The four costs in a print, and the one everyone forgets
A print job has four direct costs. Material is the filament or resin the slicer says the plate consumes, including supports, brim and purge, priced at what you actually paid per kilogram delivered. Machine time is the wear you put on the printer, charged by the hour. Electricity is the average power draw multiplied by the hours and your tariff. Labour is every minute you touch the job — slicing, plate prep, part removal, support cleanup, sanding, packing.
Labour is the one that gets left out, and it is usually the largest line. Twenty minutes of hands-on work at $30 an hour is $10, which on a typical hobby-class part exceeds the filament, the machine time and the electricity put together. A shop that prices only material and "a bit for the machine" is paying its customers to take its parts away.
Machine time is the second most misunderstood. The printer is a wasting asset: every hour it runs uses up a finite quantity of hours before it is worn out or obsolete. Charging that hour is not padding — it is the mechanism by which the machine funds its own replacement. If a $500 printer will realistically print 4,000 hours, then 12.5 cents an hour is what it costs to own, and a 40-hour print carries $5 of it.
The formula: cost, then risk, then margin
Add the four direct costs to get the cost of one attempt. Then two divisions turn that into a price.
Divide by (1 − failure rate) to get the cost per good part. This is the step people get wrong by multiplying instead. If one print in ten fails, ten attempts yield nine sellable parts, so each sellable part carries ten-ninths of an attempt's cost: dividing by 0.90 gives a 11.1% uplift, not 10%. At a 20% failure rate the uplift is 25%; at 50% the cost doubles. The gap between multiplying and dividing widens sharply as the failure rate rises, which is exactly the regime where getting it right matters.
Divide by (1 − margin) to convert cost into price. Gross margin is measured against the selling price, so a 50% margin means cost is half of price and price is twice cost. Markup is the same money measured against cost, and the two numbers are never equal above zero: a 50% margin is a 100% markup, a 60% margin is a 150% markup. Quoting a "50% markup" when you meant a 50% margin leaves you at a 33% margin, a third of your intended profit gone.
What the formula deliberately excludes is anything downstream of the part: marketplace commission, payment processing, shipping, packaging and sales tax. Those attach to the transaction rather than the print, and they need to be added on top or your margin evaporates at checkout.
Worked example: a 10-hour, 100 g part
A customer wants one part. The slicer says 10 hours and 100 g of PLA. You paid $25/kg. Your $500 printer will run about 2,000 hours before you retire it, and you spend about 5 cents an hour on nozzles and build surfaces. It draws 120 W average, and your all-in electricity rate is $0.16/kWh. You spend 15 minutes on the job by hand and value your time at $30/hour. One print in twenty of this type fails. You want a 50% gross margin.
- Material. 100 ÷ 1,000 × $25 = $2.50.
- Machine rate. 500 ÷ 2,000 = $0.25/h depreciation, plus $0.05/h maintenance = $0.30/h. Over 10 hours: $3.00.
- Electricity. 120 W = 0.12 kW; 0.12 × 10 = 1.2 kWh; × $0.16 = $0.192.
- Labour. 15 ÷ 60 × $30 = $7.50.
- Cost of one attempt. 2.50 + 3.00 + 0.192 + 7.50 = $13.192.
- Failure allowance. 13.192 ÷ 0.95 = $13.89 per good part.
- Price at 50% margin. 13.89 ÷ 0.50 = $27.77, of which $13.89 is profit.
Read the split: labour is 57% of the cost of the attempt, material is 19%, machine time 23%, electricity 1.5%. That ranking is typical of small FDM parts and it is why "filament cost × 3" — which would have produced a $7.50 price here — bankrupts print shops.
How to read the result and where to push
Cost per print hour is the number to watch across jobs. It tells you what an hour of your capacity costs to supply, and therefore the minimum an hour of machine time must earn. If your cost per print hour is $1.40 and a job ties up the machine for three days to produce a $20 part, that job is losing to almost anything else you could have printed.
Long prints shift the balance towards machine and energy. Hands-on time barely changes between a 2-hour print and a 40-hour print, so labour's share falls as print time rises, while machine and electricity scale linearly with hours. On multi-day prints, machine depreciation usually overtakes labour as the largest line.
