Introduction
The global food system is currently at a crossroads. Its adverse effects on planetary health are well known, yet it faces growing food demand within increasingly constrained conditions (Fraser and Campbell, Reference Fraser and Campbell2019; Global Panel on Agriculture and Food Systems for Nutrition, 2020). To achieve emissions reductions while pursuing development objectives, it is therefore necessary to effectively scale innovations able to strengthen the sustainability of the food system while meeting national development targets (Li et al., Reference Li, Wang, Chan and Manzini2014).
In agricultural research for development (AR4D), scaling initiatives typically begin with the goal of expanding a specific innovation or solution expected to drive positive change (i.e., emissions reductions) across extensive geographical areas (Woltering et al., Reference Woltering, Valencia Leñero, Boa-Alvarado, Van Loon, Ubels and Leeuwis2024). Scaling efforts thus focus on how to systematically transition from initial farmer adoption in pilot projects to self-sustaining and widespread uptake of innovations (Chandy et al., Reference Chandy, Hosono, Kharas and Linn2013; Cooley and Kohl, Reference Cooley and Kohl2016). However, for most food system innovations, the bottlenecks and opportunities for scaling exist not only in the innovation itself but in the non-technological factors that constitute the enabling environment for scaling (Van Loon et al., Reference Van Loon, Woltering, Krupnik, Baudron, Boa and Govaerts2020b).
Scaling initiatives frequently face constraints—such as innovation costs, market access, and lack of capital, particularly in the case of women, among others—that impact their ability to equitably achieve positive outcomes across diverse social groups (McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024; Schut, Leeuwis and Thiele, Reference Schut, Leeuwis and Thiele2020). Although agricultural technologies and management practices can enhance the lives of adopters and contribute to climate change mitigation, these innovations may also entail significant environmental and socioeconomic trade-offs over the long term or when implemented at scale (Castro-Nunez et al., Reference Castro-Nunez, Buriticá, Gonzalez, Villarino, Holmann, Perez, Del Río, Sandoval, Eufemia, Löhr, Durango, Romero, Lana, Sotelo, Rivera, Loboguerrero and Quintero2021). They can, for example, potentially lead to unintended harm or further marginalization of certain populations (Glover, Sumberg and Andersson, Reference Glover, Sumberg and Andersson2016; Schut, Leeuwis and Thiele, Reference Schut, Leeuwis and Thiele2020; McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024). Furthermore, while an innovation and intervention strategy may prove suitable and scalable in one context, it may not be in another context (Sartas et al., Reference Sartas, Schut, Proietti, Thiele and Leeuwis2020). Thus, for effective scaling aiming to contribute to food system transformation, stakeholders should also evaluate innovations based on how they align with biophysical (Wigboldus et al., Reference Wigboldus, Klerkx, Leeuwis, Schut, Muilerman and Jochemsen2016), social, and institutional circumstances (Woltering and Boa-Alvarado, Reference Woltering and Boa-Alvarado2019) as well as the both economic and technological needs of a given context. These considerations for selecting innovations have often been either completely overlooked or only partially addressed in the scaling literature (Dossou-Yovo et al., Reference Dossou-Yovo, Arouna, Benfica, Mujawamariya and Yossa2024; Khatri-Chhetri et al., Reference Khatri-Chhetri, Pant, Aggarwal, Vasireddy and Yadav2019; Wigboldus and Brouwers, Reference Wigboldus and Brouwers2016; Woltering and Boa-Alvarado, Reference Woltering and Boa-Alvarado2019). Additionally, in many Global South countries, where low-emission strategies in isolation are not necessarily prioritized, the environmental and social co-benefits of scaling innovations for low-emission food systems can motivate stakeholders and secure political support (Thornton et al., Reference Thornton, Whitbread, Baedeker, Cairns, Claessens, Baethgen, Bunn, Friedmann, Giller, Herrero, Howden, Kilcline, Nangia, Ramirez-Villegas, Kumar, West and Keating2018).
Various approaches and methods have emerged within the food system sector to tackle the complexity of scaling, such as ‘Scaling Readiness’ assessments (Sartas et al., Reference Sartas, Schut, Proietti, Thiele and Leeuwis2020), yet these often lack explicit attention to the non-technological factors that constitute the enabling environment for scaling, driven by institutional, biophysical, and social conditions, in practice (Van Loon et al., Reference Van Loon, Woltering, Krupnik, Baudron, Boa and Govaerts2020b). This paper addresses this gap and presents a multi-criteria framework for assessing the potential of food system innovations to achieve low-emission development at scale (Fig. 1). This serves as a support tool for evaluating potential opportunities, barriers, bottlenecks, and trade-offs to inform scaling strategies. Using a participatory ranking methodology, stakeholders tested the framework for innovations relevant to Nandi County, Kenya. They applied the framework in a workshop, collaborative format, with diverse actors engaged in the local context. The method draws on transdisciplinary experts to answer reflective questions throughout the assessment process. Importantly, stakeholders should apply the low-emission strategy-focused framework early in the innovation process, even before a scaling ambition is defined. After a prioritization exercise, they assessed the potential of three CGIAR innovations for achieving low-emission development at scale: improved livestock breeds and feeds, integrated aquaculture systems, and biogas technology contribute to achieving Kenya’s climate commitments outlined in its updated Nationally Determined Contributions (NDCs). These innovations not only align with Kenya’s target to reduce greenhouse gas emissions by 32% by 2030 but also advance sustainable practices that bolster resilience in its food systems and promote the achievement of other sustainable development outcomes.
Eleven criteria of the multi-criteria assessment framework for sustainable and inclusive scaling. Each of the ‘context’ sub-criteria are accounted for individually and weighted equally.
This paper is structured as follows. The ‘Conceptual framework: The multi-criteria scaling assessment framework’ section introduces the multi-criteria scaling assessment framework, with a literature review underpinning each criterion. The ‘Case study’ section provides background for the case study context, while the ‘Methodology’ section outlines the methodology applied. Within the Results section, the findings are clearly presented, while the Discussion section provides a thoughtful interpretation, highlighting key trade-offs and co-benefits. The conclusion section summarizes the key findings and opportunities for further research.
