The Complete Guide to Estimating Narrow-Web Label Jobs (Flexo and Digital)

The Estimate That Didn't Match the Invoice

A shop we'll call typical here just added its first UV digital press alongside three flexo units. Two weeks later, a repeat customer calls in with a reorder: same label, same artwork, but the run has dropped from 15,000 to 2,500 because the customer is testing a new SKU. The estimator pulls up last year's flexo quote, discounts it by a gut-feel percentage for the smaller run, and sends it out. The order ships. Six weeks later, at invoicing, someone finally runs the numbers on what that job actually cost to produce — plate charges that didn't amortize the way they used to, a makeready that ate a much larger share of a much shorter run, and a digital press sitting idle the whole time that could have produced the job at a lower total cost with no plates at all. The margin is gone, and nobody can say exactly when it disappeared, because the estimate and the actual cost were never built the same way.

This is not a story about a bad estimator. It's a story about a spreadsheet — or a habit — that was never built to answer the question "which process, and at what price, actually makes money on this job?" This guide walks through the mechanism behind every number in a label quote, flexo and digital, so that by the end you can build a defensible per-label price for any job that lands on your desk and know which press should run it.

How a Per-Colour Plate Cost and Makeready Waste Change the Per-Label Number

Flexo pricing starts with a press-speed curve, not a single number. A press doesn't run at one speed regardless of the job — speed drops as colour count rises, because more colour stations mean more registration risk and more setup complexity. A one-colour job might run near the press's top rated speed; a seven-colour job on the same press runs meaningfully slower. Any flexo cost model that ignores this and applies one flat run-speed to every job is quietly overcharging simple jobs and underpricing complex ones.

On top of run speed sits per-colour plate cost. Each colour on the job requires its own plate, and that plate cost is a fixed charge that has to be recovered somewhere — either billed directly to the customer as a one-time plate charge, or amortized into the per-label price across the run length. This is where run length starts to matter enormously: a $600 total plate cost spread across 50,000 labels adds about $0.012 to each label; spread across 2,500 labels it adds about $0.24 — twenty times more per label, even though the plates cost exactly the same to make. (These are round, illustrative numbers to show the mechanism, not a published rate card.)

Then there's makeready waste — the substrate consumed while the press is being brought up to registration, colour, and pressure before it starts producing sellable labels. Makeready is a largely fixed quantity of wasted material and time regardless of run length, which means it behaves exactly like the plate cost: a fixed cost that a longer run absorbs more easily than a short one. A press that wastes, say, 300 feet getting to colour on every job costs a 10,000-label run almost nothing per label, and costs a 500-label run a great deal per label.

The last piece is die amortization. The cutting die used to shape the label is itself a capital cost, typically owned by the converter or billed once to the customer, and it needs to be spread across the labels it will realistically produce before it needs to be replaced or before the job reorders. Getting a flexo quote right means carrying all four of these — press-speed-by-colour-count, per-colour plate cost, makeready waste, and die amortization — through to a single per-label number, not guessing at a discount off last year's price. If you want the full worked build-up of a flexo quote step by step, see how to estimate a flexo label job.

Cost Per MSI: The Substrate Math Behind Every Flexo Quote

Underneath the plate and makeready math sits the substrate itself, and in the US label industry that's priced by MSI — thousand square inches. MSI pricing exists because label substrate isn't sold or consumed by the linear foot; it's consumed by area, and a label's width, repeat length, and web layout all determine how much of that thousand-square-inch unit a given order actually uses. Two jobs that run the same number of labels can consume very different amounts of substrate if one label is a small rectangle and the other is a large wrap-around shape with more waste around the die-cut edges.

Getting MSI cost right means starting from the substrate's cost per MSI (from the material library), calculating the actual area consumed by the label footprint including gaps and trim, and multiplying through — before a single flexo-specific charge (plate, makeready, die) is even added. Skip this step, or eyeball it, and the whole estimate is built on a wrong foundation, no matter how carefully the plate and makeready math that follows is done. A full walkthrough of the MSI calculation, with worked examples, lives at cost per MSI for labels.

