A reasonable question, and one we’d rather answer directly than dodge: if a spool of filament costs around $30 and a dash vent uses a small fraction of it, why doesn’t the part cost five dollars?
The honest answer is that material is close to the smallest line item in low-volume reproduction. Here’s where the money actually goes.
The one-time cost: getting to a working model
Before a single sellable part exists, someone has to produce a validated model. For a moderately complex interior piece that’s typically 15–40 hours of work:
- Sourcing a reference part — often the hardest step. A good, undamaged original of a discontinued part can be genuinely difficult to find, and we sometimes buy several to average out their variation.
- Measurement and modelling — the bulk of it. Reconstructing intended geometry from a warped, cracked thirty-year-old sample is slow, careful work.
- Test printing and iteration — usually three to six revisions before fitment is right, each one a print cycle plus a fitting session.
- On-vehicle validation — confirming the part fits a real car, not just the reference sample.
That cost is fixed regardless of whether we sell five units or five hundred. Which leads directly to the central economic problem of this business.
The volume problem
Injection moulding is astonishingly cheap per unit — often well under a dollar for a small trim piece. But the tooling costs thousands to tens of thousands of dollars and takes weeks to produce.
That model works beautifully at 50,000 units. It’s completely unviable at 200, which is a realistic lifetime demand for a specific interior part on a specific 1990s model. Nobody is cutting a steel mould for that, which is precisely why the part was discontinued in the first place.
Additive manufacturing has effectively no tooling cost, which is what makes 200 units possible at all. The trade-off is that per-unit cost stays roughly flat instead of collapsing with volume — there’s no economy of scale to grow into.
| Injection moulding | Additive | |
|---|---|---|
| Tooling cost | $5,000–$50,000+ | Essentially none |
| Per-unit cost | Very low | Moderate, and roughly constant |
| Viable at 200 units | No | Yes |
| Design change cost | New tooling | Edit the file |
Machine time is the real per-unit driver
A dash vent is typically 4–10 hours of print time. Larger or more detailed pieces run longer.
During that window the machine produces exactly one thing. It can’t be doing anything else. Printer capacity is the genuine bottleneck in this kind of operation — not filament, not labour, not shipping. Every part competes for hours on a finite number of machines, and those machines have purchase cost, maintenance, nozzle and belt wear, and power draw amortised across every hour they run.
Failed prints are a real cost, not an edge case
Prints fail. A warped corner, a layer shift, a partial clog at hour six of an eight-hour run. In a well-tuned process the failure rate is low but never zero, and every failure consumes the full machine time and material up to the point it failed.
Then there’s quality rejection — parts that completed successfully but didn’t meet spec on inspection. Those get scrapped too.
Both have to be priced into the parts that do ship. That’s not padding; it’s arithmetic.
Post-processing is more labour than people expect
A part coming off the plate is not a finished product. Depending on the piece:
- Support removal, carefully, without marring visible surfaces
- Cleaning up seams and the first-layer edge
- Surface finishing where the part is cosmetic
- Dimensional inspection of the critical features
- Test fitting on a reference assembly for tight-tolerance parts
This is hands-on human time on every single unit, and it’s frequently the largest labour component.
Material: real, but not the headline
Engineering filaments cost meaningfully more than commodity PLA — ASA and similar UV-stable materials run several times the price of the cheapest options. That’s a deliberate choice for the reasons covered in our materials guide: a cheaper filament produces a part that deforms on a hot dashboard, which is worse than no part at all.
Even so, on a typical interior part material is a modest share of the total. Choosing the correct engineering material rather than the cheapest one changes the final price far less than people assume — and changes the part’s lifespan enormously.
Everything else
Packaging that protects a part with fine features in transit. Shipping. Payment processing, which takes a percentage of every order. Replacements under our fitment guarantee. Website, hosting, and the time spent answering fitment questions by email — which is substantial, and which we’d rather do properly than not at all.
Roughly how it breaks down
For a typical interior part at realistic volumes, the shape of the cost looks something like:
- Machine time and amortisation — the largest single share
- Post-processing labour — close behind
- Design cost recovery — spread across expected lifetime units
- Material — smaller than most people guess
- Failure and rejection allowance
- Fulfilment, fees and overhead
The exact proportions shift by part. The ordering rarely does.
Why comparing to a junkyard part isn’t like-for-like
A used original is often cheaper. It’s also the same age, the same chemistry, and the same distance into the degradation curve as the one you’re replacing. You’re buying a part with an unknown and probably short remaining life.
A reproduction in a UV-stable engineering material starts its clock at zero, in a material chosen specifically to resist the failure mode that killed the original. Whether that’s worth the difference depends on what you’re building — we laid out both sides in 3D printed vs. junkyard parts.
What brings the price down
Genuinely, demand does. Design cost amortises over more units, we can batch prints more efficiently, and machine scheduling gets less wasteful. Parts that sell steadily get cheaper to make.
This is a large part of why we encourage custom requests even when we don’t have the part. Several people asking for the same piece is what justifies the design work — and if you’re the one who asks first, you’ve effectively put it on the roadmap.