Why Is My Printing Job Losing Money?

The Job That Looked Fine on the Quote

The invoice comes back from the press floor and the number doesn't match. The job quoted at $2,140 actually cost something closer to $2,600 to run — you know this because the material draw, the labor hours, and the waste roll all got logged against the job number. Nothing on the quote was obviously wrong. The colour count was right, the substrate price was current, the run length matched the PO. So you sit there re-adding the same numbers, looking for the mistake, and the mistake isn't there. The formula was fine. One or two of the inputs to the formula weren't.

That's the pattern behind almost every job that loses money on a label press: the estimate itself is rarely broken end to end. It's usually one variable — a waste percentage, a shop rate, an unbilled change — that was wrong going in, and wrong in a way that's invisible until the job is closed out. By the end of this piece you'll have a short list of the specific places margin actually escapes on a label job, and a way to check your own last loser against each one instead of re-deriving the whole quote from scratch.

Where a Label Job Actually Loses Money

Before diagnosing your own job, it helps to know the short list of places a per-label number usually breaks. In a flexo shop quoting off a spreadsheet, it's almost always one of these four:

1. Makeready waste estimated as a guess, not a measurement — the plate-up, colour-registration, and print-to-spec waste at job start was assumed at some round percentage instead of tracked from the last comparable job.
2. A shop rate that no longer reflects the real cost of running the press — the hourly rate baked into the quote hasn't been rebuilt against current overhead in a while.
3. Die, plate, or tooling cost amortised across the wrong run length — a die cost spread over an assumed reorder volume that never happened.
4. Scope that changed after the quote went out, without the quote changing with it — a proof round, a die change, a late art revision that ate press time nobody re-priced.

Any one of these, alone, is usually a small miss. Stacked — a slightly-low waste assumption and a shop rate that's drifted and one unbilled revision — they add up to exactly the kind of job that looks fine on paper and loses money on the floor. Tracing which one (or two) hit your job is the whole exercise; you rarely need to distrust the entire quote.

The Makeready Waste You Estimated in Theory, Not in Minutes

Makeready waste is the material burned getting a job into registration and up to colour before a single sellable label rolls off the press — plate mounting, impression setting, colour matching, and the substrate consumed at press start and press end. It's a real cost, it's substrate-and-time, and on a narrow-web flexo press it can be a meaningful share of a short run's total material draw.

The trap is estimating it once, as a flat percentage, and reusing that percentage for every job regardless of colour count, substrate, or die condition. A four-colour job with tight registration tolerances on a difficult substrate does not waste the same amount as a two-colour job on a forgiving stock, and a shop that quotes both at the same "10% waste" line is quietly overcharging one and underpricing the other. The underpriced one is the job that loses money — not because the waste assumption was crazy, but because it was applied uniformly to jobs that don't behave uniformly.

The fix isn't a better guess. It's measuring actual makeready waste off completed jobs, by colour count and by press, and feeding that measured number back into the next quote for a similar job — which only works if the estimated waste and the actual waste are captured on the same job record where you can compare them. We go deeper on the mechanics of makeready waste specifically — what counts, what doesn't, and how to build a real waste table — in makeready waste on a flexo label job.

A Shop Rate That's Wrong by a Little, Every Time

Every quote runs on an assumed hourly cost to operate the press — labor, overhead, depreciation, utilities, all loaded into one number per hour. If that number is stale, every job priced off it is wrong by the same small amount, in the same direction, and because it's consistent, it's the hardest kind of error to spot. A one-off mistake shows up as an outlier. A shop rate that's 8% low shows up as a shop that's a little less profitable than it should be on almost everything — which looks like normal business, not a costing problem.

Shop rates drift for boring reasons: a new press lease changed the depreciation load, a health-insurance renewal changed the labor burden, electricity went up, and nobody rebuilt the per-hour number to reflect it. The rate that's baked into your spreadsheet template might be a year or two old, copied forward from quote to quote because it was never wrong enough, on any single job, to notice.

This is worth rebuilding on a schedule, not waiting for a loss to trigger it — the actual method for constructing a defensible loaded hourly rate (labor, overhead, and depreciation, not just a labor wage) is covered in building a budgeted hourly rate for a print shop.

The Changes Nobody Re-Quoted

The fourth culprit isn't a bad input — it's a missing one. A customer asks for a die change after the quote is approved. A proof gets rejected and a second round of plates gets pulled. An art file arrives with more colours than the RFQ specified. Somebody on the floor makes the call to run it anyway, because holding the job for a formal re-quote costs more in schedule than the change is worth, and the extra time or material gets absorbed into the job without ever touching the price.

None of that is unreasonable in the moment. It's also exactly how a correctly-priced quote turns into a job that loses money — the quote was right for the job that was quoted, and wrong for the job that actually got run. The gap between those two jobs is real cost that never made it back to the customer or the invoice.

The only reliable defense is closing the loop: logging what actually happened against the job (hours, material, waste) against what the quote assumed, and reviewing the delta before the next similar job goes out the door — which is the practice generally called estimated-vs-actual job costing. We cover the full mechanics of setting that comparison up, including what to log and how often to review it, in estimated vs. actual job costing for label converters.

How to Trace Your Own Loser: Estimated vs. Actual

Here's the actual diagnostic, in order, for the job sitting in front of you right now:

1. Pull the original quote and the job's actual numbers side by side — material consumed, hours logged, waste roll measured. If you don't have both in one place, that's the first problem to fix, not the last.
2. Check waste first. Compare the waste percentage the quote assumed against the waste percentage the job actually produced. If actual waste was meaningfully higher, you've likely found it — or found half of it.
3. Check the shop rate next. Does the hourly rate in the quote match a rate you'd defend today, rebuilt from current overhead? If it's stale, every job priced off it is quietly light, not just this one.
4. Check for unbilled scope. Was there a die change, a reprint, a revision round, anything added to the job after the quote was locked that never generated a change order?
5. Only then, re-check the base math — press speed by colour count, plate cost, die amortisation, substrate MSI pricing — because in most jobs that lose money, the base formula was never the problem.

A job rarely loses money because the whole quote was wrong. It loses money because one input — waste, rate, or scope — was wrong, and nothing caught it before the job closed out.

If margin erosion isn't a one-job problem for your shop but a pattern across several jobs a month, it's worth stepping back from the single job and looking at where margin is escaping systemically — covered in label converter margin erosion, and if you want the full mechanics of how a flexo quote is built from press speed, plate cost, makeready, and overhead in the first place, that's laid out in the complete guide to label estimating.

Building the Habit Before the Next Job Surprises You

The shops that stop losing money on individual jobs aren't the ones with a better gut feel for pricing — they're the ones who log estimated and actual numbers on every job as a matter of habit, so a variance shows up as a data point instead of a surprise three weeks later at invoicing. That habit is easier to build with a structure already in place than to invent from a blank sheet under deadline pressure.

If you want a starting point for tracking estimated versus actual by job — waste, hours, and material variance in one sheet — the Label Converter Job Costing Workbook is built for exactly that comparison, and it's a reasonable first step before moving the same comparison into a system that logs it automatically as jobs close.