Why Prusa is floundering, and how you can avoid their fate December 26, 2023 on Drew DeVault's blog

Prusa is a 3D printer manufacturer which has a long history of being admired by the 3D printing community for high quality, open source printers. They have been struggling as of late, and came under criticism for making the firmware of their Mk4 printer non-free.¹

Armin Ronacher uses Prusa as a case-study in why open source companies fail, and uses this example to underline his argument that open source needs to adapt for commercial needs, namely by adding commercial exclusivity clauses to its licenses – Armin is one of the principal proponents of the non-free Functional Source License. Armin cites his experience with a Chinese manufactured 3D printer as evidence that intellectual property is at the heart of Prusa’s decline, and goes on to discuss how this dynamic applies to his own work in developing a non-free license for use with Sentry. I find this work pretty interesting – FSL is a novel entry into the non-free license compendium, and it’s certainly a better way to do software than proprietary models, assuming that it’s not characterized as free or open source. But, allow me to use the same case study to draw different conclusions.

It is clear on the face of it that Prusa’s move to a non-free firmware is unrelated to their struggles with the Chinese competition – their firmware was GPL’d, and the cited competitor (Bambu) evidently respects copyleft, and there’s no evidence that Bambu’s printers incorporate derivatives of Prusa’s firmware in a manner which violates the GPL. Making the license non-free is immaterial to the market dynamics between Prusa and Bambu, so the real explanation must lie elsewhere.

If you had asked me 10 years ago what I expected Prusa’s largest risk would be, I would have simply answered “China” and you would have probably said the same. The Chinese economy and industrial base can outcompete Western manufacturing in almost every manufacturing market.² This was always the obvious vulnerability in their business model, and they absolutely needed to be prepared for this situation, or their death was all but certain. Prusa made one of the classic errors in open source business models: they made their product, made it open source, sold it, and assumed that they were done working on their business model.

It was inevitable that someday Chinese manufacturers would undercut Prusa on manufacturing costs. Prusa responded to this certainty by not diversifying their business model whatsoever. There has only ever been one Prusa product: their latest 3D printer model. The Mk4 costs $1,200. You can buy the previous generation (at $1,000), or the MINI (from 2019, $500). You can open your wallet and get their high-end printers, which are neat but fail to address the one thing that most users at this price-point really want, which is more build volume. Or, you can buy an Ender 3 off Amazon right now for $180 and you’ll get better than half of the value of an Mk4 at an 85% discount. You could also buy Creality’s flagship model for a cool $800 and get a product which beats the Mk4 in every respect. China has joined the market, bringing with them all of the competitive advantages their industrial base can bring to bear, and Prusa’s naive strategy is causing their position to fall like a rock.

Someone new to 3D printing will pick up an Ender and will probably be happy with it for 1-2 years. When they upgrade, will they upgrade to a Prusa or an Ender 5? Three to five years a customer spends in someone else’s customer pipeline is an incredibly expensive opportunity cost Prusa is missing out on. This opportunity cost is the kind of arithmetic that would make loss leaders like a cheap, low-end, low-or-negative-margin Prusa printer make financial sense. Hell, Prusa should have made a separate product line of white-labeled Chinese entry-level 3D printers just to get people on the Prusa brand.

Prusa left many stones unturned. Bambu’s cloud slicer is a massive lost opportunity for Prusa. On-demand cloud printing services are another lost opportunity. Prusa could have built a marketplace for models & parts and skimmed a margin off of the top, but they waited until 2022 to launch Printables – waiting until the 11th hour when everyone was fed up with Thingiverse. Imagine a Prusa where it works out of the box, you can fire up a slicer in your browser which auto-connects to your printer and prints models from a Prusa-operated model repository, paying $10 for a premium model, $1 off the top goes to Prusa, with the same saved payment details which ensure that a fresh spool of Prusa filament arrives at your front door when it auto-detects that your printer is almost out. The print you want is too big for your build volume? Click here to have it cloud printed – do you want priority shipping for that? Your hot-end is reaching the end of its life – as one of our valued business customers on our premium support contract we would be happy to send you a temporary replacement printer while yours is shipped in for service.

Prusa’s early foothold in the market was strong, and they were wise to execute the way they did early on. But they absolutely had to diversify their lines of business. Prusa left gaping holes in the market and utterly failed to capitalize on any of them. Prusa could have been synonymous with 3D printing if they had invested in the brand (though they probably needed a better name). I should be able to walk into a Best Buy and pick up an entry-level Prusa for $250-$500, or into a Home Depot and pick up a workshop model for $1000-$2000. I should be able to bring it home, unbox it, scan a QR code to register it with PrusaConnect, and have a Benchy printing in less than 10 minutes.

Chinese manufacturers did all of this and more, and they’re winning. They aren’t just cheaper – they offer an outright better product. These are not cheap knock-offs: if you want the best 3D printer today it’s going to be a Chinese one, regardless of how much you want to spend, but, as it happens, you’re going to spend less.

Note that none of this is material to the license of the product, be it free or non-free. It’s about building a brand, developing a customer relationship, and identifying and exploiting market opportunities. Hackers and enthusiasts who found companies like Prusa tend to imagine that the product is everything, but it’s not. Maybe 10% of the work is developing the 3D printer itself – don’t abandon the other 90% of your business. Especially when you make that 10% open: someone else is going to repurpose it, do the other 90%, and eat your lunch. FOSS is great precisely because it makes that 10% into community property and shares the cost of innovation, but you’d be a fool to act as if that was all there was to it. You need to deal with sales and marketing, chase down promising leads, identify and respond to risks, look for and exploit new market opportunities, and much more to be successful.

This is a classic failure mode of open source businesses, and it’s Prusa’s fault. They had an excellent foothold early in the market, leveraging open source and open hardware to great results and working hand-in-hand with enthusiasts early on to develop the essential technology of 3D printing. Then, they figured they were done developing their business model, and completely dropped the ball as a result. Open source is not an “if you build it, the money will come” situation, and to think otherwise is a grave mistake. Businesses need to identify their risks and then mitigate them, and if they don’t do that due diligence, then it’s their fault when it fails – it’s not a problem with FOSS.

Free and open source software is an incredibly powerful tool, including as a commercial opportunity. FOSS really has changed the world! But building a business is still hard, and in addition to its fantastic advantages, the FOSS model poses important and challenging constraints that you need to understand and work with. You have to be creative, and you must do a risk/reward assessment to understand how it applies to your business and how you can utilize it for commercial success. Do the legwork and you can utilize FOSS for a competitive advantage, but skip this step and you will probably fail within a decade.