3D printing on demand: €6.75 a unit, and ten sales a day before you pay yourself

The €18 sale that leaves you €6.75

Sell a printed object on a marketplace and over a third of the price is gone before you count your own time. Filament, platform fees, packaging, electricity, a failed-print allowance — that is €6.25 of an €18 sale (roughly 35%). Twenty minutes of human labour at €15/h takes another €5.00.

What is left is €6.75. That number is the entire business. Everything else in this report is about whether you can repeat it 296 times a month, and whether the €18 will still be €18 next year.

It probably will not be. The dossier’s sharpest data point: articulated dragons that sold for £14 in 2023 sell for £4.50 now. A 68% collapse in two years, in the exact niche people enter first.

Where each €18 goes

Line	Cost	Share of price
Filament (60g, wholesale-adjusted)	€1.50	8%
Marketplace fees (~11%)	€2.00	11%
Packaging and shipping consumables	€2.50	14%
Electricity	€0.15	1%
Failed-print allowance (3–8% real rate)	€0.10	1%
Your labour, 20 min @ €15/h	€5.00	28%
Left over	€6.75	37%

Two lines here are the ones amateurs delete from their spreadsheets: labour and failure. Together they are €5.10 — three quarters of the margin. Delete them and you get the internet’s favourite claim, an 80% gross margin business. Keep them and you get a job that pays €20/hour with capital risk attached.

The barrier to entry — a €300 printer — has collapsed exactly as fast as the price anyone will pay for what it makes.

B2B is the same machine with a different customer

The alternative arithmetic is genuinely better, though the numbers only reconcile if B2B labour is priced higher than the consumer-side €15/h used above — the report does not explain why, so treat the hourly rate as a separate, unstated assumption. A 150g functional part sold direct to an engineering client at €40–50 costs €6–7.50 in engineering filament, €0.30 in electricity, €0.30 in failures and €15 of labour at 45 minutes (implying roughly €20/h, not the €15/h used elsewhere). Net €18–27, with the low end closer to €18 than the €20 sometimes quoted.

	Consumer / marketplace	B2B / prototyping
Price per unit	€18	€40–50
Net margin per unit	€6.75	€18–27 (estimated)
Units for €2,000/month	~296	75–110
Units per day	~10	3–4
Platform fee	~11%	0%
Hard constraint	Being found	Being trusted

Three to four jobs a day is mechanically easy. Finding the people who need them is not. That is a sales job wearing a 3D-printing costume, and nothing about owning printers helps you do it.

The Etsy rule change that deleted a whole business model

In June 2025 Etsy removed the clause permitting sellers to print and sell licensed third-party design files. Items made with “computerised tools” must now be based on the seller’s own original design.

That killed the easiest entry route in the industry: buy a commercial licence from a Patreon designer, print, list, sell. Thousands of shops were built on exactly that. Enforcement is retroactive.

It is not cosmetic. Etsy removed 22% more listings and suspended 1.5x more sellers in 2024 than the year before. One long-standing seller, paying for legitimate merchant licences, described being caught in a sweep aimed at illegal sellers.

The lesson is structural, not about Etsy. If your compliance depends on one platform’s clause, your business can end in a policy update you did not get to read first.

Licences are easy; the fixed costs are not

No special licence or qualification is required to run a print service in most jurisdictions. This is unregulated light manufacturing. That is precisely why 20,000+ firms in Italy alone are estimated to already own the capability in-house — an unverified trade figure, but directionally the problem: many of your prospects can already do the job themselves.

The real constraints are three:

• IP exposure. Printing licensed characters, sports logos or trademarked geometry draws takedowns and, occasionally, litigation. This is enforced, not theoretical.
• Business registration and social contributions. In most European systems a sole trader owes a minimum social-security contribution regardless of revenue. In Italy that floor is roughly €4,460 in 2025 and €4,521 in 2026, reducible to around €2,900 under the flat-rate regime. Whatever the country, budget for a fixed annual sum owed at zero revenue.
• Premises rules. Opening a public-facing workshop generally requires a filing with the local authority; working purely as a subcontractor for other businesses often does not. Confirm locally.

None of this stops you. All of it means slow ramp-ups bleed money before the first invoice.

Almost everyone who did this at scale is dead

The precedent list is brutal, and the survivors survived for one identifiable reason.

Company	Model	Outcome
Shapeways (2007–2024)	Consumer marketplace + industrial	Chapter 7, July 2024, on $34.46M FY2023 revenue. Revived as smaller Manuevo BV, still serving US/UK/EU customers; original marketplace not restored
Fast Radius (2017–2022)	Industrial bureau	Closed. $24M debt vs $6.2M cash; assets sold for $15.9M
Sculpteo / BASF Forward AM	Online bureau	Insolvency under 5 months after BASF carve-out, despite 30% growth
3D Hubs (2013–2018)	P2P hobbyist network	Original model killed; entity sold to Protolabs for $280M in 2021
Printrbot (2011–2018)	Hardware	Closed. Founder cited low sales
Weerg Srl (Italy)	Industrial B2B	€13.6M revenue, €1.75M profit FY2024, 50–99 staff
Xometry	Marketplace for buyers	$545M FY2024 revenue, 34.5% marketplace gross margin, $1.0M adjusted EBITDA

Shapeways had real revenue and still went bankrupt; commentary blames SPAC pressure for growth over sustainable unit economics. Fast Radius died the same way — a capital structure sized for $635M of 2025 revenue that print-job margins could never fund. Neither failed for lack of customers.

3D Hubs is the one that should worry the reader most. Its customers bought their own cheap printers. The network evaporated, and the company survived only by abandoning the hobbyists who built it — to considerable uproar among the small operators who had grown on it.

Weerg sells tolerance-controlled parts with technical sign-off. Xometry sells procurement access to a network of at least 3,429 suppliers (FY2023 figure; the company has since reported further double-digit supplier growth in FY2024, so the current count is higher). Neither sells a print. Both survive.

What kills you

Ask one question before spending anything: why doesn’t this client just buy a printer? If the honest answer is “they haven’t got round to it” or “it’s cheaper for now”, you are 3D Hubs’ original network.

The specific failure modes:

• Price decay outruns cost decay. Filament goes from €18–30 retail to €5–11 wholesale per kg, and your selling price falls faster than that saves you.
• Volume you cannot hold. Ten consumer sales a day, every day, with no slow weeks and no acquisition cost, on a platform where 5,000+ listings counts as saturated.
• The dropout curve. An industry blog — stated, not audited — puts break-even failure at roughly 65% of shops, with 80% gone by year two. Directionally consistent with everything else here, but a single unaudited source and should be read as an estimate, not a fact.
• Platform dependence. Already changed once in 18 months.

The money is not in owning the machine. It is in being the person a small manufacturer calls when a jig has to fit tomorrow.