Once substrate cost, plate cost, makeready, and die amortization are all rolled up, the last input is the shop's own hourly rate — press time plus a loaded overhead figure that recovers rent, utilities, and support staff, not just the operator's wage. Press operator labor itself is a real, measurable cost: the U.S. Bureau of Labor Statistics reports a mean wage of $21.37 per hour ($44,450 per year) for printing press operators nationally as of May 2023. That figure is only the operator wage, though — a shop's true loaded hourly rate for quoting purposes needs to layer in overhead on top, and every shop's overhead structure is different enough that there's no single number that applies universally. The discipline that matters is consistency: use the same loaded rate every time you quote, so that press-hour cost isn't quietly drifting between estimates.

Three Ways to Cost a Digital Label Job

Flexo has one cost mechanism with several moving parts. Digital label printing has three genuinely different cost mechanisms, and confusing them is one of the fastest ways to misquote a digital job.

LEP (toner-based digital, like HP Indigo) is priced by the click. A click charge is a per-impression cost, typically tiered by volume and negotiated with the press manufacturer or reseller, and — this is the part that surprises estimators coming from flexo — it does not change based on how much ink coverage the label actually has. A label that's 90% solid coverage costs the same click charge as a label that's 10% coverage, because the charge is per impression, not per unit of ink consumed. This is structurally different from every flexo cost driver above, where more coverage and more colours generally do cost more. It also means LEP economics reward jobs with heavy, complex artwork relative to a flexo job of equivalent complexity, because the flexo plate and colour-station costs scale with complexity while the LEP click charge doesn't.

UV inkjet is priced by measured ink coverage, which is closer to how a flexo estimator's intuition already works — more ink laid down costs more, and white ink underprint in particular typically carries its own surcharge and a speed penalty because white ink lays down heavier and the press has to slow to accommodate it. Getting a UV inkjet quote right means actually measuring or estimating the coverage percentage of the artwork, not assuming a flat ink cost the way an LEP job would.

Subscription and allocation models are a third pattern, more common with certain digital press arrangements: a fixed periodic fee is paid regardless of volume, and that fixed fee then needs to be allocated across whatever labels are actually produced in the period to arrive at a true per-label cost. A subscription that costs the same whether the press runs 5,000 or 50,000 labels in a month produces wildly different per-label overhead depending on how busy that press actually is — which means the per-label cost of a subscription-model job isn't fixed at all; it depends on total plant throughput, not just the individual job.

Treating all three of these as "the digital price" is a common and expensive mistake. A shop quoting an LEP job using coverage-based logic, or a UV inkjet job using a flat click-style logic, will misprice it in a predictable direction every time. The full comparison, with the math behind each mechanism, is at digital label press cost models compared.

It's also worth noting the broader trend behind why this matters more than it used to: digital label printing is growing faster than the overall label market. Mordor Intelligence projects the U.S. print-label market's digital segment growing at 3.21% CAGR through 2031, and separately projects inkjet specifically as the fastest-growing label printing process at 5.3% CAGR through 2031 — both faster than flexo's growth in the same window. Variable-data label printing, which depends on exactly these digital mechanisms, is projected by Mordor Intelligence to grow from USD 20.53 billion in 2025 to USD 29.37 billion by 2031, a 6.14% CAGR. A 2022 industry survey covered by Packaging Impressions found 38% of converters and printers reported increased label versioning over the prior 24 months — a trend that pushes more jobs toward shorter, more customized runs where digital cost mechanics matter more, not less.

The Value Crossover Point: Where Flexo and Digital Meet

Every narrow-web label estimating guide eventually arrives at the same question: at what volume does flexo start beating digital, or digital start beating flexo? The honest answer is that there's no single universal number — the crossover point moves with colour count, label geometry, special effects like metallics, and the shop's own configured plate and makeready costs. But the concept itself is real, well understood industry-wide, and worth building into every quote rather than guessing at.

The underlying logic is straightforward once the flexo cost stack above is understood: flexo carries fixed costs (plates, makeready, die) that get spread thinner as run length grows, so its per-label cost curve declines steeply with volume and eventually flattens out low. Digital, in any of its three cost forms, carries closer to zero fixed setup cost, so its per-label cost is close to flat regardless of volume — high relative to flexo at long runs, low relative to flexo at short runs. The crossover is simply the volume where those two curves intersect.

