Nobody Said Take My Money — Part 4 of 7
By mid-2026, without ever planning it as a research program, I had run three independent demand tests on Mother Hen. Different methods, different audiences, different costs. Paid advertising against a cold audience. Two days of face-to-face selling to a warm, self-selected crowd of exact target customers. And a slow, careful read of the communities where those customers talk to each other when no one is selling to them.
Three tests. One answer. This is the post about what that answer sounded like — because the maddening thing about weak demand is that it doesn’t sound like “no.” It sounds like “wow.”
Test one: the paid campaign
Early in the venture’s commercial life, I hired an agency and ran a paid pre-launch campaign — professionally built, aimed at the actual target audience, with real money behind it. This is the most expensive and most honest kind of message test there is: strangers, cold traffic, a real ask.
It didn’t convert. Not “converted poorly” in a way that invited optimization — across the variations tried, the message never found a version of itself that made cold strangers reach for a wallet.
I’m deliberately not naming the agency or litigating the campaign’s mechanics; the execution isn’t the point, and the fault I care about here is mine. The point is what I did with the signal. A paid campaign that won’t convert against your own target audience is about as loud as early evidence gets. It should have been a halt trigger: stop building, figure out whether demand exists at all, resume only with proof in hand.
Instead I filed it under “marketing is hard” and kept building for roughly six more months. The money was spent, the lesson was in hand, and I declined delivery.
The runnable fix, which I’ll return to in Part 7: decide the kill threshold before the campaign runs, in writing. “If this misses X, development halts that day.” The number has to be committed cold, before you’re invested in the outcome — because afterward, sunk cost will quietly move any goalpost that wasn’t nailed down.
Test two: the festival booth
In early June 2026 I worked a booth for two days at a large homesteading festival in Tennessee — thousands of attendees, and not a general crowd but the crowd: people who keep chickens, lose chickens, and drive hours to spend a weekend on exactly this way of life.
Day one: heavy foot traffic, warm conversations all day. People were excited about the product — genuinely excited, unprompted-predator-stories excited. They nodded at the door monitoring, they lit up at the egg-shaped status light on the table, they told me about the raccoon that got in two springs ago. Zero sales.
Day one also handed me a moment I think about a lot. A retailer stopped by and asked the good, boring, professional questions — who handles fulfillment, what’s the margin per unit. And standing there, doing arithmetic in my head in real time, I realized I was looking at something like a hundred dollars a unit, best case, before any of the costs I hadn’t thought to count. He wasn’t hostile. He was doing basic diligence, and I didn’t have answers, and both of us could tell.
Day two: same again, this time with more industry people — coop providers and influencers, receptive across the board. Zero sales. Two full festival days in front of the most precisely targeted audience I will ever stand in front of, and I did not sell one unit.
Now the part that makes this a post about epistemics and not just a sad booth story. That night, I diagnosed the failure as pricing. I wrote it down: hardware needs roughly a 4:1 price-to-cost ratio to leave margin for a channel, mine was priced around 1.5:1, therefore the bundle was economically unviable — a pricing problem, not a demand problem. It’s a competent-sounding analysis. It has ratios in it.
It took another week and a half of flailing (Part 5) to notice the fact that demolishes it: in two full days, not one person raised the price as their objection. Nobody asked for a discount. Nobody said “too rich for me.” They said lovely things and kept walking. A pricing problem announces itself — people tell you, constantly, that you’re too expensive. What I had was a desire problem wearing a pricing costume, and I dressed it that way myself, because a pricing problem is fixable with a spreadsheet and a desire problem might not be fixable at all. Watch yourself do this. The mind protects the product.
What the booth should have been instead: a commitment test. A clipboard and a card-on-file pre-order — real card, charged the day the unit ships, cancel anytime by text. Not painless (a real card on a real form), but costless today. Ten cards out of two days is a business. Two cards is your answer, and either way you’d know — for the cost of a clipboard. I walked away with goodwill instead of data. Goodwill doesn’t compound. Data does.
Test three: the community deep-read
The third test cost nothing and, in hindsight, was available all along. I went where chicken keepers talk to each other with no vendor in the room — the forums and subreddits where they trade predator stories — and I read carefully, not for enthusiasm but for behavior.
The fear is real. That was never the issue; the threads are full of loss, and the grief in them is not mild. But the behavior underneath the fear told the actual story: these people already solve the problem, with a walk. A nightly headcount at the coop door. Free, reliable, already habitual, bundled with a moment of checking on animals they love — many of them would take the walk anyway.
My true competitor was never another monitoring product. It was a habit that costs nothing and mostly works. Against that competitor, a monitoring subscription isn’t buying safety — it’s buying permission to skip a chore that a lot of the audience doesn’t experience as a chore. The reading also surfaced something darker for the product thesis, which Part 5 picks up: the losses that actually devastate people mostly don’t happen the way my product assumed.
One answer
Line the three tests up.
Paid cold traffic: enthusiasm in the metrics that don’t cost money, nothing in the one that does. A warm room of exact buyers: two days of delight, zero transactions. The community in its natural habitat: real fear, already self-soothed for free.
Three different instruments, one reading: enthusiasm without purchase. People liked Mother Hen. People agreed the problem was real. People said kind, encouraging, specific things, at every stage, in every channel, for over a year.
And in the entire life of the product — this is the sentence the shutdown decision was eventually built on, and the reason this series is named what it’s named — nobody ever said “take my money.”
Not once. No one asked to pre-order. No one asked when they could buy it, then followed up. No one tried to press a deposit on me at the booth. I have since come to believe that this sentence is the single most useful demand instrument a founder owns, because it requires no budget and cannot be gamed: if you have genuine demand, somebody, somewhere, gets impatient and tries to give you money before you’re ready to take it. It happens to every product people actually want. If it has never happened to yours, that silence is a reading too.
Enthusiasm is what politeness looks like at scale. Purchase is the only sentence in the customer’s language that can’t be said politely. Learn to hear the difference a year earlier than I did.
Part 5: The Pivot Month — What Thrashing Looks Like From Inside.