Conceptual framework: the multi-criteria scaling assessment framework
Many tools and approaches have been employed to prioritize innovations in agri-food systems, including simulation modeling (Shirsath et al., Reference Shirsath, Aggarwal, Thornton and Dunnett2017), cost–benefit analysis (Kumar et al., Reference Kumar, Murthy, Gumma, Khan P, Khatri-Chhetri, Aggarwal and Whitbread2018), economic surplus models (Stevanović et al., Reference Stevanović, Popp, Lotze-Campen, Dietrich, Müller, Bonsch, Schmitz, Bodirsky, Humpenöder and Weindl2016), and integrated assessment modeling (Rosegrant et al., Reference Rosegrant, Sulser, Mason-D’Croz, Cenacchi, Nin-Pratt, Dunston and Willaarts2017). While these methods offer valuable insights, participatory methods are recommended because they incorporate expert and stakeholder perspectives before scaling, accurately reflect local or ‘field’ realities, and provide the flexibility to consider multiple variables and system objectives (Thornton et al., Reference Thornton, Whitbread, Baedeker, Cairns, Claessens, Baethgen, Bunn, Friedmann, Giller, Herrero, Howden, Kilcline, Nangia, Ramirez-Villegas, Kumar, West and Keating2018). For instance, Khatri-Chhetri et al. (Reference Khatri-Chhetri, Pant, Aggarwal, Vasireddy and Yadav2019) used participatory approaches to evaluate innovations based on criteria such as adaptation benefits, adoption barriers, potential incentive mechanisms, and institutional support for scaling. Similarly, Dossou-Yovo et al. (Reference Dossou-Yovo, Arouna, Benfica, Mujawamariya and Yossa2024) employed participatory methodologies considering factors such as productivity, income, climate change adaptation and mitigation potential, technology cost, technical feasibility, gender inclusivity, and market demand.
The multi-criteria assessment framework aims to support collaborative prioritization of innovations with high potential for inclusive and sustainable scaling to achieve low-emission development in food systems. It assumes that scaling innovations for low-emission food systems requires an integrated approach. The methodology promotes participatory evaluation with stakeholders embedded in the intended scaling context of innovations based on their suitability for the target context, their economic viability, and their environmental and social trade-offs and co-benefits, among others. By integrating these criteria, the framework provides rapid guidance to development practitioners and decision-makers. It helps relevant stakeholders co-evaluate innovations that can support transforming food systems through low-emission pathways while achieving sustainable development. The aim of this assessment framework is to guide developing an innovation’s scaling strategy and gaining local insights that support it. The framework uniquely integrates the concept of environmental and social co-benefits and trade-offs into the scaling strategy, in line with responsible innovation approaches (Stilgoe, Owen and Macnaghten, Reference Stilgoe, Owen and Macnaghten2013). Below, we present each criterion in detail and its basis in the literature (Fig. 1).
Context
Determining how well innovations can scale depends significantly on the political, business, social, and cultural context as well as their interactions—an idea encapsulated in the phrase ‘context is king’ (Woltering and Boa-Alvarado, Reference Woltering and Boa-Alvarado2019). In scaling for food system transformation, assessments of the contextualized ‘innovation needs’ are relevant for the identification of gaps and points of entry (Anandajayasekeram, Reference Anandajayasekeram2016). Not only does there need to be a conducive technological environment, in terms of access to machinery and inputs (technological context), innovations for food system transformation also need to align with the climatic conditions, land, soil types, flora, and fauna of the target setting (biophysical context), especially at the agricultural production level (Shilomboleni and De Plaen, Reference Shilomboleni and De Plaen2019; Westermann et al., Reference Westermann, Förch, Thornton and Körner2018). Furthermore, they should align with the local and national regulations and laws to guarantee political support for scaling (political context) (Anandajayasekeram, Reference Anandajayasekeram2016; Merrey, Loboguerrero Rodriguez and Zeppenfeldt, Reference Merrey, Loboguerrero Rodriguez and Zeppenfeldt2023; Minh et al., Reference Minh, Zwart, Appoh and Schmitter2021). Since food systems have a strong link with the larger social system (social context), there is also a need to further consider these interlinkages when scaling innovations (Anandajayasekeram, Reference Anandajayasekeram2016; McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024; Minh et al., Reference Minh, Zwart, Appoh and Schmitter2021; Shilomboleni and De Plaen, Reference Shilomboleni and De Plaen2019). Scaling involves agency and decision-making in the household and the community and amongst stakeholder networks (McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024). Schut, Leeuwis and Thiele (Reference Schut, Leeuwis and Thiele2020) thus call for a greater emphasis on the processes of learning, negotiation, and conflict resolution, practices that affect behavioral change and ultimately adoption and scaling of innovations (Schut, Leeuwis and Thiele, Reference Schut, Leeuwis and Thiele2020). Whether or not an approach suits the local social context strongly affects the ‘responsible’ scaling potential. Focus on local context, for example, supports more inclusive approaches (McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024). The four criteria that make up the ‘context’ are therefore technological, institutional, social, and biophysical (see ‘Results’ section).
Adoption and diffusion
In today’s low- and middle-income countries (LMICs), current levels and patterns of innovation uptake are inadequate to facilitate the necessary transformations in food systems (Hambrey, Reference Hambrey2017). For effective transformation to happen, food system innovations should be readily accepted or ‘fit the system’ (Woltering and Boa-Alvarado, Reference Woltering and Boa-Alvarado2019) to be dispersed among the target users. Adoption and diffusion are often slow. We cannot assume they will take place automatically if an innovation is introduced. Rather, to allow for effective adoption and diffusion, innovation efforts must analyze local perspectives on the barriers as well as the opportunities of adoption (Anandajayasekeram, Reference Anandajayasekeram2016; Minh et al., Reference Minh, Zwart, Appoh and Schmitter2021) identifying existing approaches and entrenched sociocultural practices that may hinder or enhance uptake. For example, there might be structural or normative constraints to adoption by different social groups (Rietveld et al., Reference Rietveld, Farnworth, Nawaz, Timler, Tittonell, Van der Burg and Groot2024). Furthermore, stakeholders may lack adequate access to required resources, such as machinery, inputs, and information, or they may lack decision-making power, as many women do in LMICs (Meinzen-Dick et al., Reference Meinzen-Dick, Behrman, Menon and Quisumbing2012).
Innovation cost
The start-up, operational, and opportunity costs of innovations constitute a significant factor to be considered when scaling low-emission solutions for food system transformation (Merrey, Loboguerrero Rodriguez and Zeppenfeldt, Reference Merrey, Loboguerrero Rodriguez and Zeppenfeldt2023). Innovations need to be affordable and sustainable for end users beyond project time frames (Füller, Hutter and Kröger, Reference Füller, Hutter and Kröger2021; Shilomboleni and De Plaen, Reference Shilomboleni and De Plaen2019). Therefore, a participatory assessment by diverse end users should be conducted early on to anticipate affordability and willingness to pay (Olum et al., Reference Olum, Gellynck, Juvinal, Ongeng and De Steur2020). This will help stakeholders promoting an innovation to understand differing constraints faced by farmers, especially women, with varying resource limitations.
Market access
Potential market constraints need to be addressed when developing and scaling innovations for low-emission food systems to avoid negative prices and wage effects. Market systems can provide a built-in incentive for adoption at scale through the profit motive since the expectations of future profits encourage investment (Anandajayasekeram, Reference Anandajayasekeram2016; Merrey, Loboguerrero Rodriguez and Zeppenfeldt, Reference Merrey, Loboguerrero Rodriguez and Zeppenfeldt2023; Minh et al., Reference Minh, Zwart, Appoh and Schmitter2021). However, market systems also present one of the main bottlenecks for adoption in LMICs with limited formal market infrastructure (Jack, Reference Jack2013). Infrastructural challenges also pose a threat to market access (Chamberlin and Jayne, Reference Chamberlin and Jayne2013) as well as to the mobility of, in particular, women and vulnerable groups (Kihiu, Reference Kihiu2020).
