Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / The Hardware Startup Supply Chain Blindspot
Transcript
- Lucas: Look at the schematic for almost any consumer electronics device released in the last six months. You are looking at hundreds of discrete components—capacitors, resistors, microcontrollers—that have finite lifespans. For a software company, a dependency update is a click. For a hardware startup, it can be a six-month redesign cycle. Luna: Right, because you cannot just patch the silicon when the fab changes its process node without re-qualifying the entire board. It is not like pushing code to production. Lucas: Exactly. And most founders I talk to treat their bill of materials as a static list rather than a dynamic portfolio of risks. They focus heavily on getting the first thousand units out the door, which we covered in previous episodes on pilot programs and design for manufacturability. But the real danger zone starts after year two, when those early suppliers begin to sunset parts. Luna: So we are talking about the quiet crisis that kills companies who actually succeed initially? Because if you sell your product, you suddenly realize you cannot make another one? Lucas: Precisely. Let’s look at a concrete example from the 3D printing sector. Formlabs had a massive run with their resin printers. About three years into their scale-up, they received an end of life notification for a specific power management IC used in their mainboard. It was a standard part, widely available, nothing exotic. But the manufacturer announced they would stop production within twelve months. Luna: That sounds manageable if you have cash reserves. Why did it become a crisis? Lucas: Because the alternative wasn't just buying a different chip. The new equivalent had slightly different pin configurations and thermal dissipation profiles. Their engineering team had to redo the PCB layout, re-spin the boards, and re-run all the regulatory compliance tests for FCC and CE certification. That is not a week's work. That is roughly four months of delayed revenue and increased burn rate. Luna: Four months of zero new shipments while you fix a problem that should have been flagged before mass production. Did they just absorb that cost? Lucas: They absorbed it, but they changed their procurement protocol immediately. Before that incident, their engineers selected parts based purely on performance and unit cost. Afterward, they implemented a mandatory 'obsolescence check' at the design freeze stage. Every component had to be cross-referenced against distributor databases like Digi-Key and Mouser to see if the part was marked as 'not recommended for new designs' or already in end of life status. Luna: That is surprisingly manual work for a company that size. Most teams automate what they can, but supplier data is notoriously messy. Lucas: It is messy because manufacturers don't always send these notices proactively. Sometimes you find out because your primary distributor runs out of stock and tells you there is no foreseeable replenishment. Or worse, you find out because a secondary market dealer is selling the same part at five times the price, signaling that the original factory has shifted capacity to newer products. Luna: So the signal isn't the letter from the manufacturer. The signal is the price spike in the grey market. Lucas: Spot on. And that leads us to the concept of second-sourcing. A mature hardware business never relies on a single source for any critical component. If you use a specific sensor from Company A, you need a qualified alternative from Company B that fits the same form factor and electrical specs. But qualifying a second source takes time and money. Luna: Which brings us back to cash flow. Startups are usually short on cash exactly when they need to spend it on redundant inventory or qualification testing. Lucas: Correct. This is where the 'last-time buy' strategy comes in. When you get that notice that a part is ending, you calculate exactly how many more units you expect to sell over the next eighteen months. Then you place a massive order for enough components to cover that forecast plus a safety margin. You tie up capital in inventory, yes, but you buy yourself runway to transition to the new part without stopping production. Luna: Buying inventory to save production seems counterintuitive when everyone is obsessed with lean manufacturing and just-in-time delivery. Lucas: It feels counterintuitive until you are staring at an empty warehouse and a backlog of orders. Lean works great for stable products. It breaks down when the underlying inputs are volatile. For hardware startups, resilience often looks like inefficiency on paper. You hold extra stock of critical chips. You keep older tooling around longer than necessary. You pay a premium for faster shipping because waiting two weeks for a container means missing a retail window. Luna: I remember reading about a drone startup that ran into this. They were using a custom motor driver that only one vendor made. When that vendor went bankrupt overnight, the startup was stuck. Lucas: Yes, the Flyability Elios team faced similar pressures early on. They operated in confined spaces where standard off-the-shelf parts didn't fit. Customization creates lock-in. If you build a product around a proprietary interface or a unique mechanical coupling, you have removed the option to swap vendors easily. That is