How many cycles do I need?

Two comparable cycles show a difference; three or more help you see whether it repeats. If you have only one, use it as the baseline and start measuring the next one carefully.

Should I compare with published yields?

Only with care. Published figures depend on crop, variety, climate, technology and period. Your own best cycle is usually a more honest reference.

What is the most useful single indicator?

Marketable kg per m² together with rejects. Total harvest hides grading losses that often explain the gap between cycles.

Can software do this for me?

Software helps once the records are consistent. The first comparison is usually faster in a spreadsheet built from your documents.

Does this guarantee better results?

No. It helps you choose what to investigate. Results depend on crop, variety, climate, water, market and how changes are carried out.

How to analyze several greenhouse crop cycles with the records you already have

Reviewed: 2026-10-13 · SeedMatchGroup

Short answer

Put at least two or three comparable cycles of the same crop side by side, convert every input to the same unit (per m² and per kg of marketable produce), and look for the line that moves most between your best and worst cycle. That line — not the average — usually points to the bottleneck worth investigating first.

Season reports, a harvest logbook and a tablet with yield charts on a desk inside a tomato greenhouse — Comparing cycles starts with the records already on your desk: harvest, grading, inputs and labour.

Long-running growers often have more evidence than they think: harvest sheets, packhouse grading reports, water-meter readings, fertilizer invoices, energy bills, payroll and maintenance logs. This guide shows how to turn them into one comparison table, what to do when some records are missing, and how to read the result without jumping to conclusions.

What counts as a comparable cycle

A cycle is comparable when the crop, variety type, planting window and growing area are similar enough that a difference in results can be explained by management, equipment or weather rather than by a different product.

Same crop and similar variety segment (for example, cluster tomato vs. cluster tomato).
Similar planting and harvest window, or a clear note of the season.
Same production area, or results converted to per-m².
Known changes recorded: new substrate, new controller, staff change, disease outbreak.

If your cycles are not comparable, compare each one with itself (input per kg sold) rather than with each other.

The data you need — and where to find it

Indicator	Typical source	Unit to use
Marketable yield	Packhouse or sales records	kg/m² per cycle
Grade-out / rejects	Grading reports, returns	% of harvested kg
Water applied	Water meter, irrigation controller logs	L/m² and L/kg marketable
Drain volume (if substrate)	Drain meter or manual measurement	% of applied
Fertilizer	Invoices, stock-book	cost or kg N/K per m²
Energy	Utility bills, fuel delivery notes	kWh or fuel per m² and per kg
Labour	Payroll, time sheets	hours per m² and per 1,000 kg
Failures and downtime	Maintenance log	events and hours per cycle

Keep the original documents; record in the table only what you can trace back to them.

Step-by-step: build the comparison table

List the cycles in columns (for example, Spring 2024, Spring 2025, Spring 2026).
Fill each row from the table above, with the source document noted.
Normalize: divide by area (m²) and by marketable kg.
Mark gaps as "missing" — never fill them with a guess.
Calculate the spread for each row: (best − worst) ÷ average.
Rank the rows by spread and by money impact.
Write one hypothesis for the top two rows, and the evidence that would confirm or reject it.

Greenhouse performance audit template (copy into your spreadsheet)

One row per block, crop and season. Fill only what your records support; leave the rest blank and note why.

Field	Formula / unit	Typical record
Crop, variety, planting–last harvest dates	text, dates	crop plan
Productive area	m²	site plan
Marketable yield	marketable kg ÷ m²	packhouse / sales
Grade-out	rejected kg ÷ harvested kg × 100	grading sheets
Water intensity	litres applied ÷ marketable kg	meter, controller
Fertilizer intensity	kg product (or kg N) ÷ marketable kg	invoices, stock
Energy intensity	kWh or fuel units ÷ marketable kg	bills, sub-meters
Labour intensity	hours ÷ tonne marketable	time sheets
Downtime / failures	events, hours	maintenance log

Reading the audit: compare the same crop and season across years first. A line that is much worse than your own best comparable cycle is a candidate bottleneck — then check with the diagnosis steps above.

Limitations: intensities differ by crop, variety, cycle length, climate, heating season and market grade, so do not compare against other farms' or published figures without matching those conditions. Shared meters must be split with a stated assumption.

A hypothetical worked example

Hypothetical numbers, for method only — not benchmarks for any crop or region.

Row	Cycle A	Cycle B	Cycle C
Marketable yield (kg/m²)	30	34	26
Rejects (%)	8	6	15
Water (L per kg marketable)	28	25	35
Labour (h per 1,000 kg)	40	36	48

Cycle C is worst on every row and its rejects almost double. A sensible next question is why rejects rose (a disease event, a heat period, a fertigation fault, a grading change) before assuming more area or new equipment is the answer.

Separating process, equipment and management problems

Process: the method is wrong or inconsistent (irrigation timing, pruning, harvest interval). Clue: results vary by shift, block or worker.
Equipment: the system cannot deliver what is set (vents that do not open, uneven drippers, a faulty sensor). Clue: logs show setpoints not reached, or measured uniformity is poor.
Management: decisions or planning are late (ordering, staffing, maintenance). Clue: problems follow delays, not weather or hardware.

A single bad cycle can mix all three. Record which one each piece of evidence points to.

What to do when data is missing

Use partial periods: one well-recorded month is better than a full season of estimates.
Use proxies with a clear label: pump run-hours × rated flow for water, delivery notes for fuel.
Start measuring now on one block: a water meter, a drain cup, a simple harvest-per-row sheet.
State what you do not know in any RFQ, so suppliers do not design around invented numbers.

When to bring in an adviser or lab

Ask a local extension service, agronomist or accredited lab when the comparison points to water quality, nutrient balance, root-zone disease or pest pressure. Irrigation-water tests (pH, EC, alkalinity, specific ions) are interpreted against crop and system — Penn State Extension's guidance is a useful starting point, but local specialists and regulations govern.

Your historic reports are a baseline, not a verdict. Build the comparison table, mark what is missing, test the top hypothesis on a small area, and only then decide whether you need an upgrade, an expansion or a process change.

Tools and next steps

Frequently asked questions

How many cycles do I need?: Two comparable cycles show a difference; three or more help you see whether it repeats. If you have only one, use it as the baseline and start measuring the next one carefully.
Should I compare with published yields?: Only with care. Published figures depend on crop, variety, climate, technology and period. Your own best cycle is usually a more honest reference.
What is the most useful single indicator?: Marketable kg per m² together with rejects. Total harvest hides grading losses that often explain the gap between cycles.
Can software do this for me?: Software helps once the records are consistent. The first comparison is usually faster in a spreadsheet built from your documents.
Does this guarantee better results?: No. It helps you choose what to investigate. Results depend on crop, variety, climate, water, market and how changes are carried out.

Sources

General planning guidance. Results depend on crop, variety, climate, water, regulation and market; numerical examples are hypothetical. Confirm decisions with qualified local specialists.