Financial opportunities
While farmers and other end users need access to innovations and to financial services, scaling low-emission innovations itself requires mobilizing a broad range of financial resources from public, private, international, and domestic sources. These funds are crucial for absorbing a significant portion of the scaling cost (Minh et al., Reference Minh, Zwart, Appoh and Schmitter2021; Rosenstock et al., Reference Rosenstock, Lubberink, Gondwe, Manyise and Dentoni2020). However, the burden of last-mile delivery costs—such as reaching farmers in remote areas, covering regulatory system costs for seed certification, and ensuring a steady supply from private vendors or agricultural input providers (agrovets)—is often overlooked in scaling strategies (Rosenstock et al., Reference Rosenstock, Lubberink, Gondwe, Manyise and Dentoni2020). This oversight can lead to an overestimation of impact projections, without the so-called true cost accounting that also reflects (unintentional) environmental and social harm (Michalke, Kunz and Gaugler, Reference Michalke, Kunz and Gaugler2025). Furthermore, project developers must ensure a viable ‘exit strategy’ for project-level implementation and ensure longer-term financing opportunities to allow for sustained use. Planning for financial considerations will enable innovations to continue being adapted and reach further geographies and more users beyond the scope of donor-funded or externally financed projects.
Environmental and social co-benefits
Identifying potential co-benefits of scaling innovations on the environment (land, water, air, and biodiversity) as well as on society (poverty, undernutrition, and inequality) can help to maximize the positive impacts of scaling (Anandajayasekeram, Reference Anandajayasekeram2016). For example, Morales-Muñoz et al. (Reference Morales-Muñoz, Bailey, Löhr, Caroli, Villarino, LoboGuerrero and Castro-Nuñez2022) examine the co-benefits of coordinating climate action and peacebuilding in two regions of Colombia. Their study focuses on the communal management of land and water resources and the restoration of ecosystems through climate- and conflict-smart agriculture. The authors find that communities can enhance social cohesion and promote peace initiatives by fostering climate action that intentionally facilitates dialogue on the sustainable management of natural resources.
Environmental and social trade-offs
Outcomes may be beneficial in one aspect (e.g., productivity or profit) but detrimental in others (e.g., biodiversity and gender equity) (Mausch, Hall and Hambloch, Reference Mausch, Hall and Hambloch2020). By examining the potential trade-offs and unintended negative downstream impacts of the innovation on both social and environmental dimensions (McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024; Wigboldus and Brouwers, Reference Wigboldus and Brouwers2016), stakeholders can take measures to minimize unintended effects (Anandajayasekeram, Reference Anandajayasekeram2016). McGuire et al. (Reference McGuire, Rietveld, Crump and Leeuwis2022) highlighted the need to apply unique gender and social considerations through ex-ante assessments to understand how innovations might unintentionally harm a group of people or be ineffective in achieving a social equity goal.
Economic benefits
Ideally, scaling innovations for food system transformation should improve general economic performance in that sector, increasing productivity and market development to ensure sustainability (Anandajayasekeram, Reference Anandajayasekeram2016; Bachmann et al., Reference Bachmann, Natcher, Kulshreshtha, Baco, Akponikpe and Peak2016; Rosenstock et al., Reference Rosenstock, Lubberink, Gondwe, Manyise and Dentoni2020; Shilomboleni and De Plaen, Reference Shilomboleni and De Plaen2019). However, the extent to which innovations produce economic benefits is dependent on several intersecting dimensions, such as productivity increases, market access, or fit-for-purpose partnerships, and may not be felt equitably. These considerations call for inclusive and adaptive business models for climate-smart value creation (Rosenstock et al., Reference Rosenstock, Lubberink, Gondwe, Manyise and Dentoni2020).
Case study
This study was embedded within the Low Emissions Food Systems Initiative (Mitigate+) of the global research partnership for a food-secure future (CGIAR) between 2022 and 2024, operating in Colombia, Kenya, Vietnam, and China (CGIAR, 2022). Nandi County in western Kenya was selected as a case study for applying the prioritization framework due it its diverse agricultural commodities and agroecological zones (Fig. 2). The priorities of Nandi County align with Kenya’s national policy frameworks (Habermann and Zhang, Reference Habermann and Zhang2022), including the Kenya Climate-Smart Agriculture Strategy (2017–2026), which specifically promotes livestock development.
Location map of Nandi County in Kenya.
In Kenya, emissions from the food system account for approximately 72% of the country’s total emissions. Martius et al. (Reference Martius, Wassmann, Mwambo, Pingault and Guérin2023) highlighted two main priorities for action, in line with the national priorities identified in Kenya’s latest NDC that commits Kenya to abate greenhouse gas (GHG) emissions by 32% by 2030 relative to the business-as-usual scenario of 143 Mt CO2eq (Ministry of Environment and Forestry, 2020): (i) reduce emission intensities from enteric fermentation through improved management of livestock feed and manure on pastures and (ii) encourage climate-resilient food waste management.
Nandi County’s food system is characterized by dynamic production systems, low-cost food processing, and a variety of food system actors (Jalang’o and Habermann, Reference Jalang’o and Habermann2022). Biophysical drivers such as climate change, water scarcity, and soil degradation significantly impact Nandi’s food system, with climate change effects, such as droughts, erratic rainfall, and floods, becoming increasingly common (County Government of Nandi, 2017). The location of Nandi County within Kenya is shown in Fig. 2.
The northern part of Nandi receives rainfall ranging between 1,300 mm and 1,600 mm per year, while the southern half is affected by the lake basin atmospheric conditions and receives rainfall as high as 2,000 mm per year. Annual temperatures range between 18 °C and 22 °C during the rainy season but can rise as high as 26 °C in the lowlands of the Nyando escarpment (County Government of Nandi, 2017). These conditions make the county suitable for a wide range of agricultural activities including livestock production. The national census in 2019 noted that Nandi is one of the highest-producing dairy regions in Kenya with 309,038 dairy cattle, representing about 10% of the national dairy herd, due to the favorable biophysical conditions.
Livestock production, practiced by more than 80% of households, is a major economic activity in Nandi, with semi-intensive dairy systems dominating and contributing significantly to GHG emissions (Jalang’o and Habermann, Reference Jalang’o and Habermann2022). GHGs from soil, such as carbon dioxide and nitrous oxide, are also emitted during crop production. However, monitoring capacity limits specific data on GHG emissions for Nandi County (Jalang’o and Habermann, Reference Jalang’o and Habermann2022; Oertel and Erasmi, Reference Oertel and Erasmi2016). Current reporting to the national government relies on production indices compiled from livestock and crop production reports from extension officers (Jalang’o and Habermann, Reference Jalang’o and Habermann2022).
