I must start by acknowledging of course that this is only an idea to hopefully provoke discussion and hopefully is not read as presumptuously demanding how MNT should be, ahem, reformed.
Anyway, I reckon the launch of the future models - such as the “Next” is likely to result in increased product demand that cannot be met by the current setup of hand-building the laptops in Berlin. Traditionally, this might lead to one of two outcomes:
Outsource manufacture. This would remove the charm of having your machine hand-built and, if it is outsourced to the far east, will increase the distance the product has to travel to reach the customers (in Europe), and make it harder to ensure oversight of both quality, and importantly for MNT, the ethics of the contractor.
Raise prices to constrain demand. This may turn the Reform into an executive toy rather than an accessible machine.
So how about a federation of small MNT assemblers in the key markets? These could be existing computer businesses that identify with MNT values or new organisations with novel ownership like co-operatives. Every key city could have an MNT federation member.
There would need to be some sort of governance to call the shots, either from “on high” directed by Berlin HQ, or some sort of democracy where the federation members have some say. One amongst many things that would have to be resolved.
But I think there would be many benefits. For MNT, their HQ would be freed up to do more research and development as well as dealing only with their local demand. I imagine the federation of constructors would help accelerate Reform development because they are closer to their local markets and might be in a position to channel successful modifications back upstream to the main manufacture.
When I was thinking about a Framework I suggested that they might consider recruiting “mom and pop” computer shops as dealers and service centres. However this was counter to Frameworks plans for a web-store that would supply spares and resell used components. But perhaps with Framework taking the open hardware philosophy to a whole other level it lends itself better to full service “hubs” where the machines are assembled and can deal with modifications and aftermarket support in their own markets and language.
I’d be interested to hear what others think of this idea and how it might work in practice. Assuming the Reform is open hardware already, I guess there is little to stop somebody building and selling their own, except that to do so at low volumes would be uneconomic. Perhaps the federation would be more about co-ordinating the purchase of components like chassis and circuitboards to drive prices down and paying for the rights to use the Reform trademarks in exchange for the benefits that brings?
I like this idea generally, but it does seem like it might create unnecessary complications which could drive up both overhead and risk, eating up quite a lot of the team’s time and energy and increasing costs. Especially when we consider that there will be bad and/or unreliable actors entering the mix.* I’d been pondering such a thing myself a while, though, so I felt I should chime in. It presents an attractive image of decentralized manufacture, but the very act of standing up and debugging such a system would be itself non-trivial IMO.
But, presuming two things:
Demand for the Next does outstrip the ability of MNT to build and ship these machines quickly enough to avoid significant customer or reputational loss. (This market seems very forgiving with regard to timelines so far, TBH. Open source hardware fans seem acclimated to long shipping delays.)
Margins are big enough to produce enough surplus income to justify hiring outside help if such demand does materialize. (I honestly don’t imagine they’re making enough right now to pay even themselves very much, but this is based only on my impression of how most bootstrapped early-stage companies do and on the lack of the usual telltale hiring spree on MNT’s part.)
Then it might make sense to contract with an organization which specializes in that sort of thing, one which has already done the work to source and contract with reputable workshops, and which can take on the risk of those workshops failing to meet their obligations. It might make sense for reputable, established open hardware developers to pool their resources and spin such a thing up, if they have the appetite and savings necessary to do so. I do not think this risk and overhead is a thing MNT should take on internally, at least not at this early stage.
–
*E.g., imagine what the impact might be if a workshop in another country receives parts and then does nothing with them, or claims the package was lost, or resells them for a quick buck. Legal enforcement of whatever contracts were signed could be very, very difficult, depending on the jurisdictions involved, and at any rate MNT would be losing a fair amount of money. This is a risk with any contract, but that risk multiplies with every new contract MNT has to enter into to make this vision of decentralized manufacturing happen.
TLDR: federated manufacture might sound trendy but was the norm in the olden days before “intellectual Property”, cheap global logistics and realtime communication, and it worked. And for the sake of the planet we should go back to these tried and true ways.
