Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / Hardware Startups And The Physics Of Supply Chain Risk
Transcript
- Lucas: Let’s start with a number that keeps hardware founders up at night: ninety percent of new device companies fail within their first three years, and more than half of those failures aren’t product-market fit issues—they’re supply chain collapses. We’ve talked before about reverse logistics and design for manufacturing, but today I want to drill into something even more visceral: the physics of supply chain risk. Not just the financial side, but the physical reality that your prototype on the bench is made of atoms, and those atoms are moving through a global system that is currently fracturing. Luna: Right, because you can have the most elegant CAD wireframe in the world, but if your primary capacitor supplier is in a region hit by a port strike or a natural disaster, you’re sitting on a pile of unshippable plastic. It feels less like engineering and more like geopolitical chess. Lucas: Exactly. And the difference between a resilient startup and a fragile one often comes down to one decision made six months before launch: dual sourcing. Most founders treat suppliers like vendors—transactional, interchangeable. But in today’s environment, treating them as partners with shared risk is the only way to survive. Let’s look at a concrete example. Remember when NVIDIA saw demand for their AI chips explode? They didn’t just throw money at TSMC; they restructured their entire supply chain to include secondary foundries and long-term capacity agreements. That’s not just buying chips; that’s insuring production. Luna: But for a seed-stage robotics company, you don’t have the leverage to demand priority slots from major foundries. How do you actually negotiate that kind of security without burning cash? Lucas: You don’t negotiate for priority; you negotiate for visibility. The biggest mistake I see is founders hiding their bill of materials behind NDAs until it’s too late. If you share your projected volume growth with your key component suppliers early—even if it’s just rough estimates—you give them the data to plan their own capacity. In return, they might offer you allocation guarantees that standard contracts don’t provide. It’s about trading transparency for security. Luna: So it’s basically acting like a mini-enterprise resource planning system for your own supply chain before you even have the software to track it properly. Lucas: Precisely. And this ties directly into inventory buffering. There’s a dangerous myth in hardware circles that carrying inventory is a sign of inefficiency. In reality, holding three to six months of critical component stock isn’t waste; it’s insurance. Yes, it ties up working capital, but the cost of a delayed launch due to a missing microcontroller is usually ten times higher than the carrying cost of those parts. You have to be willing to pay the premium for stability. Luna: That’s hard to sell to investors who are obsessed with burn rate. They see warehouse costs and think you’re hoarding. How do you frame that narrative? Lucas: You frame it as risk mitigation. When you present to investors, don’t just show them unit economics; show them supply chain maps. Highlight where your single points of failure are, and explain exactly how much buffer inventory you’re holding to neutralize those risks. Investors are smarter now than they were five years ago; they know that supply chain shocks are inevitable, and they reward teams that have already accounted for them in their financial models. Luna: It’s interesting because so many founders come from software backgrounds where scaling is instantaneous. Here, scaling is physical and slow. Lucas: That’s the core disconnect. In software, you add servers; in hardware, you add tooling, certifications, and raw materials. And certifications take time. You can’t just switch a safety-certified power supply because your original supplier ran out of stock. You have to recertify, which takes months. So, dual sourcing isn’t just about having two suppliers; it’s about having two suppliers whose components are cross-compatible or easily substitutable without triggering a redesign cycle. Luna: Which means your design team needs to specify generic parts wherever possible, rather than custom ones tied to a single manufacturer. Lucas: Absolutely. If you specify a proprietary connector or a custom-molded housing, you’ve painted yourself into a corner. Standardize on industry norms. Use off-the-shelf connectors, standard screw sizes, common battery chemistries. Every time you choose a standard part over a custom one, you increase your optionality. Optionality is the antidote to fragility. Luna: But doesn’t standardization kill differentiation? If everyone uses the same off-the-shelf components, how does your product stand out? Lucas: Differentiation shouldn’t come from your internal components; it should come from your integration and user experience. Think about Apple. They use standard memory chips and processors, but their value is in how they design the enclosure, the software, and the ecosystem. Your unique selling proposition is the solution you provide, not the silicon inside. Focus your R&D budget on what the customer touches, not what they never see. Luna: That’s a tough pill for engineers to swallow, though. Engineers love solving hard problems