🌾Agriculture

Field trials that actually tell you what works.

Fertilizer rates, irrigation schedules, seed varieties, pest control methods — agriculture runs on decisions made with too little data. Decision Process brings rigorous experimental design to the farm, at any scale.

Experiment Templates

Ready-to-run experiments

Fertilizer Rate Comparison

“Standard nitrogen application achieves near-optimal yield at lower input cost than high-rate application.”

Conditions

  • Low N (control)
  • Standard N
  • High N

Metrics

  • Yield per acre (lbs)
  • Input cost per acre ($)
  • Protein content (%)
Farm plotsStd N: 94% yield at 62% cost

Irrigation Scheduling A/B

“Deficit irrigation (80% ET replacement) maintains yield while reducing water use.”

Conditions

  • Full ET replacement (control)
  • 80% ET replacement
  • 60% ET replacement

Metrics

  • Yield per acre
  • Water use (acre-ft/ac)
  • Water use efficiency
Irrigation blocks80% ET: equal yield, −23% water

Seed Variety Trial

“Hybrid variety B outperforms the incumbent in drought conditions.”

Conditions

  • Incumbent variety (control)
  • Hybrid A
  • Hybrid B

Metrics

  • Yield per acre
  • Days to maturity
  • Standability rating (1–10)
Plot rowsHybrid B +18% yield (dry year)

Row Spacing Optimization

“Narrower row spacing increases canopy closure and suppresses weed pressure.”

Conditions

  • 30-inch rows (control)
  • 20-inch rows
  • 15-inch rows

Metrics

  • Yield per acre
  • Weed density (plants/m²)
  • Crop quality rating
Field strips20-inch optimal for yield+weed

Cover Crop Species Comparison

“Legume cover crops build soil nitrogen and reduce fertilizer requirements.”

Conditions

  • No cover crop (control)
  • Winter rye
  • Crimson clover
  • Radish mix

Metrics

  • Soil nitrate (ppm)
  • Subsequent crop yield
  • Input cost reduction ($)
Field plotsClover: +12 ppm N, −$40/ac cost

Worked Example

Three-rate nitrogen trial across 12 farm plots

A cooperative tests three nitrogen application rates across 12 farm plots over a single growing season. Plots are randomized to Low N, Standard N, or High N (4 plots each). Yield and input cost are recorded at harvest.

Results: yield_per_acre (lbs)

Low N — 80 lbs/ac (control)

mean: 2,840 lbs/acre

95% CI: 2,710–2,970

Standard N — 140 lbs/ac

mean: 3,210 lbs/acre

95% CI: 3,090–3,340

P(better than control) = 97%

High N — 200 lbs/ac

mean: 3,410 lbs/acre

95% CI: 3,280–3,540

P(better than control) = 99%

Standard N achieves 94% of High N yield at 62% of the input cost per acre. At 97% posterior probability, Standard N significantly outperforms Low N. The marginal return of High N does not justify the additional $38/acre input cost — Standard N is the economically optimal rate.

Explore other domains

🛒Retail🏭Manufacturing💻Digital

Run your first agriculture experiment

Private beta — tell us about your use case and we'll get you set up.