Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / The Silent Killer of Hardware Startups
Transcript
- Lucas: Most hardware founders think their biggest financial risk is a component shortage or a tooling error. But if you look at the balance sheets of the startups that actually survive past year two, the real killer is almost always something far more boring: inventory carrying cost. Luna: Right, because everyone calculates the cost of the part itself, but nobody really models what it costs to keep that part sitting on a shelf for ninety days. Lucas: Exactly. And by September of twenty twenty six, with interest rates having settled into this sticky middle ground, the cost of holding inventory has become a silent margin destroyer. I want to talk about a specific framework we use when advising new teams, starting with a company called Kinetic Robotics, which we met with last month. Luna: I remember them. They were making those compact robotic arms for lab automation, right? They seemed close to a breakthrough but kept delaying production. Lucas: They were. And the delay wasn't technical. It was financial. They had secured a manufacturing run for ten thousand units, paid for it upfront, and then sat on it while waiting for a major distributor contract to sign. They thought they were being prudent by having stock ready. In reality, they were burning roughly eight percent of their total valuation every single quarter just to store those arms. Luna: Eight percent per quarter? That sounds like they were paying rent on their own money. How did they even calculate that number? Lucas: It starts with the basic rule of thumb that most engineers ignore: inventory carrying cost is typically between twenty five and thirty percent of the inventory value per year. So if you have fifty thousand dollars worth of components in a warehouse, you are losing twelve to fifteen thousand dollars annually not because the parts broke, but because they exist. Luna: That includes warehousing, insurance, and depreciation, obviously. But does it include the opportunity cost of the capital tied up? Lucas: Yes, and that is the piece that trips people up. If you took that fifty thousand dollars and invested it in a low-risk treasury bill or used it to hire an engineer, you would earn a return. By locking it in steel and plastic, you pay an implicit tax. At current rates, that opportunity cost alone adds another four percent to your carrying burden. Luna: So when Kinetic Robotics looked at their books, they weren't just seeing a line item for warehouse space. They were seeing a massive hidden expense eating their runway. Lucas: Precisely. Their CFO realized that for every unit they held beyond the immediate need, they were effectively setting fire to three percent of its value every month. The decision to wait for the big contract was logically sound from a sales perspective, but financially catastrophic. Luna: It’s interesting how that contrasts with software startups where inventory is zero. Hardware teams have this physical weight that drags down their agility. Lucas: It’s the fundamental difference between variable costs and fixed commitments. In software, if a feature doesn't sell, you delete code. In hardware, if a batch doesn't sell, you have a mountain of metal taking up space and requiring security. The emotional attachment to having 'stock ready' is strong, but it's an illusion of safety. Luna: But isn't having stock ready actually safer when supply chains are as volatile as they've been lately? You can't build what you don't have. Lucas: That is the classic dilemma. And it brings us to the second part of the calculation: the risk premium. If you hold inventory, you protect against supply shocks. But you also expose yourself to obsolescence and damage. The key is finding the sweet spot where the cost of stockout equals the cost of holding stock. Luna: So you're suggesting there's a mathematical equilibrium point rather than a binary choice between too much or too little. Lucas: There is. And calculating it requires looking at your order cycle time versus your demand variability. If you can produce in two weeks and demand is steady, you should carry almost no buffer. If lead times are six months, you need a strategic reserve. The problem for most startups is they assume long lead times for everything, so they overstock on low-risk items. Luna: That makes sense. Most components are commodity, but they treat them like critical scarcity items. What about the initial prototype phase? Does carrying cost matter there? Lucas: Not really. During prototyping, volume is so low that storage costs are negligible. The danger zone starts when you move to pilot runs of hundreds or thousands. That is when the compounding effect kicks in. I saw a consumer electronics startup last year that held two thousand smart home hubs in a garage because they couldn't find a warehouse deal. They lost forty percent of their margin to theft and humidity damage alone. Luna: Humidity damage? That seems like a quality control issue rather than a financial one. Lucas: It is both. When inventory degrades, you write it off. A write-off is a direct hit to gross profit. It also messes with your investor reporting because suddenly your asset base shrinks without corresponding revenue. It looks like poor management of resources, even if it was just bad weatherproofing. Luna: So the lesson is to minimize the time between payment and sale. But how do you do that when you need to pay manufacturers upfront? Lucas: You negotiate payment terms that align with cash flow, not production schedules. Instead of paying thirty percent deposit and seventy percent on delivery, try to structure deals around milestones or accept longer net sixty terms in exchange for a slight discount. It frees up working capital immediately. Luna: Or you use factoring, though that comes with its own fees. It feels like every option has a tradeoff. Lucas: Always. But the alternative is letting inventory sit idle. Let me pause here for a moment to mention something important about how we