Print Job Cost Estimator

Ask a print shop for 500 business cards and the answer is one number. Behind it sit four kinds of cost and a single division, and the division matters most. Cards are not printed one at a time. They are printed in groups on a larger sheet, and how many fit on that sheet shapes nearly everything else in the quote.

Four costs and a divisor

The costs are easy to name. Paper is bought by the sheet or the pack. Ink or toner is bought by the cartridge, unless a service contract charges a flat amount for every printed side instead. Finishing is whatever happens after printing; the PrintWiki encyclopedia lists cutting, trimming, folding and binding, along with decorative work such as laminating. Setup comes last. Its trade name is makeready, the collective term for all of the operations necessary to set up a press, performed so that time and paper are not wasted once the job itself has started.

Setup is the reason unit cost falls with quantity. It is paid once, so its share of each piece is the setup cost divided by the number of pieces. Double the run and that share halves, while the paper and ink in each piece stay almost where they were.

Then there is the paper nobody keeps. Some sheets are spoiled in getting the job right, and the customer pays for them whether or not the quote says so. The contract terms of the U.S. Government Publishing Office are blunt about it: bidders are to include all items of cost for production and normal waste and spoilage in their bids.

A markup and a tax sit on top. None of this is hard arithmetic. The hard part is the two numbers that feed it: how many sheets the job needs, and how much ink each sheet takes.

What a cartridge yield measures

A cartridge is sold with a page yield, and cost per page looks like simple division, price over pages. The pages come from a test. ISO/IEC 19752 covers monochrome toner, ISO/IEC 19798 covers color toner, and ISO/IEC 24711 covers inkjet cartridges. The first of them was introduced in June 2004. Before it, makers had no shared method, and the usual description of a cartridge’s capacity was a number of printed pages at approximately 5 percent coverage.

The test is deliberately artificial. The monochrome standard fixes the preparation and the environment, the paper, the print mode and the test document. It asks for at least three printers with three cartridges each, bought on the open market, and it defines when a cartridge counts as finished. For color, ISO/IEC 24712 defines a suite of four customer type documents and one diagnostic page, used by both the color toner standard and the inkjet standard.

The page matters most, and yields are still quoted against the old figure. A United States government supply catalog of 2009 footnotes its laser cartridge yields as approximate pages based on 5 percent coverage, declared in accordance with ISO/IEC 19752, and marks its color cartridges as based on approximately 5 percent coverage too. A rated yield, then, is a count of pages printed at about 5 percent.

Real pages are not that page. Measure a flyer with a solid colored background and one channel alone can pass 80 percent. By simple proportion, a page at 80 percent takes sixteen times the toner of the test page. The same catalog warns that actual yield depends on the printer and the specific use. So the tool asks what coverage a yield was rated at, starts that field at 5, and scales the cost by measured coverage over rated coverage.

The standards say plainly what they are not. The 2025 edition of ISO/IEC 19752 notes that yield is one of the contributions to cost per page, that the document does not directly measure cost per page, and that cartridges are in most cases not the only contributors. It also sets aside production and large format machines, where the major cost of ownership is not the consumable. A yield exists to give clear and objective criteria for comparing one cartridge with another. It is a poor forecast of what any one job will use.

Coverage, measured the naive way

To know what a job will use, measure the artwork. The tool draws the image on a canvas and converts every pixel with the arithmetic the W3C publishes for uncalibrated color. Black is one minus the largest of red, green and blue, and each of cyan, magenta and yellow is what is left of its opposite once that black is taken out. The W3C’s own example pairs the color firebrick, which is red 178, green 34 and blue 34, with no cyan, 81 percent magenta, 81 percent yellow and 30 percent black. Averaged over the whole image, those four figures are the coverage per channel.

The W3C calls this conversion naive, and means it. It is necessarily approximate, the specification says, since it has no knowledge of the colorimetry of the inks, the dot gain or the colorimetry of the RGB space. A real printer converts through an ICC profile, a set of data that characterizes a device by mapping between its own color space and a standard connection space. Turning RGB into one printer’s CMYK goes through two such profiles. The format is defined precisely and the processing is not, which leaves room for variation between systems. So two printers given one file need not lay down the same ink.