Methodology
Development of the framework for selecting innovations
The multi-criteria framework for assessing scaling potential and prioritizing innovations distinguishes itself from previous frameworks by incorporating environmental and social co-benefits and trade-offs as assessment criteria to contribute to the building of a scaling strategy. The framework uses 11 criteria to assess the enabling environment for scaling innovations and their potential to transform low-emission food systems. The literature cites similar approaches, such as the scaling scan tool that uses the ‘scaling ingredients’ for assessing the bottlenecks and opportunities for achieving a scaling ambition (Jacobs, Ubels and Woltering, Reference Jacobs, Ubels and Woltering2018). While we recognize these are not exhaustive, they enable us to evaluate potential innovations for low-emission food system development.
Pre-selection of innovations for analysis
The study adopted the approach used by Khatri-Chhetri et al. (Reference Khatri-Chhetri, Pant, Aggarwal, Vasireddy and Yadav2019) for pre-selecting innovations for scoring. First, we surveyed scientists from several CGIAR organizations working on low-emission food systems. The survey yielded a list of 18 CGIAR innovations with the potential to reduce food system GHG emissions in Nandi County, originating from five CGIAR research centers: the Alliance of Bioversity International and CIAT, the International Maize and Wheat Improvement Center (CIMMYT), the International Livestock Research Institute (ILRI), the International Rice Research Institute (IRRI), the International Water Management Institute (IWMI), and WorldFish. Scientists were selected based on their research experience on low-emission food systems and were asked to suggest innovations based on three criteria: the food system GHG mitigation potential of the innovation, applicability in the Global South, and the likelihood of the innovation generating co-benefits if scaled.
From this list, three innovations or bundles of innovations were pre-selected by the research scientist working on low-emission food systems in Kenya and local government representatives from Nandi County. The selection was based on the GHG mitigation potential of the innovation or innovation bundle, their ability to deliver Sustainable Development Goal (SDG) co-benefits as well as their alignment with Kenya’s and Nandi County’s development goals. The selection of the innovation or innovation bundles were also based on their ability to contribute to the achievement of Kenyan government NDC. The selected innovations include improved livestock breeds and feeds (ILBF), integrated aquaculture practices (IAP), and biogas technology (BT). A brief description of the GHG reduction potential of these technologies is provided in Table 1.
Description of the selected innovations with their GHG mitigation potential
Data collection and analysis
Two participatory assessment workshops—held on December 7, 2022, and February 20, 2023—provided data for the study, in line with previous participatory approaches in the literature for prioritizing innovations for scaling (Jacobs, Ubels and Woltering, Reference Jacobs, Ubels and Woltering2018; Khatri-Chhetri et al., Reference Khatri-Chhetri, Pant, Aggarwal, Vasireddy and Yadav2019). The participants invited to the workshops were representatives of member organizations of the multi-stakeholder platform of climate-smart agriculture (CSA) in Kenya who had carried out field activities in the Nandi Country for at least 2 years. This specific audience was chosen for the exercise due to their expertise in CSA and food systems within the Kenyan context as well as their understanding of Nandi agri-food system dynamics and needs.
The stakeholders included representatives from Kenya’s Ministry of Agriculture and Livestock Development, NGOs working on CSA, Nandi’s local government, farmer associations, consumer protection organizations, universities, and local research organizations. Participants were invited as experts on the innovations being discussed during the scoring process. In total, the first workshop included 30 participants (of whom 13 were women), while the second included 28 participants (of whom 15 were women) from the Kenya CSA multi-stakeholder platform and research centers.
Participants were first introduced to the scoring framework. Subsequently, scientists delivered a presentation on each innovation, based on their direct involvement in the innovation’s research and development. Key topics covered during the presentations included an overview of innovation, requirements, shortcomings, and an example of the innovation’s implementation.
After each presentation, workshop participants were divided into two groups, each with a facilitator and a rapporteur from the workshop organization team. In these groups, the innovations to be scored were discussed in detail, following the criteria in the framework and the accompanying questions developed from the criteria. Stakeholders were asked to score the innovations, considering Nandi County as the target context for scaling the technology. They provided a score ranging from 0 to 3 for each criterion (Table 2). A score of 3 meant that the criterion was highly favorable for scaling, while 0 indicated unfavorable.
Results of the scoring exercise to identify potential innovations for food systems transformation based in Nandi County, Kenya
After the group discussion for each innovation, individual surveys were conducted with participants to obtain their perception of the relative importance (in percentage) of each criterion for scaling the low-emission food system innovations. The data collected was inputed into Microsoft Excel, and the average score per question, the average score per criterion, and the relative importance of each criterion for scaling the innovations was calculated. The reasons justifying each score, collected through qualitative responses, were compiled along with each criterion. In cases where the scores per question were less than 2, information on the necessary measures to ensure successful scaling was also solicited. The results of the workshops are complemented by a review of literature relevant to the potential of the innovations in Nandi County.
Results
Scoring results overview
The results of this study provide a comprehensive analysis of the potential of three CGIAR innovations to scale toward low-emission development in the food system within Nandi County, Kenya. Table 2 presents the findings from the scoring process, highlighting the average scores per criterion and perceived relative importance of each criterion to scaling. Here, we show that out of a total score of 33 points, ILBF obtained 25 out of 33 (75%), IAP 27 points (81%), and BT 26 (76%) (Table 2). We now provide quantitative and qualitative insights gathered from stakeholders for each of the criteria.
Assessing the conditions for sustainable scaling of CGIAR innovations in Nandi County
Technological context
The workshop participants identified the difficulty in accessing improved livestock breeds and fodder as a key barrier to scaling the ILBF innovation package, leaving the technological context with a low score (1.75/3). This is in line with previous reports from the Nandi County government (County Government of Nandi, 2017), which noted that key dairy production bottlenecks included inadequate feed availability, especially during the dry season, high infertility, and poor adoption of production technologies in livestock husbandry. With less than 1% of the animals under zero grazing (County Government of Nandi, 2017), the problem of available pasture becomes highly significant. Stakeholders also noted a lack of supporting or complementary innovations. They discussed that scaling ILBFs in Nandi requires more milk collection facilities and better milk collection systems, especially transportation systems. Therefore, to successfully scale this innovation, lead organizations must ensure equitable access to improved breeds and feeds and to information on their management as well as create a well-functioning milk collection system.
The results show that for IAP, the technological context of Nandi had a mid-high score (2.25/3). Nandi County lacks supporting technologies needed to scale IAP, such as fingerlings, storage support, transportation, and drying facilities. According to stakeholders, aquaculture farmers can bypass these barriers by forming cooperatives through which they can ensure the availability of these services.