I think the easiest way to think about it is the traditional business model of Head Office → Distributor → Dealership. Back in the olden days (before air travel, when communication was by letter or maybe telegram), the local agents of a company had considerable leeway, and I concede there was more need for ways to enforce trust. Nowadays with command and control more stuff can be done centrally so you have - in the IT scene - firms like Apple that run huge globe spanning operations managed centrally from its iconic “doughnut” headquarters.
With open hardware we can not only go back towards the olden days of autonomous distributors, but make it more democratic still. And of course MNT is a computer business and can apply modern IT in service of a decentralised structure, rather than as applied as in “modern” capitalism for centralising command and control.
I remember somebody telling me they worked in a bike dealership several decades ago and the bikes would come in a kit of parts. They even claimed that this included lacing the spokes, hubs and rims into wheels! So MNT - whose machines are intended as living and growing tools rather than finished consumer products - might get the distributor or maybe even the dealership to do final assembly.
I reject this as being a showstopper, particularly since we are dealing with open hardware which removes the stakes involved where you are dealing with so-called intellectual property. In the olden days companies had to trust their overseas agents (I’m British and our Victorian age was famous for its gentleman’s code of conduct that enforced the trust needed). Perhaps Open Hardware scene has its own sufficiently robust codes of conduct?
In the case of MNT the local distributors can start with kits of parts brought on pro-forma basis, and as trust increases batches of individual components can be drop shipped from their place of manufacture. And if it is found a distributor does stiff MNT in the big way you suggest, it will happen only the once!
As I understand it, a significant part of MNT’s labor costs come from quality control. E.g., DIY kits initially took them more time and effort to ship because they would first assemble the parts, test the unit, and then disassemble it again. For this to be an improvement for MNT, rather than an additional drain on time and capital, it seems they would need to sell untested kitted out components for the distributors to then test and assemble and sell. The distributors would have to buy the components up front and take on the risk of not being able to shift their inventory themselves; this would also mean a lot more wasted hardware, which would definitely not be very environment (or profit) friendly in comparison to a more “command and control” JIT manufacturing model. (This waste is why JIT manufacturing is the dominant form today.)
However, switching away from a business model where they sell directly to consumers to one where they sell to distributors might indeed be an improvement for MNT, provided they can source competent and reliable distributors who can uphold the brand’s standards. The danger, of course, is that the losses incurred by the increased wastefulness of this system will be borne by those distributors, which could make the whole enterprise a bit more fragile as the already razor-thin margins would be even thinner for those tasked with interfacing with the end users. Witness how this business model worked out for Zulily, for example.
Edit: The advantage of having all that warehoused material all around the world (provided distributors can be sourced and kept happy), though, is that it would protect the supply chain from chip shortages somewhat. There’s definitely a tradeoff between the older way of manufacturing and the modern way in terms of resilience versus efficiency. My assumption is that MNT is a small enough team, and that the market is small enough, that margins are already extremely thin, which makes me think a strategy favoring efficiency would be better for them.
I’m not sure why it is any more “wasteful” for faulty components to be identified at the dealership level than at a centralised assembly shop at Berlin. A dud component is a dud component, and it could be agreed to ship it back to HQ for refund and post-mortem.
I worked in the bicycle scene and there are small scale hobbyist distributors for boutique brands, we are talking of home business scale. I can think of a couple that would have been considered on any objective level totally incompetent and unreliable, but somehow things muddled though. The model of manufacturer/brand ->distributor → dealership is reliable and robust.
It’s not the faulty components I’m referring to with regard to increased wastefulness. It’s the up-front purchasing of components based on the estimates of multiple distributors. I presume there will be a not insignificant number of distributors who will purchase more than they can sell, resulting in those devices sitting unused. A JIT manufacturing model, though more vulnerable to supply chain shocks, does allow for a tight fit between production and demand, and thus less waste.
Right, but you did have to muddle through to reach that point. There was waste and there were distributors causing extra headaches. My point isn’t that this is a totally unworkable system that no industry should ever use. My point is that it does come with additional overhead, and my suspicion is that this additional overhead isn’t worth the improved supply chain resilience at this point in time.