Before you commit capital to a glasshouse, it helps to test your assumptions in a structured way. A simulation or planning model, whether a detailed tool or a well-built spreadsheet, lets you change one input at a time and see what happens to the result. It cannot tell you whether a site will work or whether a buyer will sign. It can show you which assumptions the plan depends on most, and that is often the most useful thing it does.
This article is about planning models for controlled environment agriculture (CEA) and heated glasshouses. It is not a review of games. For a comparison between farming games and real growing, see our article on farming simulators and real greenhouses.
Start with the inputs you can defend
A model is only as good as the numbers you feed it. Before you open any tool, write down the inputs you are using and where each one came from. Typical inputs include:
- the floor area, growing system and number of crop cycles per year
- the crops you plan to grow, and the time from sowing to harvest for each
- expected yield per square metre, taken from growers or research rather than from the tool’s default
- heating and energy assumptions, based on the climate in your area and the insulation of the structure
- labour hours for each stage of the crop, and the cost of the people who will do the work
- packaging, transport, certification and other costs that are easy to forget
- the price you expect to achieve, based on conversations with buyers
Mark each input as one of three things: confirmed by a supplier quote, taken from a published source, or your own estimate. Most plans look sound until you count how many of the key numbers are in the third group.
Run a sensitivity check
Once the base case is in place, change one input at a time and record the effect on your outcome. Common outcomes to track are gross margin per square metre, the number of months to recover the capital, and the cash balance at the lowest point of the year.
Pay particular attention to the inputs that move the result most. In many glasshouse plans these are the achieved selling price, the yield, and the energy cost. A small change in one of these can change the answer more than any amount of fine tuning elsewhere.
Then run a worst-case version. What happens if the first crop takes longer than planned, the heating system fails in January, or a buyer pays late? A plan that only works when every assumption goes your way is not a plan. It is a hope.
Test the timing, not just the totals
Many models show annual totals, which hide the real pressure of a growing business. A glasshouse can look profitable across the year while running short of cash in the months when you are buying seed, paying for heating and waiting for the first harvest.
Build a monthly cash flow. Show when money goes out and when it comes in, including payment terms with buyers and any credit you will need from suppliers. The lowest month in that schedule is the one your funding has to cover.
Compare the model with people who grow
A model should be checked against the experience of growers, not used in place of it. Speak to growers in your region about their real labour costs, the yields they achieve on each crop, and the points in the season when they struggle. Ask what they wish they had known in the first two years.
Those conversations often change the model. A labour figure that looked generous may turn out to be too low once you count the hours spent on training, pruning and harvesting, or the time lost when a member of staff leaves in the middle of a season.
What a model cannot tell you
Even a careful model leaves things out. It will not tell you:
- whether the site has planning permission for the structure you want, or how long the process will take
- whether the grid connection or gas supply is adequate for the heating you need
- how well the crop will do in your particular microclimate, including wind exposure and frost pockets
- whether the market for your product will hold up, or how a buyer will treat you in a bad season
- how your team will cope with a disease outbreak or a seasonal labour shortage
These questions need site visits, written quotes, professional advice and early conversations with buyers. A simulation can help you decide which questions to ask first.
Practical steps before you commit
- Write the assumptions down. Keep a single page with each key input and its source.
- Run the base case, then the downside. Record the result of each change.
- Build a monthly cash flow. Find the lowest point and check it against your funding.
- Check two or three inputs with growers. Pay particular attention to yield, labour and price.
- Get written quotes for the big items. Structure, heating and glazing are the usual ones.
- Review the model after your first season. Replace estimates with real figures and rerun it.
The aim is not to produce a perfect forecast. It is to find out early which parts of your plan are fragile, so you can strengthen them before the money goes in.
Key points
- A planning model is a thinking tool. It shows the effect of your assumptions but does not prove that a plan will work.
- Write down the source of each key input and treat your own estimates with care.
- Run a sensitivity check and a worst-case scenario, and focus on the inputs that move the result most.
- Build a monthly cash flow, because the lowest month is often the real test.
- Check the model against growers and written quotes before you commit capital.
- Review the model after the first season, and replace estimates with actual figures.
For practical help with a planning review or a greenhouse project, our consultancy team can work through the assumptions with you, and our article on greenhouse business plans sets out what a complete plan should cover.