Your Job Costing and Your P&L Are Both Right — They're Just Answering Different Questions
There is a specific conversation that happens in converter plants a few times a year. The plant manager pulls up the MIS job margin report after a difficult month and says the numbers look reasonable. The controller pulls up the P&L and says they don't. Both of them are looking at real numbers. Neither of them is wrong. They are just looking at systems that were never designed to agree with each other — and most of the time, nobody has mapped where they diverge or what that gap is hiding about actual margin.
This piece walks one composite job through the cost lifecycle to show exactly where the divergence happens, why it happens by design, and what it obscures about the margin decisions you make every day.
---
The Job: A Short-Run Folding Carton With a Rerun
Imagine a 15,000-unit carton run for a regional food brand. It's a six-color job with foil stamp and glue, running on a 40-inch offset press. The customer ordered the same item eight months ago; you're pulling the original estimate and adjusting quantity. The job ships, invoices, and closes. Six weeks later, the customer calls back with a change order — slightly different copy, same structure — and you run it again as a rerun, 12,000 units.
That's not an unusual week in a folding carton shop. It's also a job that will touch every divergence point worth understanding.
---
Divergence Point 1: The Machine-Hour Rate That Was Set From Last Year's Assumptions
Your estimate prices press time at a Budgeted Hourly Rate — a BHR — built from a snapshot of your cost structure: equipment depreciation, maintenance contracts, labor burden, facility allocation, energy. That snapshot was accurate when it was taken. The question is when it was taken.
BHRs are often set once per year, sometimes less frequently. In the intervening time, equipment gets added or retired, maintenance costs change, labor rates shift, and energy costs move. The BHR on your estimating system is the denominator of all of that — and when it's wrong, every job estimated from it inherits the error silently.
For a long-run commodity job, a stale BHR might wash out. For a short-run carton with significant make-ready, where press time is a large fraction of total cost, a BHR that's 8–12% low means you've been quoting thin margins as full margins. You won't see it in any individual job report. You'll see it, eventually, in the P&L — or you won't, because the variance gets absorbed into overhead and disappears.
*Ask your numbers: When was the BHR on each press last recalculated against actual costs? How much has labor burden, maintenance spend, or energy changed since then?*
---
Divergence Point 2: Make-Ready and Setup Scrap — Invisible Unless Someone Badges It
Your estimate for the carton job carries a make-ready allowance: a number of sheets to reach color, a set-up time, a scrap percentage. That allowance was built from historical averages, or from the original job, or from the estimator's experience with similar work.
On the floor, what actually happens to make-ready time is determined by whether operators badge non-productive time against a reason code. If they do, setup time is visible. If they don't — and in many shops, they don't consistently — breakdown time, calibration, and changeover time get absorbed into whatever the press was running when the clock was ticking. The result is a permanently understated picture of true machine utilization, and a make-ready cost that the shop-floor system never cleanly captures.
Setup scrap compounds this. On high-SKU label and carton programs, a meaningful share of total substrate waste — industry practitioners cite figures in the range of 15–25% on high-SKU programs — is attributable to setup, not production. If that scrap is allocated to overhead rather than to the job, the job looks cleaner than it is, and the overhead rate quietly inflates.
For the rerun eight months later, your estimate reuses the original make-ready allowance. If the original allowance was wrong, the rerun inherits the error without anyone reviewing it.
*Ask your numbers: How is non-productive press time captured? Is setup scrap tracked separately from run scrap? Does the job record reflect what actually happened, or what the estimate expected?*
---
Divergence Point 3: Material Issued vs. Material Yielded — and What Happens to the Difference
Substrate and ink are booked in the MIS at estimated yield. What actually leaves the warehouse is issued quantity. Those two numbers are not the same job, and the gap between them — over-issue or under-issue — is reconciled at period close, not at job close.
Ink is the cleaner illustration: ink cannot be accurately attributed to individual jobs by design. When a color is mixed from existing stock and runs across multiple jobs in sequence, there is no clean method for dividing total ink consumed per job. The practitioner-recommended approach is to book estimated ink cost as actual at the job level, then reconcile total estimated ink cost against total ink purchases over a period — annually, in some plants. That means individual job ink margins are always theoretical. You are comparing your estimate to itself, not to what was actually consumed.
Substrate over-issue gets reconciled at period close, which means it lands in a period that may not match the period in which the job was invoiced. The job closes, revenue is recognized, and the material variance arrives later, in a period where the job is already gone.
*Ask your numbers: How is substrate over/under-issue handled at close? Does the variance get allocated back to the job, or does it disappear into a materials variance account? What does your ink reconciliation cycle look like — and how old is the last one?*
---
Divergence Point 4: Tooling Amortization Across Runs the ERP Sees as Separate
The carton job has a die. The die was charged to the customer on the first run. Eight months later, on the rerun, it runs again. In the intervening time there may have been a die remake, a plate change, or a revised cutting score. Those costs arrive as invoices — sometimes from outside vendors — after the jobs that incurred them have already closed.
The MIS and the ERP handle this differently. The ERP sees tooling costs as capital or prepaid expense, allocated by rule across time or across a predetermined number of runs. The MIS may or may not track tooling against specific jobs at all; in at least one major MIS platform, the recommended approach is a separate tooling cost/sales analysis rather than booking tooling costs against individual production jobs. The two systems are deliberately answering different questions.