with custom solutions. Lucas: It is. But the market rewards reliability over novelty in the foundational layers. A founder who builds a robust, standardized platform can iterate faster because they aren’t waiting for custom tooling to open. Speed is a competitive advantage, and standardization accelerates it. Luna: Let’s pivot to the geographic aspect. With tensions rising globally, relying on a single country for manufacturing is becoming a liability. Lucas: It’s no longer a liability; it’s an existential threat. We’re seeing a shift toward nearshoring and friend-shoring. Companies are moving assembly from distant overseas factories to regions closer to their end markets or allied nations. For a hardware startup, this might mean assembling in Mexico instead of China, or Eastern Europe instead of Southeast Asia. It reduces lead times and lowers political risk, even if labor costs are slightly higher. Luna: Does that impact your quality control processes? Moving to a new region often means a new workforce with different skill sets. Lucas: It absolutely does. Quality control must be embedded in the process, not inspected at the end. You need to send your own quality engineers to the factory floor during the ramp-up phase. Don’t rely on the contract manufacturer’s QA team alone. Build checklists, automated testing rigs, and statistical process controls that leave no room for ambiguity. Quality is a habit, not an inspection. Luna: So it’s about transferring knowledge, not just products. Teaching the factory how to make your product correctly. Lucas: Exactly. Documentation is your intellectual property. Your work instructions, your test protocols, your acceptance criteria—they must be so detailed that any trained operator can follow them. If your product quality depends on one senior technician knowing a trick in their head, you have a single point of failure. Codify that knowledge. Luna: I’m curious about the role of digital twins here. Earlier episodes covered how startups use them for design, but how do they help with supply chain risk specifically? Lucas: Digital twins allow you to simulate supply chain disruptions before they happen. You can model scenarios: What if my primary supplier delays shipment by four weeks? What if a raw material price spikes by twenty percent? By running these simulations, you can identify bottlenecks and test mitigation strategies virtually. It turns reactive crisis management into proactive planning. Luna: That sounds expensive to set up. Is it realistic for a small team? Lucas: Not necessarily complex simulations. Even basic spreadsheet models that link lead times to inventory levels can serve as a primitive digital twin. The goal is visibility. If you can see how a delay in Component A affects the final assembly date, you can make informed decisions about air freight versus sea freight, or whether to hold extra stock. Luna: It’s about building intuition through data, even if that data is simple. Lucas: Precisely. Now, let’s pause for a moment. If these conversations are useful for what you're building or running, and you want to support independent journalism that stays ad-free, consider buying us a coffee. You can find us at buy me a coffee dot com slash fexingo. It helps keep these deep dives going. Luna: Yeah, it really does. We appreciate every bit of support that lets us dig into these operational details with you. Lucas: Back to the topic. Another layer of risk is regulatory compliance. As you scale, you face different standards in different markets. CE marks in Europe, FCC in the US, CCC in China. Each has its own testing requirements and certification bodies. Luna: Which can create parallel supply chains if you’re producing separate batches for different regions. That doubles your complexity. Lucas: It does. The smart move is to design for global compliance from day one. Use components that are already certified for multiple markets. Design enclosures that meet the strictest environmental standards upfront. It costs more initially, but it saves you from retooling and recertifying later. Compliance is a design constraint, not an afterthought. Luna: So, treat regulations as part of the spec sheet, not as a hurdle to clear at the end. Lucas: Exactly. And finally, let’s talk about relationships. Your supply chain is a network of people. Treat them with respect. Pay invoices on time. Communicate changes early. A supplier who likes you will go to bat for you when things go wrong. A supplier who sees you as a transactional client will prioritize bigger, more reliable customers. In a crisis, goodwill is currency. Luna: That’s such an underrated point. Human dynamics drive industrial outcomes. Lucas: Always. So, to wrap this up: resilience in hardware starts with dual sourcing, moves through strategic inventory buffering, relies on standardization for flexibility, and is cemented by strong human relationships. It’s not glamorous, but it’s the foundation that allows innovation to actually reach the market. Luna: And it reminds us that building a physical product is a marathon of logistics, not just a sprint of invention. Lucas: Right. The question for founders today isn’t just 'Can we build it?' but 'Can we deliver it consistently, despite the chaos?' That’s the real test of maturity. Luna: Something worth keeping in mind as you navigate your next prototype phase. Lucas: Indeed. Thanks for listening.