Why failed prints belong in the price

What a failure actually costs

A failed print does not just waste the plastic already laid down. It consumes four things at once, and only one of them is material:

• Filament. Usually partial, occasionally the whole model plus supports and brim.
• Machine hours. The printer was occupied for the entire duration of the failure, including the hour before anyone noticed it had gone wrong.
• Your time. Noticing, scrapping, cleaning the plate, re-slicing, restarting.
• Queue position. The machine that was going to print a paid job next is now printing a free one.

Material is the visible loss and usually the smallest. The hours are the expensive part, because hours do not come back.

The arithmetic: you must produce more than you sell

A failure rate is a multiplier on everything you do. If 8% of the prints you start are scrapped, then to deliver 100 good parts you must start about 109.

Consider a part that costs $3.84 to make when it succeeds. At an 8% failure rate, every good part you ship effectively carries $3.84 ÷ 0.92 = $4.17 of cost — an extra $0.33 per part, or 8.7% on top of everything.

At a 20% failure rate, common while tuning a difficult geometry or a new material, the same logic gives $3.84 ÷ 0.80 = $4.80. That is $0.96 per good part, a quarter more than the successful-print cost. Setting the failure rate to zero does not remove the cost. It just moves it out of the spreadsheet and into your margin.

Measuring your own failure rate

There is no useful industry figure, and anyone quoting one is selling something. Yours is measurable in about five minutes a month.

1. Count the prints you started in the last 30 days. Slicer history or a print-management tool will show this.
2. Count the ones you scrapped as unusable — not the ones with a cosmetic flaw you sold anyway.
3. Divide. That is your real rate for the work you actually take on.

Do it separately for each material if you print more than one. Failure rates are material-specific and geometry-specific: a printer that runs PLA at 4% may run a flexible filament at 25%, and averaging the two produces a number that is wrong for both.

Where failures actually come from

Failures cluster. Four causes account for most of them, and each is addressable without new equipment:

Cause	What it looks like	What reduces it
First layer	Part detached or shifted hours in	Level (or probe) the bed, clean it, slow the first layer
Wet filament	Bubbling, poor layer bonding, random blobbing	Dry the spool, store sealed with desiccant
Geometry	Tall thin parts wobbling, overhangs collapsing	Rotate the model, add support, slow down
Wear	Under-extrusion appearing gradually	Check the nozzle and extruder tension on a schedule

Wet filament is the one people skip, because the spool looks fine and the failure looks random. It is neither — it is usually a spool that has been open for months.

How to price for it without hiding it

Two legitimate approaches, and they produce the same price:

Gross up at the end. Compute your successful-part cost, then divide by (1 − failure rate) to get the cost you actually carry. Simple, and easy to explain to a customer who asks why the price is what it is.

Price the failures as a line. Show a failure allowance as its own row in the breakdown — the print cost calculator does this. Slightly more work, but it makes the number visible, and visible costs are the ones people remember to check.

What does not work is quoting a rate you have not measured, then defending it when someone asks how you arrived at it.

When to fix the process instead

Pricing around failures is only correct while the failures are genuinely variable. There comes a rate at which the better answer is to stop printing until the process is fixed.

A useful test: if a single cause accounts for more than half your failures, it is a process problem, not an environmental one. A first-layer issue that recurs every third print is not something to charge customers for — it is a bed preparation problem with a solution that costs nothing but an evening. Once the rate is down, the price can come down, which is a competitive advantage you bought with maintenance rather than with margin.

Important assumptions

The 8% default is the calculator's example figure, not a measured average. The whole argument of this page is that your own number is the only one worth using, and that it must be counted rather than estimated.

The arithmetic assumes failures are independent and that a failed print is scrapped entirely. Partial failures — prints that finish but have a cosmetic flaw you can still sell — do not consume a full material allowance, and it is fair to exclude them from the rate.