What this means in practice: the job-level margin your MIS shows for the rerun does not include the full tooling picture. The ERP's view does — but it's allocated by a rule that may not reflect how the die was actually used or when the remake was incurred.
*Ask your numbers: How is die and plate cost tracked across reruns? When a remake is invoiced after job close, where does that cost land? Is tooling margin tracked at all, and at what level?*
---
Divergence Point 5: Gang-Run Cost Splitting
Your press runs two carton jobs together in a gang — different SKUs, shared substrate, shared finishing path. The press time and substrate are real and measurable. Allocating them to individual jobs is not.
There is no established industry convention for how to split gang-run costs when substrates, finishing paths, and run configurations differ across jobs in the gang. The approach used by any given plant is an internal rule — divide by impressions, by square inches, by units, by revenue — and each rule produces a different margin per job. The job that looks best under one allocation looks different under another.
This is not a software problem waiting for the right feature. It is a structural problem in the underlying cost model. The physical run and the job record are not the same unit of work.
*Ask your numbers: How are gang-run costs split? Is the allocation method documented and consistent? Do different estimators or different MIS users apply the same rule?*
---
Divergence Point 6: Unit-of-Measure Conversions That Quietly Change the Number
The press thinks in MSI or linear feet or impressions. The warehouse thinks in sheets or rolls. The ERP thinks in units or cartons or pallets. The customer invoice is in units.
Somewhere between the press room and the general ledger, conversions happen — and each conversion introduces a rounding decision, a yield assumption, or a waste factor. A substrate yield expressed in MSI converts to sheets, then to good units, then to invoiced quantity. At each step, the denominator of your cost-per-unit calculation shifts slightly.
For a short-run job, these small shifts in denominator are not trivial. A yield assumption that's 3% optimistic on a 15,000-unit run changes the cost per unit more than most estimators would expect.
*Ask your numbers: Trace the unit conversion from press output to invoiced quantity on a recent job. Where do yield assumptions enter? Where is the last place you can verify the conversion against something physical?*
---
Divergence Point 7: Post-Close Costs — Outsourced Finishing, Freight, and Change Orders
The carton job ships. The invoice goes out. The job closes in the MIS. Two weeks later, the outsourced foil-stamping vendor sends their invoice. The freight bill arrives after that.
These costs land in the ERP in a period when the job is already invoiced and the revenue is already recognized. There is no recovery path. The variance — the gap between what the estimate assumed and what was actually billed — either posts to a job that is already closed, or it disappears into a period cost that looks unrelated to the original job.
For the rerun, if the finishing vendor's rates have changed since the original estimate, and no one has updated the estimate, the rerun is priced off a cost structure that no longer exists.
*Ask your numbers: How are post-close vendor invoices handled? Do they post back to the job, or to a variance account? When were outsourced finishing and freight rates last updated in your estimating system?*
---
The Real Cost of the Divergence: Decisions Made on Numbers That Were Never Reconciled
None of the above is a failure of either system. The MIS job margin report is doing what it was designed to do: it runs on standard costs and gives you a fast, operational view of how the job performed against the estimate. The ERP P&L is doing what it was designed to do: it records actual financial transactions, matches revenue to cost in the correct period, and produces auditable financials. These are not the same job, and they are not supposed to be.
The problem is when the operational numbers are used to make decisions that require the financial numbers — and no one has mapped where they diverge.
The specific decision this matters for most is quoting. If your estimators are calibrating future quotes off past MIS job margins, and those job margins are running on stale BHRs, understated make-ready, theoretical ink costs, incomplete tooling, and missing post-close costs, then the estimate for the next job is wrong in the same direction as the last one. You are not learning from actuals. You are learning from a model that was built from assumptions and never corrected against what actually happened.
The detection mechanism in most plants is post-mortem: a monthly variance report, a job margin review run at period close, a controller who notices a pattern and asks questions. That is structurally late. By the time the pattern is visible, the jobs that created it are already invoiced, and the margin recovery window has closed.
For short-run and repeat work, the problem compounds silently. The original estimate — carrying the original BHR, the original make-ready allowance, the original substrate yield — gets reused without adjustment across every rerun. If the original estimate was wrong, every rerun is wrong in the same direction, and the divergence between your job costing and your P&L grows with each repeat cycle.
---
Questions Worth Asking Your Own Numbers
- Pick three closed jobs from the last 90 days. Compare the MIS job margin to the gross margin on the corresponding invoice line in the ERP. What is the gap, and what is it made of? - When was the BHR on each major cost center last updated? What has changed in actual costs since then? - How is setup scrap captured — at the job level, at the shift level, into overhead? Does the answer differ by press or by shift? - What happens to substrate and ink variance at period close? Does it trace back to a job, or does it disappear? - How are outsourced finishing and freight costs matched to jobs when vendor invoices arrive after job close? - For repeat work, what is the process for reviewing and updating an estimate before it is reused?
These are not questions with comfortable universal answers. They are questions worth sitting with your own numbers to answer specifically, because the gap between your job costing system and your P&L is real, it is structural, and it is the place where margin decisions get made on information that was never quite what it appeared to be.
---
*If any of these questions surface patterns worth thinking through, we're glad to compare notes. Twenty-five years in converter plants hasn't made us certain about everything — but it has made us precise about where the hard questions tend to live.*
If any of these questions surface patterns worth thinking through, we're glad to compare notes.