3D printing beginner mistakes that cost money, with the math
Every figure below is a worked example with its inputs shown, not a market average. Swap in your own numbers and the reasoning still holds.
The inputs behind the math
- Filament: a $20 spool of PLA per kg, or $0.02 per gram.
- Machine hour: $0.17, adding the wear of a $600 printer over a 6,000-hour life ($0.10), maintenance ($0.05) and 120 W of electricity at $0.17/kWh ($0.02).
- Labor: $30 per hour.
- The example part: 85 g of PLA, 6 hours of printing and 20 minutes of prep and finishing. It costs $1.70 in material + $0.12 in electricity + $0.90 in machine time + $10.00 in labor + $0.27 failure reserve (10%) = $12.99. With a 60% markup on cost, $20.78.
Mistakes at the printer
1. Not drying filament
What happens: wet filament prints with stringing, bubbles and weak layers, and the part looks bad or breaks in use.
What it costs in the example: a 250 g, 8-hour PETG print (priced like PLA) lost: 250 g × $0.02 + 8 h × $0.17 = $6.36. Drying the spool for 8 hours in a 100 W dryer costs $0.14 of electricity.
How to avoid it: dry PETG, TPU and nylon before printing and store them sealed with desiccant. Temperatures and times for each material are in how to dry wet filament.
2. Guessing the first layer
What happens: with the nozzle too high (wrong Z offset), the first layer does not stick and the part comes loose mid-print; too low, the nozzle drags on the plate and can clog.
What it costs in the example: the example part lets go halfway: 42.5 g × $0.02 + 3 h × $0.17 = $1.36, plus the plate if the nozzle scratches the surface.
How to avoid it: calibrate the Z offset or run the printer's calibration, wash the plate and watch the first few minutes. The step by step is in first layer not sticking.
3. Prying parts off with a metal scraper
What happens: forcing a metal scraper under a part on a hot plate, at a steep angle, can scratch or chip the PEI, and every mark becomes a spot where the first layer will not grip. Two related mistakes: Bambu Lab warns that acetone damages textured PEI, and Prusa says PETG can damage smooth PEI if printed without glue.
What it costs: a new build plate, whose price varies a lot from printer to printer, plus the prints that fail before you spot the damage.
How to avoid it: let the plate cool and flex it, as Prusa advises; reach for the scraper only when flexing is not enough, and keep it almost flat. Clean with warm water and dish soap without moisturizers, as Bambu Lab recommends, and put glue stick on smooth PEI before printing PETG.
4. Skipping maintenance
What happens: a worn or half-clogged nozzle, a loose belt and dirty rods turn into failures that look random. Abrasive filament (carbon or glass fiber, metal-filled or glow-in-the-dark) wears out a regular nozzle quickly, according to Bambu Lab.
What it costs: a new nozzle or hotend, which varies by model, plus the prints lost before you find the cause. The calculator sets aside $0.05 per machine hour for maintenance, $300 over 6,000 hours; if your spare parts cost more, raise that number.
How to avoid it: follow the manufacturer's maintenance schedule (Bambu Lab and Prusa publish theirs) and use a hardened steel nozzle for abrasive filament.
Mistakes in the slicer
5. 100% infill "for strength"
What happens: the part uses far more filament and time, and strength does not rise in the same proportion. For parts that carry a load, extra walls usually add more strength per gram than extra infill.
What it costs in the example: working from the volume (3 walls of 0.42 mm, 0.8 mm top and bottom, PLA at 1.24 g/cm³), a 50 × 50 × 20 mm block weighs 62 g solid and about 18 g at 15% infill. That is 44 g more: $0.88 per part, or $44 over 50 parts, before the extra print time.
How to avoid it: start from the slicer default (15% in Bambu Studio and PrusaSlicer, 20% in Cura), add walls, and raise infill only if a test calls for it. See the guide to infill percentage.
6. Supports everywhere
What happens: supports set to Everywhere (in Cura) grow even on top of the part itself; they waste filament and removal time, and leave marks.
What it costs in the example: inputs: a miniature with normal supports everywhere uses 35 g of support and 15 minutes of cleanup; reoriented, with tree supports from the build plate only, 8 g and 5 minutes. The difference is 27 g × $0.02 + 10 min × $30/h = $5.54 per part, or $277 over 50 parts.
How to avoid it: rotate the part to reduce overhangs and use tree supports from the plate only (On build plate only in Bambu Studio and OrcaSlicer, Support on build plate only in PrusaSlicer, Touching Buildplate in Cura). See tree vs normal supports.
7. Ultra-fine layers for everything
What happens: 0.08 mm layers smooth out gentle curves, but multiply the layer count. On a functional part, the difference usually matters little.
What it costs in the example: input: a part that prints in 4 hours at 0.20 mm takes 9 hours at 0.08 mm (2.5 times the layers; check the time in your slicer). The extra 5 hours cost 5 × $0.17 = $0.85 per part: $42.50 and 250 printer hours over 50 parts.
How to avoid it: use 0.20 mm as the default and fine layers only on small, detailed parts; when only one section needs detail, use variable layer height (Variable layer height in Bambu Studio and PrusaSlicer, Adaptive Layers in Cura). See the guide to layer height.
