them shows what is already committed on the constraint.

Infinite-capacity scheduling. Many ERP and MRP systems schedule backward from the requested due date and assume the work center can absorb the load, because they don't model finite capacity. So the system confirms a date that quietly stacks a third job onto a machine already running two. The number looks authoritative. It's fiction.

Spreadsheet or whiteboard. The load lives in the scheduler's head or in a sheet that's a day or two stale. The quoter and the scheduler aren't looking at the same picture, so the promise and the plan drift apart — and the drift is invisible until the date arrives. This gets sharply worse the more customers you're juggling across the same machines, which is the everyday reality of running multiple customers' jobs through shared resources.

Here's math you can check against your own shop. A shop running two shifts, Monday to Friday, has roughly 80 production hours per machine per week against the 168 hours on the calendar. The moment committed backlog on a cell exceeds the available hours before a promised date, that date is already late — you just don't know it yet. A quote desk that can't compare committed hours against available hours is guessing, and job shop lead time quoting done on a guess is how clusters of late jobs get manufactured.

Picture a 30-employee contract machining shop with one five-axis cell that everything tough has to pass through. The estimator quotes a new job at the shop's standard three-week lead time. What the estimator can't see is that the cell already has 110 committed hours queued against roughly 80 hours a week of capacity. The job is late before the PO is even signed — not because anyone on the floor did anything wrong, but because the date was set against capacity that didn't exist. Every shop has a version of that cell. The only question is whether the quote desk can see it.

What a confident due date actually requires

A due date you can defend rests on three things, none of them optional:

• Real backlog visibility. You need to know what's committed on each constraint resource right now — not a standard lead time, the actual queue.
• Honest job duration. Setup plus run plus realistic queue time, not best-case cycle time with no allowance for the jobs ahead of it.
• A deliberate buffer. Every shop has variability. A confident date includes a buffer sized to your real variability, placed on purpose rather than bolted on in a panic.

The deeper point: due date promising in manufacturing is a capacity question, not a calendar question. "How long does this kind of part take" is the wrong question. "When can this part start, given everything already in front of it, and how long will it take from there" is the right one. That reframe — from part duration to resource availability — is the whole game, and it's the same logic that drives capacity planning across the shop.

The buffer deserves its own emphasis. Unplanned downtime runs 35% more expensive than planned downtime (Arda Cards 2026), and it lands without warning. A due date with zero slack assumes a perfect run from quote to ship, and no shop runs perfectly. The buffer isn't padding to cover slow work. It's the price of variability you already know exists.

What a bad quote actually costs the shop

The cost of a date the shop can't keep isn't just one unhappy customer. It compounds.

When a promised date collides with reality, somebody has to resequence — bump job A to save job B. That decision carries a price. A scheduling conflict that reaches the floor costs $250–$1,000 per incident in machine restart, resequencing, and lost capacity (Product Brief). And the conflict you created by over-promising doesn't stay contained: pulling one job forward pushes another back, and now you're choosing which customer to disappoint. The healthier move is to keep those conflicts off the floor in the first place, which is a scheduling discipline you build before the date is ever promised.

The second-order cost is your own quoting credibility. A shop that misses dates starts padding every quote defensively — quoting six weeks for four weeks of work because nobody trusts the four-week number anymore. That defensive padding loses you jobs to faster competitors, and it trains customers to expect you to be late. Reckless optimism loses jobs after the fact; defensive padding loses them before you ever start. Either way you lose.

Quoting accuracy is a system property, not a talent

It's tempting to treat production quoting accuracy as a skill — find the estimator with the best instincts and lean on them. That works until that person is on vacation, or until the shop grows past what one person can hold in their head.

Accurate quoting is a property of the system, not the individual. A shop quotes well when the quote desk can see real load, when job durations reflect what actually happens on the floor, and when the buffer policy is written down instead of improvised. Remove any of those and accuracy collapses back to luck, no matter how good the estimator is.

That's the encouraging part. You don't need a better gut. You need a better picture — and a picture is something you can build.

Building quoting discipline: a working method

Closing the gap between the quote desk and the floor is a process change before it's a software change. This method works regardless of what tool you use:

1. Quote against committed load, not standard lead times. Before promising a date, look at what's already on the constraint the job will run through. If you can't see that in seconds, fix that first.
2. Identify your real constraint. In most job shops, one or two resources govern throughput. Quote against those, not the shop average. A date that ignores the bottleneck is a guess wearing a suit.
3. Separate "earliest possible" from "promised." The earliest a job could finish and the date you commit to are not the same number. Put your buffer between them, deliberately.
4. Quote and schedule from the same picture. This is the single highest-leverage fix. When the quoter and the scheduler look at the same live load, both reckless optimism and defensive padding disappear, because the guesswork they were each compensating for is gone.
5. Close the loop. Track quoted date against actual ship date. If you're consistently late on a class of work, your standard assumptions for that work are wrong — fix the assumption, not just the job. That feedback loop is the engine of steady on-time delivery improvement.

None of this requires enterprise software. A disciplined shop can run this with a clean spreadsheet and a daily standup. What it requires is that the promised date come from the shop's real load instead of a number someone hopes is still true.

A confident date is a decision, not a hope

A confident due date isn't optimism and it isn't padding. It's a number that comes from what your shop can actually do, given what it's already committed to do. Get that right and on-time delivery stops being a recovery exercise and becomes a quoting decision you make on purpose.

The discipline matters more than the tool — but the right tool makes the discipline close to automatic. When your quote desk and your schedule are the same live picture, a quoter can see committed load on every machine before promising a date, and the date stops being a hope. That's the core of what Visual Machine Scheduler does: a drag-and-drop view of every job on every machine, so the date you quote is the date the shop can keep.

Want to see your real load before you promise the next date? Start a free trial — no credit card required, 14-day trial.

If you'd rather start with the method than the software, the planning templates and tools in our store walk through the capacity-based quoting workflow above.