The Relay Tax
Since enterprises treat hardware as a liability shield, startups can disrupt legacy monopolies by wrapping cheap silicon in certified compliance armor.
By Alan Reed
Sparked by Show HN: I replaced a $120k bowling center system with $1,600 in ESP32s · discussion
I saw a post on Hacker News recently by a guy who replaced his local bowling alley's $120,000 legacy scoring system with $1,600 worth of ESP32 microcontrollers. A non-technical administrator was facing a six-figure bill just to bridge a physical pin mechanism with a digital screen. The legacy manufacturer held a monopoly on the specific interface between the heavy machinery resetting the pins and the monitors displaying the score. A hacker swooped in, bypassed the proprietary system, used cheap silicon triggering mechanical relays, and solved the entire problem over a few weekends.
This dynamic happens every day across elevators, commercial HVAC controllers, and factory floors. The bowling alley anomaly is a perfect distillation of a universal law governing heavy physical industries — and a blueprint for a certain type of billion-dollar company. I call this structural bottleneck the Relay Tax.
Why does this happen? The physical world relies entirely on mechanical switches. Heavy machinery moves through kinetic energy, hydraulics, and high-voltage electrical currents. The modern world, conversely, demands digital interfaces. Bridging these two domains creates a rigid mathematical asymptote of friction. Whenever a business needs to translate a physical action into a digital state, a legacy monopoly stands at the chasm charging an extortionate toll. The hardware bridging the gap is conceptually trivial. A small logic board flips a mechanical relay, taking a low-voltage electrical signal and outputting a high-voltage mechanical action. Yet the vendors controlling these chokepoints price their hardware as if they are selling a bespoke supercomputer.
The implication is that these pricing models are completely divorced from the underlying bill of materials. To understand why enterprise systems grew so bloated, you have to model these B2B environments as isolated evolutionary ecosystems. They are textbook cases of island gigantism. In biological terms, when a species is cut off from a mainland and shielded from the brutal, highly optimized bloodbath of mainland predators, it inevitably grows massive and slow. The B2B legacy vendors experienced the exact same physics. Since they were geographically and operationally isolated from the hyper-competitive consumer tech ecosystem where smartphones ruthlessly garbage-collect any supplier shipping lame hardware, these enterprise vendors grew fat. Their machinery became giant, expensive, and grossly inefficient because no apex predator could cross the water to enforce market discipline.
Why do smart founders fail to disrupt this? Naive hackers look at commercial HVAC controllers or manufacturing assembly lines, see outdated technology, and assume the buyer is simply ignorant. The hacker calculates the computing power of the legacy box, realizes it has less processing bandwidth than a disposable vape, and concludes they can capture the market by throwing bare consumer microcontrollers at the problem. You compile an incredibly elegant, orthogonal software architecture to replace their bogus legacy system, offer it to the enterprise buyer at a ninety percent discount, and assume rational economics will force them to buy it. They say no.
The reality of the enterprise buyer's decision introduces a completely different structural incentive. The administrator at a bowling alley or a commercial shipping dock fundamentally evaluates purchases as liability transfers rather than hardware acquisitions. The ocean protecting these bloated legacy islands relies entirely on regulatory compliance rather than technical superiority. The legacy vendor charges a massive premium because they possess NRTL safety ratings and rigorous governmental certifications. The enterprise buyer is purchasing the strict guarantee that if a mechanical relay fails and the building catches fire, someone else absorbs the lawsuit — a literal firewall against bankruptcy.
When a hacker walks in with an exposed circuit board, they are mathematically offering a catastrophic vector for institutional risk.
So if the buyer actually treats hardware as a liability shield, your hardware startup must entirely rethink its objective function. Bypassing this monopoly requires treating the silicon as a solved problem and focusing entirely on engineering a bureaucratic wrapper. The winning algorithm for this specific market is straightforward: take commoditized, ridiculously cheap consumer technology and enclose it within an enterprise-grade, certified exoskeleton. You take advantage of the massive economies of scale generated by the consumer electronics industry, strip out the consumer-facing interface, and harden the casing. Companies are already doing this by putting Raspberry Pis inside a certified box to bypass the friction of industrial PLC controllers. They wrap five dollars of bare silicon in five hundred dollars of compliance armor, and the enterprise buyer eagerly signs the check.
And this strategy scales predictably across almost all heavy 20th-century industries. If you want to build a massive company in this space, you have to stop acting like a traditional software engineer and start acting like an evolutionary biologist introducing a foreign species to an isolated island. Instead of attempting to engineer a better version of the giant, flightless bird, you simply introduce a rat. The rat in this scenario is the globally commoditized ESP32 microcontroller. The legacy monopolies have spent decades optimizing for a completely different environment, building increasingly Byzantine compliance structures while completely ignoring the collapsing cost of basic compute. They have trapped themselves in a local maximum.
The opportunity lies entirely in regulatory arbitrage. A smart founder identifies a specific, heavily gated mechanical chokepoint. It could be the sensor data output of industrial HVAC systems, the dispatch logic of municipal elevators, or the mechanical scoring relays of independent bowling alleys. You map the exact compliance parameters needed to operate in that physical space. Then, you spend absolutely nothing on developing the core logical processor — the silicon itself is treated as a disposable, zero-cost primitive.
Because you are not burdened with legacy research and development costs, you deploy your startup capital exclusively on the armor needed to smuggle that cheap silicon into the enterprise. You hire the expensive lawyers, you pay for the agonizingly slow Underwriters Laboratories testing, and you over-engineer the physical plastic casing so that it survives being hit with a forklift. You make the outside look like the heavy, expensive enterprise crap the buyer expects, while the inside runs on a clean step function powered by a chip meant for a smart lightbulb.
Stop trying to invent entirely new technology. Find a bloated monopoly, buy a five-dollar microcontroller, wrap it in a certified industrial liability shield, and flip the switch.