8. Printing a big part without a test
What happens: a tight fit, an undersized hole or a lifting corner only shows up at the end of a many-hour print.
What it costs in the example: input: a 500 g, 18-hour part with a fit that comes out too tight. 500 g × $0.02 + 18 h × $0.17 = $13.06. A test slice with that fit, 10 g and 40 minutes, would have cost $0.31.
How to avoid it: cut the model in the slicer (the Cut tool) and print only the critical section. On big ABS or ASA parts, test a corner too, since that is where warping shows up first.
Buying and pricing mistakes
9. Buying ten colors at once
What happens: the color you use runs out, and the rest sit open, absorbing moisture, waiting for a project that may never come.
What it costs in the example: ten $20 spools are $200 up front. If three colors get used and seven sit, that is $140 tied up, and every idle spool will need drying before use.
How to avoid it: buy the colors your next projects need and leave the rest until an order calls for them.
10. Not counting failures in the price
What happens: every printer fails now and then. If the price does not cover it, each failure comes out of your profit.
What it costs in the example: with a 10% reserve on material, electricity and machine time, the example part carries $0.27 per sale. If failures happen at that rate and the price has no reserve, $13.50 every 50 sales comes out of your pocket.
How to avoid it: log how many prints fail each month and use that rate as your reserve. The guide to failure rate pricing walks through the math.
11. Selling at twice the material cost
What happens: the "double the filament" rule ignores electricity, machine time, failures and your time, which in the example part are worth more than the filament itself.
What it costs in the example: the part's material is $1.70, so double is $3.40. The full cost is $12.99: every sale loses $9.59, or $479.50 over 50 sales.
How to avoid it: add up every cost and put your margin on top. The guide to how much to charge explains the math, and keychain pricing applies the same logic to small parts.
12. Forgetting marketplace fees
What happens: the fee comes out of the price the customer pays. Listing at your direct-sale price lets the fee eat the margin.
What it costs in the example: input: a 20% total fee. Listed at $20.78, the part pays $4.16 in fees and leaves $16.62; profit drops from $7.79 to $3.63. Over 50 sales, $208 comes out of the margin.
How to avoid it: divide the price by what is left after the fee: $20.78 ÷ 0.80 = $25.98. Fees on Etsy, eBay and Amazon Handmade depend on the platform and the sale; the guide to selling fees covers them.
Summary: what each mistake costs in the example
| Mistake | Cost in the example | How to avoid it |
|---|---|---|
| 1. Wet filament | $6.36 per lost print (250 g, 8 h) | Dry it ($0.14) and store it sealed |
| 2. Guessing the first layer | $1.36 per part that lets go halfway | Calibrate Z offset, wash the plate |
| 3. Metal scraper on PEI | A new plate (price varies) | Let it cool and flex the plate |
| 4. No maintenance | Nozzle or hotend (price varies) plus failures | Follow the manufacturer's schedule |
| 5. 100% infill | $0.88 per part; $44 over 50 | Slicer default and more walls |
| 6. Supports everywhere | $5.54 per part; $277 over 50 | Reorient, tree supports |
| 7. 0.08 mm layers on everything | $0.85 and 5 h per part; $42.50 over 50 | 0.20 mm as the default |
| 8. Big part, no test | $13.06 per lost part | A $0.31 test slice |
| 9. Ten colors at once | $140 tied up in 7 spools | Buy as orders come in |
| 10. Failures not priced | $0.27 per sale; $13.50 over 50 | A reserve at your real rate |
| 11. Twice the material cost | $9.59 loss per sale; $479.50 over 50 | Full cost plus margin |
| 12. Forgotten fee | $4.16 per sale; $208 over 50 | Divide by what the fee leaves |
The order in the chart depends on the inputs: if your labor rate is high, supports climb; if you sell cheap parts, the fee weighs more. The pattern does not change: a mistake that repeats on every part, even a small one, adds up to more than a one-off failure.
Frequently asked questions
Which of these mistakes should I fix first?
The pricing ones (10, 11 and 12), because they repeat on every sale and take an afternoon to fix with the calculator. Next, drying filament and dialing in the first layer, which prevent two common causes of failed prints.
Does a printer with auto bed leveling prevent first-layer mistakes?
It helps but does not solve it alone. Leveling compensates for an uneven bed, and some printers also set the Z offset by themselves; even so, a greasy plate or wet filament still makes parts let go.
Do I need a hardened steel nozzle?
Only for abrasive filament: carbon or glass fiber, metal-filled or glow-in-the-dark. For regular PLA, PETG, ABS and TPU, the nozzle that came with the printer does the job.
How big a failure reserve should I use without any history?
Use a provisional reserve, like the 10% in this guide's examples, and replace it with your real rate after logging about 30 prints: if 3 failed, that is 10%; if 6 failed, 20%.
How do I tell whether a failure is the printer or the filament?
Change one thing at a time. Print the same file with a spool that has worked before: if the failure goes away, the problem is the filament (moisture, temperature, wrong profile); if it stays, it is the printer or the slicer. Write down every change: that log is how you get to your real failure rate.