For BT, participants noted that although there was an urgent need for the innovation, Nandi County has limited supporting technologies and services available after deploying the BT technology. These complementary services and technologies are crucial for biogas users. As a result, this innovation also received a mid-high score for this criterion (2.25/3). The supporting technologies and services include biogas chambers as well as repair and maintenance facilities within the county. This finding aligns with another study which noted that inadequate post-installation support and a lack of sufficient technicians and artisans are factors restricting the adoption of biogas technology in Nandi, despite the county’s great potential (ETC Group, 2017).
Biophysical context
The biophysical context received a high score (3/3) for the ILBFs innovation package, indicating that the technology is well-suited to the biophysical environment in Nandi County. Most parts of the county receive substantial rainfall (>1,200 mm), yielding favorable biophysical conditions for the innovation.
This criterion received a high score (2.75/3) also for IAP. Nandi County, with its 8 dams and 12 major rivers, has significant potential for fish farming and fishing (County Government of Nandi, 2017). However, this potential remains untapped due to the substantial investment and capital required for boats, harvesting gear, and skilled personnel (County Government of Nandi, 2017). For BT, the biophysical environment of Nandi is highly suitable due to high livestock rearing and availability of cow dung (the main feedstock for biogas digesters). The livestock sector is crucial in Nandi, contributing to more than 94% of enteric methane production (Ndung’u et al., Reference Ndung’u, Bebe, Ondiek, Butterbach-Bahl, Merbold and Goopy2018), thereby making biogas technology a suitable measure for reducing methane emission at the household level.
Institutional and social context
The institutional and social context for ILBF received the highest score possible (3/3). The ILBF innovation package aligns well with the county’s policies and fits seamlessly with the cultures and lifestyles of potential adopters. An example of such a policy is the Kenya Climate-Smart Agriculture Strategy (2017–2026), which promotes livestock development in Nandi County and across the country. The ownership of cows plays a significant role in Nandi culture. Productivity gains associated with ILBF attract livestock keepers to this innovation (Ndung’u et al., Reference Ndung’u, Bebe, Ondiek, Butterbach-Bahl, Merbold and Goopy2018).
The high score (3/3) for IAP indicates that existing laws and policies support this innovation. Respondents noted that the Kenyan government has promoted IAP in various ways, such as including IAP in national climate action plans and training youths in aquaculture management to enhance fish production. Hambrey (Reference Hambrey2017) also noted that the government of Kenya is supporting improved production efficiencies and the reduction of post-harvest losses. Workshop participants identified the social setting of Nandi County as favorable for scaling, with a score of 2.75/3. However, respondents indicated the need to consider cultural differences in fish consumption within the county.
The national and Nandi County governments promote biogas, indicating favorable institutional conditions (3/3) for BT innovations. One notable initiative is the creation of the national biogas program by the Kenyan government. According to stakeholders, BT aligns with Kenya’s NDCs and the National Climate Change Act. BT is also supported by Kenya’s Climate-Smart Agriculture Strategy and is listed in the national climate change policy. However, stakeholders noted that the technology is not socially appealing. Farmers face challenges in transporting cow dung from grazing spots to processing facilities, with limited resources available for zero grazing. Additionally, there is limited social acceptance of the technology among men, posing a problem to its implementation.
Adoption and diffusion
This study found that the potential adoption and diffusion of ILBFs is unfavorable, receiving a low score of 1.75 out of 3. Contributing factors to this score include a lack of expensive technical support and limited access to new breeds and fodder. Although platforms such as DigiFarm and DigiCow provide digital access to information on farm inputs and cow management, they either ignore the specific needs of ILBFs or have limited reach (Jalang’o and Habermann, Reference Jalang’o and Habermann2022). Furthermore, improved livestock breeds often require artificial insemination, which is expensive and labor intensive (Jalang’o and Habermann, Reference Jalang’o and Habermann2022). To overcome these adoption barriers, stakeholders emphasized the need to increase the number of farmer cooperatives to raise funds and facilitate capacity building. This approach can help disseminate improved breeds and information on optimal feeding practices more effectively.
The potential adoption of IAP was found to be low, with a score of 1.25 out of 3. Workshop participants said the low score is due to several factors, including limited capacity for implementing IAP, lack of access to (organic) feeds, low awareness about the technology, high initial investment capital, and limited market linkages. Addressing these challenges requires the provision of extension services and the implementation of financial mechanisms for scaling, such as low-interest loans and grants.
The widespread adoption of biogas as cooking fuel and lighting could significantly reduce GHG emissions and alleviate energy poverty. However, the initial cost of constructing biogas units remains a major barrier to widespread adoption (Mwirigi et al., Reference Mwirigi, Balana, Mugisha, Walekhwa, Melamu, Nakami and Makenzi2014). Stakeholders highlighted that the adoption potential of biogas technology is relatively low, with a score of 1.75 out of 3, particularly among young people who perceive the maintenance process as difficult and aesthetically unappealing. Similar findings were noted in Kiambu County by Wamuyu (Reference Wamuyu2014). Another factor hindering biogas adoption in Nandi is the need for customized cooking stoves, which adds to the financial burden since households willing to adopt biogas technology must purchase these specialized stoves. However, stakeholders also pointed out that the provision of slurry from biogas digesters, which serves as a source of bio-fertilizer, is a co-benefit that motivates household adoption.
Innovation cost
The cost criterion for the ILBF innovation received a low score of 1.125 out of 3. Stakeholders perceive the cost of purchasing new breeds and improved feeds to be high for the average farmer in Nandi. Factors contributing to the high start-up costs include the expenses for purchasing semen, seeds for fodder, and extension services. Practicing zero grazing may bring unfavorable trade-offs, including difficulties to allocate land for feed crops. Ongoing managerial costs also add to the burden, including maintenance of fodder, harvesting, labor for feeding and milking, veterinary services, irrigation, and processing.
Currently, livestock keepers apply zero grazing to less than 1% of animals in the country. They graze 75% of animals in the county on natural pasture, with the remaining animals partly grazed and partly housed (Jalang’o and Habermann, Reference Jalang’o and Habermann2022). Additionally, dairy cows in the county consist mainly of exotic breeds (Friesians and Ayrshire) and their crosses with local breeds, which aligns with the views of stakeholders in this study. Respondents identified the use of government subsidies and increased mechanization as potential solutions to reduce start-up and operational costs. According to the County Government of Nandi (2017), dairy production can be doubled with adequate investment, as it is currently far below its potential.
Stakeholders also noted the expensive IAP startup costs for the creation of ponds, purchasing feed, fingerlings, and providing energy. Operational costs, such as maintenance, labor, and feed purchases are also unaffordable for many people, explaining the low score of 1.75/3. This finding is consistent with the findings of Tran et al. (Reference Tran, Ogello, Outa, Muthoka and Hoong2023), who also observed a high initial cost of implementation, which may be challenging for small-scale farmers to afford. Additionally, Tran et al. (Reference Tran, Ogello, Outa, Muthoka and Hoong2023) noted a lack of technical expertise and infrastructure to support the use of IAP technologies.
