Eight years ago I built a garden.
Year one was tomatoes and basil, and it was not impressive. Each season after added something — berries, orchard fruit, row crops — and with them an understanding of how much food a small piece of land can truly produce, which is considerably more than I would have guessed and more than most people believe.
Before long we had more than we could eat. So I started dropping care packages at friends’ doors, and the message that came back was the same every time: the best they’d ever had.
One evening in mid-August, harvest bins at my feet, the question landed. If this helped me personally, financially, and emotionally — why couldn’t it do the same for more people?
That question is the company. Everything since has been an attempt to answer it without being naive about why nobody has.
The honest reason people don’t grow food
They try. That’s the part that gets missed in every piece of writing about this.
An enormous number of households start a garden, and a large share of them quit inside two seasons — not because they didn’t care and not because they’re lazy. They quit because the first year of growing food is a brutal information problem dressed up as a leisure activity.
You don’t know your frost dates. You don’t know that the thing you planted needs six hours of sun and the spot you chose gets four. You don’t know that you missed the window for the second planting because nothing told you the window existed. You get one feedback cycle per year, the consequences arrive months after the decision, and by the time you understand what went wrong the season is over and you’ve concluded that you’re bad at this.
You’re not bad at it. You were handed a system with a twelve-month latency and no instrumentation and told to figure it out.
I have spent my career on exactly that class of problem — slow feedback, invisible failure, consequences arriving long after the decision that caused them — and the answer is never to try harder. It’s to build something that closes the loop.
Modern barn-building
Here’s the frame I settled on, and I’ll defend it.
Communities used to raise the things a household couldn’t build alone. Not out of sentiment — out of arithmetic. Some structures take more hands than one family has, so the neighbors showed up, the barn went up, and the household could feed itself for a generation. The barn was infrastructure and the raising was a delivery mechanism.
Food production at household scale is that kind of structure again. It takes a bit more capability than most people currently have, most of that capability is knowable, and almost none of it is being transferred.
So: design, build, grow, give. Build the physical system properly — beds engineered to last rather than to look good for one season. Then stay for the part where people quit, which is the first year, and hand over enough capability that they don’t need you for the second one.
The part that makes it a worse business
I want to name this because it’s the load-bearing decision and it costs money.
The decision filter is four lines. Strengthen community. Build something essential. Transfer skill, not dependency. Scale locally before globally.
The third one is expensive. A company that succeeds by making its customers competent is structurally worse at recurring revenue than one that succeeds by making them reliant. That is not a clever positioning statement, it’s a real constraint on the business model, and I have looked hard at whether I actually mean it.
I do. Partly because it’s the thing I want to be true of anything I build, and partly for a less noble reason: dependency-based businesses in this category are already served, and the household that wants to be capable is not. There is a market argument and a values argument and they happen to point the same way, which is the only kind of principle I fully trust.
Why a platform, and why not more of one
The software exists because the information problem is the real problem. A bed designer that produces a real materials list and a real quote, so nobody’s first interaction is a vague conversation about what this might cost. A coach that tells you it’s time — in your zone, for your beds, in plain language, before you’ve missed the window rather than after.
What it isn’t is a social network for gardeners, a marketplace, or a subscription that gets more valuable the longer you fail to learn anything. Those are all better businesses. They’re also all versions of the thing I said I wouldn’t build.
Where it actually stands
In development. Pre-launch. Not open for business.
The brand is done. The designer works end to end. The plant knowledge is real, and thinner in places than I’d like. The operations side is staged rather than running. The compliance work is what’s gating the door, and I would rather be late than sloppy about it.
There’s no traction section here because there’s no traction yet. A stats band reading zero installs advertises that you’re new; one with invented numbers is worse. The only proof I have is qualitative and it’s the founder’s proof: eight seasons, more food than we could eat, and a stack of care packages that came back as the best they’d ever had.
That’s enough to build on. It isn’t enough to claim anything, and I’m not going to.
■