keep this show running. If these conversations about operational efficiency and financial modeling are useful for what you're building or running, we rely on listener support to keep the content ad-free and deeply researched. You can support the show directly at buy me a coffee dot com slash fexingo. It helps us dig deeper into these unglamorous but critical topics. Thanks for considering it. Luna: We appreciate the support. It allows us to spend time on cases like Kinetic Robotics instead of just skimming the surface. Lucas: Back to the math. Once you have your carrying cost percentage, you need to apply it to your specific SKU mix. Not all inventory carries the same risk. Fast-moving items might justify higher stock levels because the turnover rate offsets the holding cost. Slow-movers are pure drag. Luna: So you segment your inventory. ABC analysis, maybe? Classifying items by value and velocity. Lucas: ABC analysis is the standard starting point. A-items are high value, high velocity. You manage those tightly. C-items are low value, low velocity. You might just order them in bulk to save on administrative effort. The mistake is treating all C-items the same. Some C-items are obsolete tech that needs to be cleared out ASAP. Luna: Clearing them out means taking a loss, but it stops the bleeding of carrying costs. Is there a rule of thumb for how fast you should turn over inventory? Lucas: For hardware startups, I aim for an inventory turnover ratio of four to six times per year. That means your average inventory sits for two to three months. Anything longer than three months starts to feel expensive, especially in the current macro environment. Luna: Three months seems tight. Especially if you're dealing with custom molded parts that have long lead times. Lucas: It is tight. But remember, you don't need to hold the finished goods for three months. You can hold the raw materials or sub-assemblies closer to the production line. The further you push inventory downstream, the cheaper it is to store. A bare PCB is cheap to store. A fully assembled, boxed, and labeled device is expensive. Luna: So the goal is to keep things as generic and unassembled as possible until the last possible moment. Delayed differentiation. Lucas: Exactly. Delayed differentiation is your best friend here. By keeping products in a semi-finished state, you reduce the risk of obsolescence and lower the storage footprint. If a design tweak is needed, you haven't wasted labor on assembly yet. Luna: That changes the entire manufacturing strategy. It shifts the focus from efficient mass production to flexible final assembly. Lucas: It does. And for startups, flexibility is more valuable than raw efficiency. You don't need to run a factory at full capacity; you need to be able to pivot quickly without being anchored by tons of unsold finished goods. Luna: Do investors care about this metric? Or do they just look at top-line revenue? Lucas: Sophisticated investors definitely care. They look at days sales of inventory. If your DSI is rising while revenue is flat, they know you're accumulating dead weight. It signals poor operational discipline. It’s a red flag that suggests you’re selling less efficiently than you think. Luna: So it’s not just an internal accounting exercise. It’s a signal to the market about your health. Lucas: Absolutely. And correcting it can unlock significant value. When a startup reduces its inventory footprint, it frees up cash that can be used for R&D or marketing. It improves the return on invested capital. It makes the business look leaner and more agile. Luna: It turns a liability into an asset. Which is ironic, since inventory is technically an asset on the balance sheet. Lucas: Until it becomes a burden. The trick is to view inventory not as wealth, but as deferred expense. It’s money waiting to be spent, trapped in metal and plastic. The sooner you convert it back to cash, the better your position. Luna: That’s a powerful way to reframe it. Deferred expense rather than stored value. Lucas: It changes how you make decisions. Instead of asking 'do we have enough stock?', you ask 'how much stock are we willing to pay to hold?' It puts a price tag on every unit sitting in the warehouse. Luna: And hopefully, that price tag forces a harder conversation about whether that unit will actually sell before it depreciates. Lucas: It should. Because in hardware, depreciation is not just an accounting entry. It’s physical reality. Batteries degrade, plastics yellow, connectors corrode. Time is literally costing you product quality. Luna: So speed to market isn't just about brand recognition. It's about preserving the integrity of the asset itself. Lucas: Exactly. Every day a product sits on a shelf, it loses a fraction of its potential value. Minimizing that time is one of the most effective ways to protect your margins. Luna: Do you think this mindset will shift as ai driven demand forecasting becomes more accessible to smaller players? Lucas: It already is shifting. Tools that predict demand spikes are becoming cheaper and easier to integrate. But technology only helps if you have the operational discipline to act on the data. Having a forecast is useless if you still commit to large batches based on gut feeling. Luna: So the tech is enabler, but the culture change is the driver. You need founders who trust the numbers over their intuition. Lucas: That’s the hard part. Intuition built the company; data saves it. Bridging that gap is where the real growth happens. Luna: A challenging transition, but necessary for sustainability. Lucas: Necessary for survival. As the market matures, the winners won't be the ones with the coolest prototypes. They'll be the ones with the cleanest balance sheets and the fastest inventory turns. Luna: Which leaves us with a simple question for any founder listening: how many months of inventory are you currently paying to store, and could you survive if you cut it in half tomorrow? Lucas: A good litmus test. If the answer is no, you have work to do before your next product launch.