Participants also noted the high cost of BT, at $600–$900 per household, making it less affordable for lower-income households. This is consistent with other studies in the literature (Mwirigi et al., Reference Mwirigi, Balana, Mugisha, Walekhwa, Melamu, Nakami and Makenzi2014; Nguu et al., Reference Nguu, Ndivo, Aduda, Nyongesa and Musembi2014; Patman and Moronge, Reference Patman and Moronge2015; Wachera, Reference Wachera2009), which have reported high initial installation costs as a major hindrance to adoption. To overcome these challenges, stakeholders highlighted the need to explore other financial mechanisms, such as incentives, subsidies, extension of loan facilities to small-scale farmers, and social investments, to aid in the scaling of this innovation.
Market access
The ‘availability of markets’ criterion received a high score (3/3) for the ILBF innovation package, indicating that markets for milk and beef are highly accessible in this context. According to the County Government of Nandi (2017), both Nandi County and Kenya are high consumers of dairy products. For instance, milk is sold through both formal and informal markets. In informal markets, milk is sold to intermediaries, brokers, milk hawkers, urban center consumers, and other farmers or hotels. In formal markets, milk is marketed through farmer organizations.
The markets for IAP products are also promising, with stakeholders awarding a full score of 3/3. Respondents reported a high demand for fish in Kenya, and the government promotes fish markets through sensitization campaigns on fish consumption. The market for biogas remains strong in Nandi (3/3), as cooking energy and organic fertilizer are in high demand, especially in rural and peri-urban areas. However, the limited supply of materials and their costs are factors that restrict the potential market value.
Financial opportunities
Smallholder farmers seeking to purchase improved breeds and seeds face a significant bottleneck in access to low-risk financing options as well as to entering into service provider arrangements. The financial opportunities for scaling ILBF recorded a mid-score of 2 out of 3. Finance is crucial for scaling technologies, especially among smallholder farmers. When finance is absent, scaling technologies among these farmers becomes difficult. Bisheko and Rejikumar (Reference Bisheko and Rejikumar2023) also noted the lack of access to finance as one of the barriers to scaling innovations in Sub-Saharan Africa and Asia. Similarly, Van Loon et al. (Reference Van Loon, Woltering, Krupnik, Baudron, Boa and Govaerts2020a) reported that farmers had low access to finance for scaling agricultural innovations in Zimbabwe and Mexico. To overcome financial challenges for scaling, stakeholders identified the need for governments, international donors, and private finance schemes to subsidize the scaling of innovations or provide loans for scaling high-risk innovations.
For scaling IAP (2.5/3), stakeholders noted that financial opportunities exist through loans from government agencies, programs, donors working on low-emission development, and the private sector. But these opportunities remain underexplored. For BT, stakeholders highlighted the need to explore other financial mechanisms, such as incentives, subsidies, extending loan facilities to small-scale farmers, and attracting social investors to aid in scaling this innovation. However, the possibility of leveraging finance from these sources and other external funding is low, which explains why financial opportunities received a mid-high score of 2.5 out of 3. Despite the availability of various financial mechanisms, there is a need for sensitizing farmers, as stakeholders believe that many in the sector are unaware of these different streams of income for scaling.
Environment and social co-benefits
Co-benefits have become a major topic in climate mitigation discourse and are often cited in climate-related decision-making as factors that can significantly influence the outcomes of direct cost–benefit evaluations (Intergovernmental Panel on Climate Change, n.d.). The criteria for environmental and social co-benefits received a score of 2.25 out of 3 for ILBF. The potential of improved fodder to trigger carbon sequestration and reduce per-unit GHG emissions from dairy production were identified as environmental co-benefits. A social co-benefit noted was the anticipated increase in income due to the increased production of milk and meat, leading to improved living standards.
IAP recorded a high score of 2.85 out of 3, with several environmental and social co-benefits identified. These include the potential to reduce resource use, such as water and fertilizer, diversification of income sources for adopters, and improved nutrition. According to Hambrey (Reference Hambrey2017), aquaculture can help offset GHG emissions through carbon sequestration.
BT has many co-benefits, such as the conservation of forests through reduced logging as a source of fuelwood, reduction in GHG emissions, and soil conservation through fertilizer application. Stakeholders also noted that this innovation has the potential to reduce conflicts between households, as they will not have to compete over wood fuel for cooking. This aligns with the findings of Jain et al. (Reference Jain, Newman, Nizhou, Dekker, Le Feuvre, Richter and Thompson2019), who reported that biogas technology offers environmental and social co-benefits such as climate change mitigation, contribution to a circular economy, and improvement in health and sanitation through better solid waste management.
Environmental and social trade-offs
When determining an optimal mitigation measure to limit GHG emissions (Verspecht et al., Reference Verspecht, Vandermeulen, Ter Avest and Van Huylenbroeck2012), stakeholders should consider both potential ‘winners and losers’. Trade-offs that might arise when scaling ILBF include the prevalence of insects and the use of chemical sprays in fodder, the risk of water scarcity due to livestock intensification, and the potential loss of local, more-resilient breeds. For IAP, stakeholders noted that introducing fish species might negatively impact native species in rivers. The workshop participants also highlighted some environmental and social trade-offs with scaling BT, such as the disruption of social interactions during activities such as firewood fetching and the aesthetically unappealing process of managing biogas digesters. Scaling organizations must be aware of these trade-offs to maximize the positive impacts and minimize the unintended consequences of scaling on lives, livelihoods, and the environment in Nandi. The average score across all three innovations for this criterion was 2 out of 3.
Economic benefits
This criterion received a score of 2.25/3 for ILBF. According to participating stakeholders, while the adoption of ILBF production may not necessarily reduce production costs, productivity is likely to increase. However, many factors affect productivity and profits, such as farm management, scale of production, and business models, which were not examined in the survey. Therefore, an in-depth study of the factors affecting profit margins is needed before drawing conclusions about the economic benefits of innovation.
Stakeholders further anticipated high IAP returns as economic benefits of scaling, due to increased productivity and the potential rise in employment. However, they indicated a need for detailed risk evaluation and cost–benefit analysis before implementing the technology. The economic benefits of BT were perceived by stakeholders as high (3/3). The ability of BT to save time and replace inorganic fertilizer and firewood, which appear to be cheaper, were highlighted as key benefits. Stakeholders also noted that project promoters of anaerobic digestion can obtain additional funding from carbon credits by using appropriate monitoring, verification, and reporting methodologies to demonstrate the GHG mitigation potential of the technology.
Our findings indicate that for ILBF, the demand (availability of a market) for the product promoted by the innovation was rated by participants as having the highest relative importance to scaling (11.85%), while social context was given the lowest relative importance (5.17%). This disparity suggests a need for greater awareness among partners regarding the significance of understanding ‘for whom’ scaling is pursued and highlights the importance of expanding the use of social suitability studies. For IAP, financial opportunities for scaling were rated the highest (12.29%), while the biophysical context received the lowest score (5.59%). This underscores the need to identify bankable business models for IAP. For BT, technology cost was rated the highest in relative importance (10.97%), while social context, once again, received the lowest rating (7.62%).