There is no universal crossover point that applies to every press, substrate, and colour count combination — it should be calculated for the job in front of you rather than looked up as a single number. What the broader market data does show is the shape of the effect: Mordor Intelligence notes that flexo delivers a low cost-per-thousand beyond roughly 10,000 units in the broader U.S. print-label market, which is consistent with the same underlying logic from the other direction: flexo's fixed-cost curve keeps flattening as volume climbs well past the crossover zone.

What this means practically: don't quote a job on "we've always run jobs like this on the Mark Andy" — quote both processes side by side, using each one's real cost mechanism, and let the crossover fall out of the math rather than out of habit. The full mechanics of finding your shop's own crossover point, with worked comparisons, are at flexo to digital crossover quantity.

The crossover isn't a fixed number you memorize — it's a calculation you run, because it moves every time colour count, geometry, or plate cost changes.

This distinction matters more as the underlying market shifts. Apex International's analysis, covered by Label and Narrow Web, projects the global flexo press market growing from USD 9.93 billion in 2025 to USD 12.49 billion by 2030, while the global digital press market grows from USD 22.0 billion to USD 36.9 billion in the same window — both processes are growing, but digital press capacity is scaling faster, which means more shops will have both processes on the floor and more jobs will need this side-by-side comparison, not fewer.

From Estimate to Actual: Why Job Costing Closes the Loop

An estimate is a hypothesis. It says "this job should cost X to produce, given the plate cost, makeready waste, run rate, and overhead we assumed going in." The only way to know whether that hypothesis was right is to compare it to what the job actually cost once it ran — actual press time, actual material consumed, actual makeready waste, actual click or coverage charges if it ran digital.

Most small converters don't close this loop consistently. It's not that estimated-vs-actual comparison is hard conceptually — it's that doing it by hand, job after job, in a spreadsheet that wasn't built for the comparison, is tedious enough that it gets skipped when the shop is busy, which is exactly when it matters most. The result is that a shop can carry a systematic pricing error — an underpriced plate charge, an overly optimistic run-speed assumption, a die amortization schedule that never gets revisited — for months without anyone noticing, because nothing forces the estimate and the actual back into the same room.

Closing that loop is also the only reliable way to know whether your shop's own crossover point, hourly rate, and makeready assumptions still match reality. A shop that reviews estimated-vs-actual on every job builds, over time, a genuinely calibrated cost model instead of one anchored to whatever assumptions were true when the spreadsheet was first built. The mechanics of setting this comparison up, and what to do when a variance shows up, are covered at estimated vs actual job costing.

Building a Narrow-Web Label Estimating Guide You Can Trust

None of the math above is exotic. Press-speed-by-colour, per-colour plate cost, makeready waste, die amortization, and MSI substrate pricing are all teachable, well-understood mechanics — the same is true of click charging, coverage-based ink cost, and subscription allocation on the digital side. What separates a shop that prices consistently from one that guesses off last year's quote isn't access to secret information; it's whether the same mechanism gets applied the same way, every time, across every job and every press on the floor.

The label market itself gives every shop good reason to get this right. Future Market Insights projects the U.S. label market growing from USD 16.1 billion in 2025 to USD 23.2 billion by 2035, a 3.7% CAGR, with pressure-sensitive labels holding 39.4% of U.S. label demand in 2025. Flexo remains the dominant process — Harper Corporation puts flexo's share at roughly 80% of labels printed, while Mordor Intelligence separately puts flexo's 2025 U.S. print-label market share at 45.63% — a growing market where both the volume and the process mix keep shifting under an estimator's feet.

A spreadsheet built by one estimator, on one afternoon, three years ago, generally can't keep up with that — not because Excel is a bad tool (it's free, familiar, and it works, right up until a cell reference is wrong or the person who built the formulas leaves), but because it was never designed to carry a full flexo cost stack, three separate digital cost mechanisms, and a crossover comparison side by side, consistently, across every quote a shop produces. Building that discipline by hand is possible. Building it consistently, across every estimator on staff, every job, every time — that's the harder problem, and it's the one worth solving next.

If you want the underlying formulas in one reference rather than scattered across a workbook, the Estimating Reference & Formula Guide collects the flexo, MSI, and digital cost-model math from this guide into a single working document. And if you'd rather get the next piece of this series — the worked flexo build-up, the MSI walkthrough, the digital model comparison, and the crossover calculator — as it's published, that's what the newsletter is for.