a strategic choice, but it must be acknowledged as a risk multiplier. Luna: So customization saves weight and space but destroys flexibility. It is a trade-off every founder has to make consciously. Lucas: Absolutely. And here is the thing most technical founders miss: they think the problem ends once the new part is qualified. It doesn't. The new part might have a different failure mode. Maybe it gets hotter under load. Maybe it is more sensitive to voltage spikes. You have to test it rigorously in the field, not just in the lab. Warranty claims can skyrocket during this transition period if you aren't careful. Luna: Which circles back to the warranty topics we discussed a few episodes ago. So obsolescence management is actually a customer service issue too. Lucas: It is entirely a customer service issue. If you switch components and the device fails at eighteen months instead of thirty-six, your reputation takes a hit. Customers don't care that you switched suppliers due to a factory closure. They just know their expensive gadget broke sooner than advertised. Luna: This feels like a discipline that separates the hobbyists from the serious businesses. Is there a tool that helps with this, or is it mostly gut feeling and spreadsheets? Lucas: There are enterprise PLM systems like Arena or PTC Windchill that integrate supply chain data. But for a startup, those are overkill and too expensive. Most successful teams build their own lightweight dashboards. They pull data from distributor APIs daily. They set alerts for price changes and stock levels. It is unglamorous backend work, but it is the immune system of the company. Luna: Unsung heroes of the hardware world. The people who watch the spreadsheet so the engineers can dream up the next invention. Lucas: Exactly. And I want to pause here briefly because this kind of operational diligence requires resources. If these conversations help you build a more resilient business, consider supporting the show directly. It keeps our research deep and independent. You can find us at buy me a coffee dot com slash fexingo — that's F. E. X. I. N. G. O. Luna: We appreciate any support that helps us keep digging into the details that matter for builders. Lucas: Let's go back to the drone example. Once that startup survived the supplier bankruptcy, they adopted a policy called 'standardization by exclusion.' They looked at every non-critical part in their drone and asked: can I replace this with something generic? If yes, they forced the change. Luna: Reducing the variety of parts reduces the complexity of the supply chain. Fewer SKUs to manage, fewer potential points of failure. Lucas: Right. They realized that their custom casing was fine because it defined their brand identity. But the screws, the cables, the basic connectors? Those should be commodity items sourced from whoever has them in stock locally. They cut their bill of materials complexity by forty percent in one quarter. Luna: That is a huge leverage point. Most founders hoard custom parts because they think it makes the product feel special. But internally, it is a logistical nightmare. Lucas: It feels special to the user, maybe. But it costs millions in hidden overhead. Standardizing on common interfaces also makes repair easier. If a customer can buy the replacement cable from any electronics store, the brand wins loyalty. If they need a proprietary cable only available from the manufacturer, the brand loses trust. Luna: So supply chain strategy is actually brand strategy in disguise. The way you build determines how you are perceived when things go wrong. Lucas: I couldn't agree more. We tend to silo operations away from marketing, but in hardware, they are inseparable. Your ability to deliver on time, to maintain quality across component swaps, to offer long-term support—all of that is dictated by how you managed your BOM years earlier. Luna: It makes me wonder if we are teaching engineering students the right things. Are we teaching them to optimize for performance, or for longevity and replaceability? Lucas: Probably performance. Curriculums are slow to adapt to market realities. But the industry is starting to push back. Investors are asking tougher questions about supply chain risk during due diligence. They don't just want to see a cool prototype anymore. They want to see a map of where every screw and chip comes from, and who else could provide them if needed. Luna: That is a shift from romanticizing the garage workshop to professionalizing the back office. It sounds less sexy, but it is what actually builds lasting companies. Lucas: It is the difference between a product launch and a sustainable business. One gets you on the news. The other keeps you open for the next decade. Luna: As more hardware startups face these exact headwinds, do you think we will see consolidation in the component supplier space too? Lucas: Likely. As demand for specialized chips grows, the smaller fabricators might get bought out by the giants. That reduces the number of alternatives available to startups. It makes that second-sourcing strategy even more critical, because your options are shrinking globally. Luna: So the advice stands: diversify early, standardize aggressively, and never assume a part will be there forever. Lucas: Never assume anything about the physical world. Everything decays, changes, or disappears. Your job is to plan for the transition before it hits you.