Big flat parts shift it towards material. A 900 g part at $25/kg carries $22.50 of filament, and material becomes the line worth optimising — through infill, wall count and orientation rather than through cheaper filament, since bargain filament raises the failure rate and failures are expensive.
The failure rate is the highest-leverage input. Going from 20% failures to 5% cuts the risk uplift from 25% to 5.3%. That is a bigger saving than any plausible reduction in filament price, and it comes from bed adhesion, dry filament and first-layer discipline rather than from spending money.
When you sell through a marketplace, put this price through the Etsy fee calculator or the eBay fee calculator before you commit to it, and use the markup vs margin calculator if your supplier quotes you in markup. For the material side in isolation, the filament length and weight calculator converts between grams, metres and dollars.
Machine rate per hour by purchase price and expected life
| Printer price | 2,000 h life | 4,000 h life | 8,000 h life | 20,000 h life |
|---|---|---|---|---|
| $300 | $0.150 | $0.075 | $0.038 | $0.015 |
| $500 | $0.250 | $0.125 | $0.063 | $0.025 |
| $1,000 | $0.500 | $0.250 | $0.125 | $0.050 |
| $2,500 | $1.250 | $0.625 | $0.313 | $0.125 |
| $6,000 | $3.000 | $1.500 | $0.750 | $0.300 |
| $20,000 | $10.000 | $5.000 | $2.500 | $1.000 |
Each cell is simply price ÷ life hours. Industrial machines justify their rate through reliability and unattended running, not through a lower hourly cost.
Margin is not markup
Margin measures profit against the price; markup measures the same profit against the cost. Price = cost ÷ (1 − margin), and price = cost × (1 + markup). A part costing $10 sold at a 50% margin is $20; the same part at a 50% markup is $15, which is only a 33% margin. If a customer, a supplier or a marketplace quotes a percentage, always ask which base it is on before you agree to it.
Pricing mistakes that kill print businesses
- Charging filament times three. It ignores machine wear, electricity and every minute of your time, and it produces a price that is roughly a quarter of cost on small parts.
- Multiplying by the failure rate instead of dividing. At 20% failures the correct uplift is 25%, not 20%. The error grows as failures grow.
- Pricing your own labour at zero. If you would not do the cleanup for free for a stranger, do not do it for free for a customer.
- Forgetting marketplace and payment fees. Commission, listing and processing fees typically remove a tenth or more of the sale price, and they come out of your margin, not the customer's.
- Quoting before slicing. Print time and material weight can double with a change of orientation or support strategy. Slice first, then quote.
- Ignoring the queue cost of long prints. A 60-hour print does not just cost 60 machine hours; it blocks every other job you could have run in that window.
Resin, SLS and where this model stops working
The cost structure here is written for FDM, but the arithmetic transfers with two substitutions. For MSLA resin printing, replace filament price per kilogram with resin price per litre and material use in millilitres, and add two lines the FDM model does not have: isopropyl alcohol for washing and the LCD panel, which is a wear item with a life measured in hundreds of hours and needs its own per-hour charge alongside depreciation. Resin post-processing also takes far more hands-on time — wash, cure, support removal and sanding — so labour's share rises further.
For SLS and MJF, cost is driven by powder and packing density rather than by time, because the machine builds a whole chamber at once. The right unit there is cost per litre of build volume divided among the parts nested in it, and refresh ratios for used powder complicate the material line. This calculator will not model that honestly.
Two structural costs sit outside this model entirely. Overhead — rent, insurance, software subscriptions, the accountant — is not caused by any one print, and is normally recovered by loading it into the machine rate or the labour rate rather than by adding a line per job. And design time, when you are modelling the part rather than printing someone else's file, is a separate service that should be quoted separately; the freelance hourly rate calculator is the right tool for that half of the job.
Finally, a price is not a quote. This tells you the floor below which you lose money. What the market will pay depends on lead time, finish quality, whether you carry the risk of a refit, and what the customer's alternative is. Knowing your floor is what lets you discount deliberately instead of accidentally.
