A Leaner, More Nuanced Approach to Classifying Smallholder Farmers

When studying small-scale producers, accurate classification into segments is essential. It helps in providing targeted recommendations and uncovering detailed insights into behaviors, challenges, and impacts within each farmer group. While various frameworks exist for this purpose, 60 Decibels (60dB) needed a model that would align with its lean data approach, where gathering comprehensive farm profiles isn’t always feasible.

Traditional Segmentation Approaches 

Historically, farmer classification frameworks have relied on detailed socioeconomic and farm profile data. Notable examples include those from AGRA/EPAR and CGAP. 

Though valuable, these frameworks often require granular farm data that can be difficult to gather quickly. Common segmentation dimensions—such as income, crop type, infrastructure access, or digital literacy—yield valuable perspectives but aren’t always viable for lean data studies.

Developing a New Segmentation Approach with Limited Data

60 Decibels frequently relies on reported land size as a proxy for farm scale and commercialization level. However, relying on land size alone has limitations. It can fail to correlate with productivity or market orientation, especially across diverse smallholder contexts. After engaging with partners like CGAP and other experts, we identified the need for a new classification approach that offers robust insights without requiring extensive data collection.

The 60dB Methodology: Agency and Commercialization as Core Dimensions

Our new segmentation approach centers on two critical dimensions: Commercialization and Agency. By evaluating farmers’ investments, use of labor, and market orientation, this model creates a more precise classification aligned with 60dB’s lean data principles.

Commercialization (75% Weighting)

• Hired Labor: Evaluates the extent to which farmers use hired labor, indicating operational scale.
• Investment in Equipment: Considers if farmers rent or own equipment as an indicator of input investment.
• Infrastructure Investment: Examines whether farmers invest in farm infrastructure, reflecting growth orientation.
• Produce Consumption vs. Sales: Assesses the proportion of produce sold versus consumed, a direct indicator of market orientation.

Agency (25% Weighting)

• Price Setting: Measures who sets the sale price, with self-determined pricing indicating higher autonomy.
• Price Perception: Evaluates the farmer’s satisfaction with selling prices, offering insights into bargaining power and agency.

Calculating Farmer Scores and Classifying Segments

Each of the indicators within the two dimensions is scored based on predefined categories. For instance, high use of hired labor earns up to 12.5 points, while infrastructure investment and consumption ratios similarly contribute to the overall Commercialization Score.

The Agency Score reflects price setting and perception factors, each with a potential 12.5 points. Summing the two scores yields a total classification score, segmenting farmers into three distinct groups:

• Subsistence Farmers (Score < 35):
  Farmers who consume most of their produce and have low market orientation or investment levels. For example, a subsistence farmer consumes a large portion of their harvest and shows limited investment in inputs.
• Pre-Commercialized Farmers (Score 35 - 60):
  These farmers sell a substantial portion of their produce and exhibit some market engagement but vary in investment intensity. A pre-commercialized farmer might sell part of their harvest, have moderate agency, and invest in labor or infrastructure occasionally.
• Commercialized Farmers (Score > 60):
  Farmers who treat their farms as businesses, selling most of their produce and investing in labor, equipment, or infrastructure. A commercialized farmer typically has higher agency and a strong market orientation, maximizing input usage to drive productivity.

Findings: Digital Farmer Service (DFS) Engagement Across Farmer Segment

This refined segmentation framework not only enhances 60dB’s ability to classify farmers with limited data but also generates actionable insights for each segment. With clearer distinctions among subsistence, pre-commercialized, and commercialized farmers, this methodology enables the design of tailored programs and interventions that address the unique challenges of each group. It’s a lean, efficient, and impactful model, proving that sometimes less data, when applied thoughtfully, can lead to more powerful insights.

Applying our segmentation model to smallholder farmers’ use of digital farmer services (DFS) in our study of Kenyan farmers revealed clear trends based on commercial orientation:

• Information & Advisory for Planting/HarvestingCommercialized and pre-commercialized farmers are more likely to use DFS for planting and harvesting advice. This group’s higher market orientation drives demand for insights that support productivity.Subsistence farmers show less engagement with advisory services, likely due to smaller scale and limited market focus.
• Market & Price InformationAll farmer segments actively use price and market information services, likely due to the wide availability of free price SMS services and government advisories in Kenya. Reliable, accessible market information is a priority for all farmers to optimize sales.
• Tools for Inputs, Equipment, Credit & InsuranceCommercialized and pre-commercialized farmers use DFS for inputs, equipment rentals, credit, and insurance more frequently than subsistence farmers, reflecting their focus on growth and risk management.

These insights suggest that DFS development should target the specific needs of each segment. While universally valuable price information services reach all farmers, advisory tools and access to inputs and credit are particularly impactful for commercially oriented farmers. This segmentation helps ensure that DFS offerings align with each farmer’s unique needs and goals.

For additional details, see our report, Digital Farmer Services in Kenya: The Farmer Perspective.