Nobody Said Take My Money — Part 1 of 7

On June 18, 2026, I wrote the decision record that shut down Mother Hen, a smart chicken-coop monitoring business I had spent more than a year building.

At the moment I shut it down, the platform was running without a hitch at two beta sites — one in Tennessee, one in Colorado, about 1,500 miles from my desk. The firmware updated itself over the air. The devices provisioned themselves through the cloud with no field pairing. The alerts arrived on time, every time, including the ones that mattered at 3 a.m. In the entire life of the product, the technology never gave me a reason to quit.

Number of paying customers, lifetime: zero.

Twelve days before the shutdown decision, I stood at a festival booth for two days in front of a steady stream of exactly the people this product was built for — chicken keepers who had lost birds, who told me predator stories unprompted, who nodded at every feature. I sold nothing. Not one unit. And in two days of conversations, not one person raised the price as the reason.

This series is the post-mortem. I’m writing it because the version of this story that usually gets published is the survivorship version — the pivot that worked, the near-death experience before the up-round. The more common outcome, the one where a competent person builds a working product for a market that turns out not to exist, mostly goes undocumented. That silence costs other founders real years. It cost me one.

The thesis, up front

This is not a story about engineering failure. I want to get that off the table in the first post, not out of pride — though there’s some of that — but because it’s the load-bearing fact of the whole series. Everything worked. The hardware worked, the cloud worked, the alerts worked, the economics on paper worked.

The failure was that I never tested whether anyone would pay, until after I had built everything. Every mistake in this series is a variation on that one sentence. The engineering competence didn’t just fail to save the business — it actively made the failure more expensive, because every technical problem I solved felt like progress, and none of it was.

What Mother Hen was

The pitch: you keep chickens. Predators — raccoons, foxes, owls — kill chickens, mostly when a coop door that should be closed isn’t. Mother Hen watched the coop for you. A cellular-connected controller at the coop tracked the door. A small battery-powered sensor tracked temperature, humidity, and door state. An egg-shaped status light sat in your kitchen and glowed one color when everything was secure and another when it wasn’t. If something was wrong — door open after dark, dangerous cold, sensor gone silent — you got a text message. 24 hours a day, no quiet hours, because a raccoon doesn’t observe quiet hours.

Cellular, not WiFi, because coops sit at the far end of the property where WiFi doesn’t reach. No app to install; a web portal for the details, text messages for the things that couldn’t wait.

The economics: a complete kit cost me $311.52 to build — $209.02 for the controller, $71.97 for the egg light, $30.53 for the door/climate sensor. The plan was to sell the hardware essentially at cost and make the business on a $29.99/month subscription. Breakeven was 48 subscribers. Not 48,000. Forty-eight.

I never got one.

What I built before asking the question

An inventory, because the size of the build is the size of the lesson:

Roughly 24,000 lines of C firmware across four codebases, running on ESP32 microcontrollers under ESP-IDF and FreeRTOS — one unified codebase that compiled to four different device binaries with about 70% shared code. Thirty-four AWS Lambda functions across three backend services: a REST API, an event-driven telemetry and notification layer, and a risk engine that classified every event into four severity tiers and decided who got woken up and when. Cloud-based fleet provisioning. Over-the-air firmware updates with rollback protection. A customer web portal. A versioned alarm specification. Two patent filings — 19/439,421 and 63/953,780.

The monthly AWS bill for all of it: between one and seven dollars. People assume hardware startups die of infrastructure costs. Mine could have run for a decade on what I spent on the booth fee for the festival where I sold nothing.

I list all of this for one reason: every item on it was built before I had a single piece of evidence that a stranger would pay for the product. Not weak evidence — none. The build list was long and the evidence list was empty, and I did not notice that imbalance for more than a year, because the build list kept growing and growing feels like progress.

Where the story goes

The next six posts, in order:

Part 2 — The n=1 trap. Mother Hen was built for one friend with one real problem. The root cause of everything: I treated “the problem exists” as proof that “the problem generalizes,” and never tested the difference. What that test costs when you run it on purpose: about ten conversations.

Part 3 — The year of building. How engineering competence becomes a hiding place. The roadmap was full of build gates and had not one demand gate. If you’re a builder, this is the post that will read most like a mirror.

Part 4 — Three tests, one answer. A paid ad campaign, a festival booth, and a community deep-read — three independent demand tests, one converging verdict: enthusiasm without purchase. This is also the post where the title of this series earns its name.

Part 5 — The pivot month. June 2026, week by week: a new strategy document nearly every day, each one individually defensible, all of them collectively a symptom. What thrashing looks like from inside while it’s happening.

Part 6 — Killing it well. How the shutdown decision actually got made — written as a formal decision record, on evidence instead of exhaustion — and the one final demand signal that arrived only after I pulled the plug.

Part 7 — The playbook. The whole failure, reordered into the sequence I’d run next time. Everything this series describes was learnable for under $500 and a month of conversations. The point of the series is that you spend the $500.

One more thing about the title. Nobody Said Take My Money is not a joke and not quite a complaint. It’s the shortest accurate summary of what happened, and it’s the sentence the shutdown decision ultimately turned on. By Part 4, you’ll know exactly where it came from.

Part 2: The n=1 Trap — My Friend Was the Hypothesis, Not the Proof.