Discussion
The role of context-specific non technological factors
The participatory assessment used in this study highlights that non-technical dimensions play a decisive role in shaping the scalability of agricultural innovations. For the innovations considered in our study, ILBF obtained a score of 24 out of 33 (75%), IAP obtained 27 points out of 33 (81%), and BT 26 out of 33 (76%) reflecting their scalability potential. These outcomes offer insights into the enabling conditions required and adjustments needed for scaling these technologies effectively ILGF, IAP and BT Innovations which demonstrated strong biophysical suitability and mitigation potential were constrained by high cost of implementation, limited access to finance and perceived limited acceptance, reducing their scalability potential despite favorable agronomic indicators.
These findings support systemic accounts of agricultural scaling that technological performance alone is insufficient to predict scaling success, and rather that scaling depends much more on perceived complexity and ‘enabling conditions’ such as governance, social norms and economic structures, than only on the innovation itself (Brouwers, Reference Brouwers2016; McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024; Thornton et al., Reference Thornton, Whitbread, Baedeker, Cairns, Claessens, Baethgen, Bunn, Friedmann, Giller, Herrero, Howden, Kilcline, Nangia, Ramirez-Villegas, Kumar, West and Keating2018; Wigboldus et al., Reference Wigboldus, Klerkx, Leeuwis, Schut, Muilerman and Jochemsen2016; Woltering et al., Reference Woltering, Valencia Leñero, Boa-Alvarado, Van Loon, Ubels and Leeuwis2024). For instance, Khatri-Chhetri et al. (Reference Khatri-Chhetri, Pant, Aggarwal, Vasireddy and Yadav2019) focused on adaptation benefits, barriers to adoption, potential incentive mechanisms, and institutional support for scaling. Similarly, Dossou-Yovo et al. (Reference Dossou-Yovo, Arouna, Benfica, Mujawamariya and Yossa2024) emphasized productivity, income, climate change adaptation and mitigation potential, technology costs, technical feasibility, gender inclusivity, and market demand as key factors. Merrey, Loboguerrero Rodriguez and Zeppenfeldt (Reference Merrey, Loboguerrero Rodriguez and Zeppenfeldt2023) underscored the importance of mission-oriented communication systems, strong partnerships (local, national, and international), supportive policies and institutions, access to financing, and market demand as essential components of the enabling environment for scaling. Although these factors may vary slightly between studies, they all underscore the need to understand the specific contextual realities where the innovation is being scaled. This idea is further supported by Woltering and Boa-Alvarado (Reference Woltering and Boa-Alvarado2019) who reiterate the importance of considering context-specific factors for the success of any scaling venture.
The analysis further reveals that policy frameworks and institutional incentives create both enabling and constraining conditions for innovation uptake. In Nandi, innovation alignment with county development strategies and existing agricultural programs such as the Kenya Climate-Smart Agriculture Strategy (2017–2026), Kenya’s NDCs and the National Climate Change Act facilitated stakeholder buy-in and coordination, while gaps in limited public financing mechanisms were perceived to slow implementation. Cultural norms and livelihood priorities also influenced adoption dynamics, particularly where innovations required changes in household labor allocation, gender roles, or consumption practices. Socio-economic factors such as land tenure security, access to credit, and farmers’ risk tolerance emerged as critical determinants of investment willingness and sustained adoption. Together, these non-technical factors shape not only whether innovations are adopted, but also who benefits, at what scale, and under what conditions, raising important equity and inclusiveness considerations for scaling strategies.
Importantly, the application of a multi-criteria assessment framework generated insights that would not have been captured through economic methods of innovation assessment such as a conventional cost–benefit analysis. A conventional cost–benefit analysis in Nandi may prioritize short-term financial returns like productivity gains and input–output efficiency, favoring technologies with rapid ‘payback’. However, this only offers a partial view because it assumes stable institutions and uniform farmer behavior, overlooking structural constraints, inequality, gender and distributional effects, environmental externalities, governance feasibility in a complex smallholder system (Klerkx, Van Mierlo and Leeuwis, Reference Klerkx, Van Mierlo, Leeuwis, Darnhofer, Gibbon and Dedieu2012; Wigboldus et al., Reference Wigboldus, Klerkx, Leeuwis, Schut, Muilerman and Jochemsen2016; van Wijk et al., 2020). By contrast, the multi-criteria framework enhances investment and policy decisions by integrating economic, social, institutional, environmental, and governance factors. It reveals hidden adoption barriers, supports trade-off analysis between short-term profitability and long-term sustainability, and guides complementary interventions.
The value of accounting for co-benefits and tradeoffs for sustainable food system transformation
A key contribution of this paper is the assessment of environmental and social co-benefits and trade-offs, which are often overlooked in the food system innovation scaling literature (Morales-Muñoz et al., Reference Morales-Muñoz, Bailey, Löhr, Caroli, Villarino, LoboGuerrero and Castro-Nuñez2022). Organizations need to actively consider these co-benefits in each scaling effort to maximize positive impacts on livelihoods, ecosystems, and communities in the target areas, thus promoting more responsible scaling (McGuire et al., Reference McGuire, Al-Zu’bi, Boa-Alvarado, Luu, Sylvester and Leñero2024). In the broader context of climate change mitigation, particularly within low-emission food systems, co-benefits have become a focal point and are frequently emphasized in climate-related decision-making, as they can significantly influence the outcomes of traditional cost–benefit analyses (Intergovernmental Panel on Climate Change, n.d.). By incorporating a co-benefits approach when scaling low-emission food system innovations, stakeholders can adopt a holistic framework that not only supports ambitious climate mitigation policies but also addresses non-climate objectives simultaneously. Therefore, it is essential for any scaling initiative to understand the co-benefits of low-emission food system innovations. Doing so will help build a robust knowledge base, facilitate more effective mitigation strategies, and identify synergies with other development outcomes (Cohen et al., Reference Cohen, Cowie, Babiker, Leip and Smith2021).
Furthermore, not all scaling efforts result in positive outcomes (Herrero et al., Reference Herrero, Thornton, Mason-D’Croz, Palmer, Benton, Bodirsky, Bogard, Hall, Lee, Nyborg, Pradhan, Bonnett, Bryan, Campbell, Christensen, Clark, Cook, de Boer, Downs and West2020). In some cases, scaling can lead to unintended consequences for environmental and social systems (Castro-Nunez et al., Reference Castro-Nunez, Buriticá, Gonzalez, Villarino, Holmann, Perez, Del Río, Sandoval, Eufemia, Löhr, Durango, Romero, Lana, Sotelo, Rivera, Loboguerrero and Quintero2021; Mausch, Hall and Hambloch, Reference Mausch, Hall and Hambloch2020). Our study demonstrates that while the evaluated innovations represent key breakthroughs in Nandi’s food system, they can also generate significant trade-offs and unintended negative impacts on other SDG indicators in the region, such as increased poverty, reduced social inclusion, rising inequality, and challenges in healthcare, among others. These trade-offs and unintended consequences must be carefully addressed to achieve genuine food system sustainability.
