Smallholder farming systems in sub-Saharan Africa are highly heterogeneous, resulting in differences in farmers’ capacities, priorities, and responses to sustainable intensification (SI) innovations. This study assessed farmers’ perceptions of six SI innovations and evaluated their suitability across four farm typologies in the Mtunthama Extension Planning Area, Central Malawi. A mixed-methods participatory approach involving focus group discussions and farmer-led innovation ranking was used across 44 trial farmers. Qualitative data were analysed using thematic content analysis, while preference rankings were converted into weighted standardised scores for comparison across farm types. Kruskal–Wallis tests were applied to assess whether maize grain yield differed significantly across farm typologies by treatment. Farmers generally perceived SI innovations positively, identifying improved yields, soil fertility, food security, dietary diversity, and climate resilience as major benefits. Labour requirements, input constraints, management complexity, and limited training were identified as key barriers. Innovation preferences differed across farm typologies, reflecting variations in household resource endowment and livelihood strategies. Although conventional ridging with continuous sole maize received the highest proportion of first-choice rankings, conservation agriculture incorporating legumes, particularly maize-cowpea double-row strip cropping, achieved the highest overall preference scores across farm types. Yield analysis revealed significant differences across farm typologies for CA Maize-Lablab Rotation (H = 9.823, p = 0.020) and Conventional Maize (H = 12.834, p = 0.005), with resource-constrained farmers consistently recording lower yields. Field days attended by 253 independent observers confirmed community-level preference for cowpea-based CA innovations, with maize-cowpea strip cropping receiving the highest vote share at both sites. Post-project follow-up revealed that 68 out of 253 field day observers (26.9%) independently adopted cowpea-based conservation agriculture in the following season, of whom 58.8% were women. Adoption was driven by the early maturity of the cowpea variety under climate variability and access to a ready vendor market. These findings demonstrate that integrating farm typology analysis with participatory innovation assessment and post-project tracking improves the targeting of SI interventions and supports the design of context-specific innovation packages more likely to achieve sustained adoption than blanket technology recommendations.

Smallholder farming systems in Sub-Saharan Africa are heterogeneous in resource endowment and livelihood objectives, posing challenges for effective agricultural interventions. This study characterises farm system diversity in Mtunthama Extension Planning Area, Kasungu District, Malawi, and evaluates the suitability of sustainable intensification innovations across farm types. Data were collected through focus group discussions and household surveys involving 43 farming households. Maize was the focus crop because it is the staple food crop of the whole of Malawi and the crop with which legumes are rotated or intercropped in the sustainable intensification options evaluated. Principal Component Analysis was used to construct a farm typology, while innovation suitability was assessed using a multi-criteria decision analysis framework. Four distinct farm types were identified: High-Resource-Endowed farms (HRE), Moderately resource-endowed farms with Diversified Cropping systems (MDV), Moderately resource-endowed farms with Mixed crop-livestock orientation (MCL), and Resource-Constrained Maize-dominated farms (RCM). The multi-criteria assessment revealed substantial variation in suitability across types. Conservation agriculture recorded the highest overall score (0.68), followed by intercropping (0.54), while strip cropping showed the lowest (0.25). MDV achieved the highest overall innovation score (0.66), whereas MCL and RCM showed limited uptake. These findings demonstrate that sustainable intensification effectiveness is shaped by resource endowment and trade-offs between productivity, income, and sustainability. The decoupling of maize yield and household income suggests intensification pathways are context-specific. The framework provides a useful approach for identifying suitable practice combinations for each farm type and for designing targeted interventions.

Smallholder farming systems in sub-Saharan Africa are highly heterogeneous, resulting in differences in farmers’ capacities, priorities, and responses to sustainable intensification (SI) innovations. This study assessed farmers’ perceptions of six SI innovations and evaluated their suitability across four farm typologies in Mtunthama Extension Planning Area, Central Malawi. A mixed-methods participatory approach involving focus group discussions, key informant interviews, and farmer-led innovation ranking was used across 44 trial farmers. Qualitative data were analysed using thematic content analysis, while preference rankings were converted into weighted standardised scores for comparison across farm types. Kruskal-Wallis tests were applied to assess whether maize grain yield differed significantly across farm typologies by treatment. Farmers generally perceived SI innovations positively, identifying improved yields, soil fertility, food security, dietary diversity, and climate resilience as major benefits. Labour requirements, input constraints, management complexity, and limited training were identified as key barriers. Innovation preferences differed across farm typologies, reflecting variations in household resource endowment and livelihood strategies. Although conventional ridging with continuous sole maize received the highest proportion of first-choice rankings, conservation agriculture incorporating legumes, particularly maize-cowpea double-row strip cropping, achieved the highest overall preference scores across farm types. Yield analysis revealed significant differences across farm typologies for CA Maize-Lablab Rotation (H = 9.823, p = 0.020) and Conventional Maize (H = 12.834, p = 0.005), with resource-constrained farmers consistently recording lower yields. Field days attended by 253 independent observers confirmed community-level preference for cowpea-based CA innovations, with maize-cowpea strip cropping receiving the highest vote share at both sites. Post-project follow-up revealed that 68 out of 253 field day observers (26.9%) independently adopted cowpea-based conservation agriculture in the following season, of whom 58.8% were women. Adoption was driven by the early maturity of the cowpea variety under climate variability and access to a ready vendor market. These findings demonstrate that integrating farm typology analysis with participatory innovation assessment and post-project tracking improves the targeting of SI interventions and supports the design of context-specific innovation packages more likely to achieve sustained adoption than blanket technology recommendations.