Starting a 3D printing service business: machines, pricing, channels, the first year

To start a 3D printing service business, pick one process you can run well (usually FDM, sometimes resin), buy several identical, reliable machines rather than a mix, price every job from your own costs with a minimum order, and choose one or two sales channels where buyers already look for parts. The first year is mostly about learning what your jobs really cost and finding the business customers who come back.

Is there a market for a 3D printing service?

Yes, and it is the part of the industry that is growing fastest. The Wohlers Report 2026 put total additive manufacturing revenue for 2025 at about $24.2 billion, up 10.9 %, with printing services making up 48 % of it. Services revenue grew 15.5 % while printer sales grew 3.6 % (Wohlers Report 2026, via 3Dnatives). More money now flows through shops that print for others than through the sale of printers.

That growth is shared with large competitors. Online marketplaces have taught engineers to expect a price seconds after they upload a file: Xometry reported second-quarter 2026 revenue of $229 million, up 41 % on the year (Xometry Q2 2026 results). A new shop does not compete with that on scale. It competes on being close, specialised, quick to answer and easy to talk to.

Which process should you start with?

Start with what your first customers will buy. A shop that serves product designers needs different machines from one that serves dentists or a motorsport team. The table compares the polymer processes most service businesses begin with, in general terms.

If you are unsure, the comparison in FDM vs SLA vs SLS sets out which parts suit which process. Most new shops start with FDM because the machines are cheap to buy and forgiving to run, then add resin when they keep turning away detail work.

How many machines, and which ones?

Several identical machines beat a varied collection. Identical printers share spare parts, settings, a maintenance routine and a single set of tested material profiles. When one fails, the job moves to the next without anyone re-slicing it. A mixed fleet multiplies everything you have to know.

Buy for utilisation, not for peak demand. A printer that runs a third of the time costs nearly as much to own as one that runs every night, and your machine-hour rate depends on the hours you actually sell. Add machines when the queue regularly runs more than a few days long, not before. The article on print farm pricing covers how capacity and rate interact once you run many printers.

How should you price 3D printing jobs?

Price from costs, never from a competitor’s price per gram. A price per gram treats a solid block and a thin-walled enclosure of the same mass as the same job, when one prints in an afternoon and the other needs supports, a slow outer wall and twenty minutes of cleaning.

A sound price has five parts:

1. Material: grams or millilitres used, including supports and purge, at what you pay plus waste.
2. Machine time: hours on the printer at a machine-hour rate that recovers the machine, power, maintenance and your share of rent and software.
3. Labour: preparation, plate removal, support removal, post-processing and packing, at a labour rate.
4. Failure allowance: a percentage for prints that fail and must be run again.
5. Margin: what the business keeps.

On top sits a minimum order value, because every order carries a fixed amount of handling (reading the request, checking the file, packing, invoicing) whatever its size. The 3D printing cost calculation article goes line by line, and the machine-hour rate article explains the rate itself.

Two things stand out. The labour line is larger than the machine line, which is why shops that price by the gram lose money on small, fiddly parts. And the part costs more to handle than to print, which is the case for a minimum order. If you want to try your own numbers, the machine-hour rate calculator and the 3D printing cost calculator do the same arithmetic.

Where will the first orders come from?

Rarely from nowhere. The first orders usually come from people who already know you, and the job of a channel is to make it easy for them, and the people they mention you to, to get a price and order.

Selling manufacturing services online compares these channels in detail, and marketplaces or your own website looks at the trade-off between a marketplace’s volume and owning the customer.

Whatever the channel, answer fast. A buyer who sends the same file to three shops often orders from the first sensible answer. An instant quote on your own site does this without you at the keyboard: Quotometry, for one, prices FDM, resin, SLS and MJF from a shop’s own machines, materials and rates (see the page for print farms).

What does the first year look like?

A realistic first year has four phases, whatever the calendar says.

1. Set up and measure. Get the machines reliable, settle on a short list of materials, and record, for every job, the hours on the printer, the minutes of hands-on work and whether it failed. These records are worth more than any spreadsheet you build in advance.
2. First repeat customers. One or two business customers who order every month change the business more than a hundred one-off orders. Look after them: fast answers, consistent quality, delivery dates you keep.
3. The capacity decision. When the queue is regularly full, decide whether to add machines of the same kind, add a second process, or raise prices. Raising prices is often the right first move.
4. Review prices against reality. Compare what you quoted with what jobs actually took. Correct the machine-hour rate, the labour minutes and the failure allowance, rather than adjusting quotes one at a time.

Through all four, watch how many of your quotes become orders. The article on quote win rate explains what to count and what a low or a very high rate is telling you.

What mistakes do new print services make?

• Pricing by the gram. Simple to explain, wrong for most parts. It underprices labour-heavy work.
• No minimum order. Small orders arrive in large numbers and each one loses money.
• Quoting every job by hand. It does not scale, and an answer that takes a day loses orders to one that takes a minute.
• Buying a mixed fleet. Every new model brings its own profiles, spares and failure modes.
• No terms for customers’ files. State who owns the design, how files are stored and when they are deleted, before a business customer asks.
• Ignoring post-processing time. Sanding, painting and support removal are where hours disappear.

None of these is fatal on its own. Together they explain why many print services stay busy without making money. A business that knows its real cost per job, and prices from it, can grow one machine at a time.

Questions people ask

• Print farm
• Service bureau
• Machine hour rate
• Minimum order value
• Utilisation
• Post-processing