Limitations
Conducting this study was not without its challenges and limitations. One of the key difficulties encountered was the perceived complexity of the numerous criteria involved. The assessment framework has 11 criteria, each of which can be interpreted in multiple ways and depends on the availability of technical expertise and sufficient contextual information as shown in section ‘Conceptual framework: The multi-criteria scaling assessment framework’.
While the study provides valuable insights into the potential of these innovations to contribute to low-emission food systems and future scaling efforts, it offers only a snapshot of opinions and interpretations that may evolve over time. Adoption behavior, institutional capacity, climate risks, market incentives, and policy frameworks evolve over time. This means that future changes such as improved extension services, new subsidies, climate shocks, or market shifts could alter the feasibility and attractiveness of innovations in ways not captured by the current analysis.
Furthermore, the workshop relied on secondary data presented and contributions to scalability, limitations and trade-offs discussed by the workshop participants were anecdotal or based on their expertise in similar locations. Although structured facilitation and triangulation reduce bias, personal experience, institutional interests, as well as group dynamics may still have influenced scoring outcomes. This bias introduces uncertainty and means that results reflect informed, though subjective perspectives rather than purely objective measurements. As a result, certain indicators rely more heavily on qualitative assessment rather than robust statistical validation. The results should therefore be complemented by formal analyses, such as through foresight studies and scenario modelling.
Also the pre-selected innovations by a group of scientists and local government representative from Nandi County prior to the workshop may differ from the innovation that would be selected by other stakeholders in Nandi County. It is thus critical to ensure that the needs and preferences of intended users are accounted for early in the design process, rather than only at the final stages of delivery, which may be too late to adjust significantly and ensure sustained adoption.
Lastly, the analysis focuses on one geographical and institutional context (Nandi County, Kenya). While this allows for deep contextual understanding, the results may not automatically apply to other regions with different agro-ecological conditions, market access, cultural norms, governance systems, or policy environments. Adoption barriers, institutional readiness, and economic viability may differ substantially elsewhere, so the findings should be interpreted as context-specific rather than universally representative. The framework therefore needs to be seen as adaptable to local contexts.
Conclusion
This study demonstrates the value of applying a participatory multi-criteria framework to assess the scalability of agricultural innovations beyond technical performance alone. By systematically incorporating institutional, policy, socio-cultural, financial, markets and environmental dimensions, the approach provides a more realistic understanding of the enabling conditions and constraints that shape scaling pathways in complex food systems. Applying the framework, we aimed to identify the key conditions for facilitating the scaling of selected CGIAR innovations (ILBF, IAP, and BT) toward achieving low-emission food system transformation and sustainable development in Nandi County, Kenya. The case of Nandi County illustrates that innovations with strong technical potential may face significant barriers related to context specific institutions, policy, socio-cultural, financial, markets or even environmental dimensions.
The study concludes that significant scaling challenges stem from both technological and non-technological factors shaping the enabling environment for these innovations. Overall, our findings thus position ‘scaling’ less as a linear process of dissemination (Rogers, Singhal and Quinlan, Reference Rogers, Singhal, Quinlan, Stacks and Salwen2014), and more as a system-building challenge, where technological performance must be matched by institutional capacity, affordability pathways, and socio-cultural fit for adoption to become durable (Wigboldus et al., Reference Wigboldus, Klerkx, Leeuwis, Schut, Muilerman and Jochemsen2016). This approach improves strategic alignment with enabling conditions and societal priorities, increasing the likelihood of inclusive, scalable, and durable outcomes (Geels, Reference Geels2004; Hermans et al., Reference Hermans, Sartas, Van Schagen, Van Asten and Schut2017; Thornton et al., Reference Thornton, Whitbread, Baedeker, Cairns, Claessens, Baethgen, Bunn, Friedmann, Giller, Herrero, Howden, Kilcline, Nangia, Ramirez-Villegas, Kumar, West and Keating2018).
Beyond its empirical findings, the study contributes methodologically by illustrating how participatory multi-criteria assessment can strengthen strategic decision-making for AR4D programs. Compared to conventional cost–benefit approaches, the framework enables the identification of trade-offs, distributional impacts, and institutional feasibility considerations that are critical for responsible and sustainable scaling (Ewell et al., Reference Ewell2025, submitted; Woltering et al., Reference Woltering, Valencia Leñero, Boa-Alvarado, Van Loon, Ubels and Leeuwis2024). However, this approach also instroduces methodological challenges, including the subjectivity of scoring, the need for facilitation capacity, and potential variability in stakeholder perceptions, highlighting the importance of transparent processes and complementary analytical tools. While the approach requires careful facilitation and transparency to manage subjectivity, it offers a replicable pathway for integrating diverse stakeholder perspectives into innovation prioritization and investment planning.
To ensure these findings reflect the practical realities, further on-the-ground technical feasibility studies and validation workshops with a broader, and more localized audience are essential. This will help to address critical issues within the enabling environment for scaling and determine how these issues can be effectively managed. By integrating this framework early in the scaling process and engaging a diverse group of stakeholders, we can effectively capture their perspectives on the technological and non-technological factors that shape the contextual enabling environment. This approach is especially crucial in contexts like Nandi County and the broader Global South, where innovations aimed at reducing emissions often receive limited support from stakeholders. Furthermore, these stakeholders, as intended users and practitioners, should be engaged throughout the innovation development, piloting and delivery process (Ewell et al., Reference Ewell2025, submitted). Future applications across different value chains and geographic contexts can further refine the framework and support the development of coordinated, inclusive, and climate-aligned scaling strategies to support both low-emission and development-oriented agri-food system transformation.
Acknowledgements
The authors would like to thank all participants in the two workshops for their invaluable contributions regarding the potential scaling of key low-emissions innovations relevant for Nandi County, Kenya. Funding was provided through the Scaling for Impact and Climate Action Science Programs of the CGIAR, and the authors are grateful for the support of CGIAR Trust Fund contributors: www.cgiar.org/funders. The authors acknowledge the editorial support provided Glenn Hyman of the Alliance of Bioversity International and CIAT Science Writing Service. Finally, the authors would like to thank Daniel Aja for providing the map of the study area.
Author contribution
Data analysis: G.A.A., H.E.; Data collection: G.A.A., M.V., J.M.S., L.C., K.S., N.T.; Formulation of research question: G.A.A., M.V., A.C.-N.; Interpretation of findings: G.A.A., H.E., M.V., J.M.S., L.C., K.S., N.T., A.C.-N.; Review, inputs, and final draft: G.A.A., H.E., M.V., J.M.S., L.C., K.S., N.T., A.C.-N.; Sourcing for funding: A.C.-N.; Study design: G.A.A., H.E., M.V., J.M.S., L.C., K.S., N.T., A.C.-N.; Writing of article (first draft): G.A.A., H.E.
Competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Open access
