Insights into the costing of on-farm testing: A case study of common bean tricot trials by the National Agricultural Research Organization (NARO)

This report presents a comprehensive activity-based costing analysis of on-farm citizen science trials using the tricot (triadic comparisons of technologies) approach, focusing on a baseline case study of common bean (Phaseolus vulgaris) trials conducted by the National Agricultural Research Organization (NARO) at the National Crops Resources Research Institute (NaCRRI) in Uganda. Utilizing the 1000FARMS platform’s standardized tricot costing framework, this evaluation dissects the financial and operational mechanics of decentralized, on-farm testing to establish a transparent baseline for institutional planning and future project scaling.

KEY OPERATIONAL PARAMETERS
The baseline evaluation models a structured seasonal trial consisting of 120 participating farmers evaluating 24 distinct bean varieties. Each farmer manages three decentralized plots under local agronomic conditions, yielding a total network size of 360 plots across the target trial landscape. The institutional setup involves a multi-tiered execution strategy combining specialized NaCRRI scientific researchers (Breeders, Research Officers, Research Assistants), technical support personnel, and decentralized extension field agents who handle direct, localized interactions.

CORE FINANCIAL STRUCTURE AND INSIGHTS
An analysis of the trial’s financial architecture reveals a critical structural imbalance between centralized oversight and field execution across the primary cost pillars:
• Labor Costs (37%): Personnel wages are the largest driver of trial expenditures. This category is heavily dominated by the high internal daily rates of centralized institutional research staff rather than decentralized extension networks, highlighting a strong operational dependency on centralized supervision for routine trial management.
• Travel expenses (27%): Field logistics represent the true heavyweight of the operational budget, split between Per Diem (25%) and Travel Allowances (2%). Notably, centralized research team travel alone accounts for 51% of all travel expenses due to frequent, repetitive monitoring trips, severely limiting the program’s geographic elasticity and scalability.
• Material Costs (7%): Physical trial inputs—encompassing seed preparation, packaging, communication bundles, and localized participant training supplies—account for a smaller share of the total expenditure. This demonstrates that field logistics and personnel allocation, rather than physical inputs, determine the trial’s financial profile.

STRATEGIC COST DRIVERS AND SENSITIVITY
Sensitivity and scenario analyses demonstrate that the trial’s financial sustainability is disproportionately governed by institutional behavior rather than material scale. Increasing the number of participating farmers introduces highly favorable economies of scale, dramatically reducing the per-farmer marginal cost. Conversely, minor increases in research team field monitoring frequencies or fuel price indices cause significant, compounded budgetary expansions due to the high variable cost of mobilizing centralized institutional staff.

KEY STRATEGIC RECOMMENDATIONS
To transform on-farm testing from an expensive, centralized project-based model into a lean, highly scalable national framework, this report recommends a paradigm shift across three core areas:
1. Transition to Decentralized Local Delivery: Institutionalize a localized operational model where routine data collection, localized seed distributions, and day-to-day farmer engagements are fully delegated to decentralized field agents and local organizations already operating within the target sub-counties.
2. Implementation of Lean Research Supervision Protocols: Restructure the centralized research team’s role from operational execution to strict quality assurance. Centralized multi-role research caravans should be replaced by a single, agile trial manager utilizing a localized quality audit checklist (evaluating plot layouts, digital data entry integrity in ClimMob, and random verification protocols) during a heavily reduced calendar of field visits.
3. Hub-and-Spoke Geographic Clustering: Restructure distribution networks by clustering participating farmers into geographic hubs of 10-20 individuals each. Scheduling training, material payouts, and data collection hub-by-hub rather than farm-by-farm dramatically curtails logistical bottlenecks, minimizes vehicle wear, and leverages community cohesion to ensure long-term data quality.

By implementing these efficiency measures, NARO-NaCRRI can safeguard data integrity, reduce per-unit testing costs, and successfully scale the tricot approach across broader agro-ecological zones to deliver climate-resilient varieties to smallholders faster and more cost-effectively.