01
Deciding what to produce
This decision does more to determine success than any other, and it should be made from three directions at once.
What the land actually suits, which is soil, water, climate, aspect and scale. Fighting the land is a permanent cost. What you can sell, which means who's within reach, what they'll pay for, and what nobody else is supplying well. And what you can bear to do repeatedly for years, because enthusiasm carries an operation for about eighteen months and the routine carries it after that.
The commercial pattern that works at small scale is high value per unit of area and per hour, with a route to market that doesn't involve competing on price. That points towards perishable, high-quality, or unusual, and away from commodity bulk.
Diversify, but not endlessly. Several enterprises spread risk and smooth cash flow; too many fragment attention and equipment and mean nothing is done well. Two or three that share infrastructure and complement each other seasonally is a common shape that works.
02
Infrastructure, in the order that pays
The commonest expensive mistake is building the wrong thing first, and the right order is fairly consistent.
Water first, always: supply, storage and distribution to where it's needed. Nothing else works without it and retrofitting it around finished layouts costs more than doing it first.
Then access and fencing, because they determine how everything is used and are disruptive to change later. Then the working buildings, storage, shelter, a place to process and pack. Then, usually last, the house, and the number of operations wrecked by building the house first is considerable.
Build for what you'll do in three years, not for what you might do in twenty. Oversized infrastructure ties up capital that the enterprise needs, and under-built infrastructure that can be extended is nearly always the better bet.
And walk the layout before committing. Where things sit relative to each other decides how far you walk every day for the next decade, and that distance is the largest hidden labour cost on most holdings.
03
The numbers that decide it
This is where honesty matters most, and where optimism is most expensive.
Work out cost of production per unit, properly, including your own labour at a real rate. An enterprise that only works because the operator isn't paid isn't viable, it's a subsidy, and knowing that is what lets you decide whether it's a subsidy you're happy to give.
Then cash flow, which is a separate question from profitability. Income arrives with the harvest and costs arrive continuously, and the gap between them is what kills operations that were profitable on paper. Enterprises with different timing, or something that produces income year-round, exist largely to bridge that.
Distinguish capital from operating cost, and be conservative about how quickly capital pays back. Equipment that's idle most of the year is capital sitting still, which is why hiring, sharing and contracting frequently beat owning at small scale.
And build in a bad year. Every operation meets weather, illness, a market collapse or a failure, and one with no reserve turns an ordinary setback into an exit.
04
Labour, and the limit nobody plans for
The binding constraint on most small farms is not land or capital. It's hours, and specifically the hours of one or two people who cannot be replaced.
Map the labour through the year before committing to an enterprise mix. Peaks that collide are the problem: two enterprises that both demand everything in the same six weeks are effectively one enterprise you can do badly. Complementary timing is worth more than higher margin.
Design tasks for repetition. Anything done daily should be short, close and easy, because a five-minute inefficiency in a twice-daily job is many days a year. This is where layout pays back.
Plan for absence. Illness, injury and needing to leave for a week happen, and an operation where nobody else knows how anything works cannot absorb them. Written procedures and someone else who can do the essentials are infrastructure as much as a shed is.
And take the off season seriously. Operations that run flat out all year burn out the people in them, and that's the most common reason good holdings are sold.
05
Getting produce to people
Growing it is half the job and the half that gets all the attention. The other half decides whether the enterprise exists.
Direct sale keeps the margin and costs time and relationship: markets, gate sales, subscription boxes, and direct supply to kitchens. It suits small scale precisely because the volumes are too small to interest a wholesaler and the quality is too high to be paid for by one.
Wholesale is the reverse: less time, far less margin, and volume requirements that a small holding usually can't meet consistently. It works as a route for surplus rather than as the primary channel.
Processing changes the economics. Something preserved, dried, cured or otherwise turned into a product has a longer window, a higher value and different rules attached, and it converts a glut into income and not waste.
⚠ Selling food is regulated everywhere, and the requirements differ sharply between raw produce, processed goods, meat and dairy. Find out what applies before building an enterprise around it.