A black only job is measured differently, as one minus the relative luminance that the W3C defines in its accessibility guidelines, a measure taken in linear light. The two methods disagree about gray. A mid gray pixel of value 128 counts as 49.8 percent black in the color conversion and 78.4 percent in the luminance one. They answer different questions, and both are estimates.

The sheet decides

Imposition is the old word for arranging pages on a sheet. The bookbinding dictionary of Etherington and Roberts defines it as the plan of arrangement of pages so that they read consecutively when the printed sheet is folded, and adds that the purpose of having different schemes is to obtain maximum economy in printing. For cards and flyers there is no folding, only repetition. PrintWiki’s entry on step-and-repeat describes exactly that: several copies of one label on a large stock, cut into individual labels during finishing.

The large stock is the press sheet, a full-size sheet of the actual paper used for the job, and common ones come in fixed sizes. The Printer Working Group’s list of standard media names has Letter at 8.5 by 11 inches, Ledger at 11 by 17, and sheets of 12 by 18 and 13 by 19 inches. On the metric side Markus Kuhn’s reference gives SRA3 as 320 by 450 mm, a raw format slightly larger than A3 that is meant to be cut down to its end size.

Three things take space away. A bleed is a printed image which runs off the edge of a page, made by deliberately trimming through the ink, so every piece needs a border of extra image. A gutter, in this tool, is the gap left between neighboring pieces. And a press cannot print everywhere: on a sheetfed press the gripper margin, usually 3/8 to 1/2 inch wide, is reserved for the mechanism that carries the sheet.

Here is the worked example, the one the tool shows when it opens. Take 3.5 by 2 inch cards on a Letter sheet with no margin, bleed or gutter. Upright, two fit across and five up the sheet, which makes 10. Turned a quarter turn, four fit across and three up, and the same paper gives 12, not 10. For 500 cards that is 42 sheets instead of 50, and 10.2 percent of each sheet is waste.

Add a 0.125 inch bleed, move to a 12 by 18 inch sheet with a 0.25 inch unprintable edge, and neither plain grid is best. Upright gives 21 and turned gives 20, but twenty turned cards leave a strip wide enough for three upright ones, and the mixed sheet holds 23. Counted as 21, 20 or 23 to a sheet, the same 500 cards need 24, 25 or 22 sheets. The parent sheet, not the ink, has just moved the paper bill by about a tenth.

What a browser estimate leaves out

The layout search is honest about its limits. It tries every arrangement that straight cuts across the whole sheet can separate, because a guillotine cutter brings one long knife down through the stack. Arrangements that interlock like a pinwheel sometimes hold one more piece, and the tool will not find them. Nor does it know the work-and-turn schemes that, according to the dictionary, cut the number of sheets to be printed by half on a larger and more expensive press.

The coverage figure is an estimate twice over. The conversion is naive, as already admitted, and coverage is not the same thing as ink. An inkjet printer also spends ink on itself. Ink dries in the nozzles of a head that sits unused, and nearly every inkjet clears them by firing the nozzles and wiping the head, with the ink itself as the cleaning fluid. That ink never reaches a page. None of that is visible in a picture of the artwork.

Scale matters too. Yields are rated on office machines; the color toner standard is not for printers whose minimum printable size is A3 or larger. The tool scales ink by area from a Letter or A4 page and cannot vouch for a wide format machine. Where a contract charges by the printed side, that rate replaces the whole ink calculation, and coverage stops mattering to the bill at all.

Every rate in the quote is typed by the person using it. There are no built-in prices. A price that was right somewhere last year is wrong here today, while a sheet count is true everywhere.

So treat the result as a well-argued first draft. The geometry is exact for the layouts it considers. The money is exactly as good as the rates typed in. The ink is a reasoned guess with its assumptions on show, and the first real sheet off the machine outranks it.