Journal · 2026.09 · 12 min

Let Sales Lead Production — Choosing a Greenhouse Type from the Market Back

The traditional model — production-led selling — grows the crop first and looks for buyers afterwards. Open-field farming can sometimes absorb that; protected agriculture cannot: the investment is several times higher, fresh produce cannot be held in inventory, and farm roads across Taiwan often will not admit a truck or carry its axle load. For export there is more: the destination country’s quarantine-greenhouse standard has to be settled at the planning stage. Decide whom to sell to before deciding what to build — skip that step and what you will be worrying about later is where the revenue comes from.

“What kind of greenhouse can I build on this land?” My own view starts from the other end: “Who are you going to sell it to?” Without an answer there is no basis for choosing a type — because a greenhouse’s specification is never decided by the parcel of land. It is decided by the market.

“Production-led selling” is the logic of traditional manufacturing: make the product first, put it into the warehouse, then let the sales team and the distribution channels move it. Industry can afford this because industrial goods keep — what does not sell can be discounted, re-specified, or simply left to wait.

Farm produce has none of those options. Quality begins counting down at the moment of harvest; cold storage slows that clock but never stops it; the harvest window is set by a biological clock, not by the order book. Open-field growing carries low capital exposure, so a season lost is a season survived. Protected agriculture costs several times — sometimes tens of times — more, so carrying the same “grow first, sell later” logic into a greenhouse means staking far heavier capital on the same uncertain outcome, with even the fallback of holding inventory taken away.

The way I read a production area comes from how I was trained. I came up through the dust of my family’s greenhouse sites, then from 2017 to 2024 worked in industrial construction — manufacturing plants, technology-sector facilities and logistics warehouses. The habit those years build is this: on a plant site layout, the first thing you draw is not the building but the vehicle movements — how raw material comes in, how finished goods go out, which route a truck of a given gross weight takes, whether the turning radius clears, what design load the pavement and the floor slab have to carry. In industrial construction these are baseline questions; a drawing that has not settled them does not get approved. Coming back to greenhouses in 2024 and reading Taiwan’s production areas with the same eyes, I found the weakest link was never the greenhouse structure. It was the physical side of selling — road access, vehicles, collection and packing, cold chain — which almost nobody treats as a design problem.

Even selling only domestically, produce still has to leave the farm before it becomes money. The road systems serving most of Taiwan’s agricultural production areas are immature — on many sites the first thing we look at isn’t the land. It’s the road:

• Large vehicles cannot get in: the access road is too narrow for refrigerated trucks and material deliveries, so everything has to be transloaded onto light vehicles — trips and product losses both multiply.
• Trucks cannot make the turns: junction angles and turning radii on farm roads are too tight, so a truck jams in the last hundred or two hundred metres and, at peak dispatch, the whole site waits on one vehicle.
• Road load ratings do not match: farm roads were never designed for the axle loads of crop and material haulage, so heavy traffic brings pavement settlement and break-up — and the cost and the downtime of repairs usually land back on the operator.

None of this is a “sort it out later” item. Logistics feed straight back into where the greenhouse sits, how the entrances and vehicle circulation are laid out, and how large the collection and packing yard has to be — and sometimes into whether the parcel suits your target market at all. Test the road at the same time as you choose the site and the greenhouse type.

Domestic channels — wholesale markets, supermarket and hypermarket supply contracts, foodservice and institutional catering, home-delivery e-commerce — each demand different grades, volumes, delivery windows and packaging, and each works back into a different cropping schedule and level of environmental control. Fix the channel first and the specification has a basis.

The moment the target is export, the whole question moves up a level: variety choice and maximum residue limits (MRLs) follow the destination country’s rules, packaging and cold chain follow the voyage — and the item most often overlooked is the greenhouse itself. Many destination countries accept fresh produce only from a registered, approved quarantine greenhouse or insect-proof screen house — in phytosanitary terms, an approved place of production — with written standards for insect-screen mesh size, double-door entry vestibules, floor treatment and management records.

If the greenhouse is not designed and built to the destination country’s quarantine-greenhouse standard from the outset, what has to be made good later is not paperwork but construction: the wrong mesh size means re-screening whole elevations, a missing entry vestibule means rebuilding the doors and the circulation around them, non-compliant floors and drainage mean demolition and rework — each several times the cost of building it right the first time. Whatever cannot be brought into compliance pushes you into fumigation, consignment-by-consignment inspection and other expensive clearance procedures, or the consignment is refused entry and returned. All of it is avoidable cost that one round of drawings could have removed.

• Fix the market: domestic or export, which channel, who the main buyer is — ideally with a letter of intent or contract-farming terms in hand before the investment.
• Fix the specification: work the buyer’s quality grades, supply volumes and delivery windows back into variety, cropping schedule and pack-out targets.
• Fix the cultivation: from the cropping schedule and quality targets, work back to the growing method and the level of environmental control — this step, not the land, decides what class of greenhouse you need.
• Fix the site: test the parcel against logistics and quarantine — can the trucks get in? Will the road carry the load? If you are exporting, can the quarantine specification actually be built here?
• Fix the greenhouse type: only now are form, specification and construction cost a question you can actually calculate.

If you are weighing a new greenhouse investment, bring your target market, channel terms and site conditions together — we will start from “who will buy it” and work backwards with you. If the numbers hold up, then we talk drawings.

← More greenhouse insights