LASER PULSE
Long-term Assistance and Services for Research (LASER)
Partners for University-Led Solutions Engine (PULSE)
The Impact of Blockchain Technology on Food Insecurity through 
African Indigenous Vegetables in Western Kenya
Final Report
Jessica Agnew, Ralph P. Hall, Joseph Mwangi, Daniel Sumner, and Nurvitria Kristofikova
SUPPLEMENT TO AGREEMENT NO. AID-7200AA18CA00009
AOR Name: Kevin Roberts
MAY 2022
This publication was produced for review by the United States Agency International Development (USAID). It was 
produced for the LASER PULSE Project, managed by Purdue University. The views expressed in this publication do 
not necessarily reflect the views of USAID or the United States Government.
2 | Page
Virginia Polytechnic Institute and State University 
Center for International Research, Education, and Development
526 Prices Fork Road
Blacksburg, VA 24061
Egerton University 
Nakuru-Mau Narok Road
Njoro, Kenya 
About the Project
Exploring the Use of Blockchain Technology to Improve Food Security Through African Indigenous Vegetables in 
Western Kenya is one of the first projects undertaken to investigate how blockchain technology (BCT) can be 
deployed in horticultural value chains with the express purpose of improving food and nutrition security among 
all value chain actors. The project also focuses on understanding how digital platforms using BCT will secure 
the place of women and youth in the value chain. This report documents the impacts of this twelve-month 
project. Additional information can be found on the project website. This project was made possible through 
a LASER PULSE grant, funded by USAID.
Project Partners
Virginia Polytechnic Institute and State University – Research Lead
Egerton University – Research Partner
AgUnity – Research Translation Partner
About LASER PULSE
LASER (Long-term Assistance and SErvices for Research) PULSE (Partners for University-Led Solutions Engine) 
is a five-year, $70M program funded through USAID’s Innovation, Technology, and Research Hub, that delivers 
research-driven solutions to field-sourced development challenges in USAID interest countries. A consortium 
led by Purdue University, with core partners Catholic Relief Services, Indiana University, Makerere University, 
and the University of Notre Dame, implements the LASER PULSE program through a growing network of 
2,500+ researchers and development practitioners in 61 countries.
About this Publication
This publication was made possible through support provided by the Center for Development Research, U.S. 
Global Development Lab (Lab), U.S. Agency for International Development, through the LASER PULSE Program 
under the terms of Cooperative Agreement No. 7200AA18CA00009. The opinions expressed herein are those 
of the author(s) and do not necessarily reflect the views of the U.S. Agency for International Development. This 
report was peer reviewed by the LASER PULSE Research & Publication Working Group. 
Recommended citation: Agnew, J., R.P. Hall, J. Mwangi, N. Kristofikova, and D. Sumner. 2022. The Impact of 
Blockchain Technology on Food Insecurity through African Indigenous Vegetables in Western Kenya. Virginia 
Tech, Blacksburg http://hdl.handle.net/10919/110444. 
This work is licensed under a Creative Commons Attribution-Non-Commercial 4.0 International License. 
3 | Page
Contents
Abbreviations.......................................................................................................................................................................................... 6
Executive Summary............................................................................................................................................................................... 7
Blockchain technology has the potential to transform the entire AIV value chain. .................................................. 9
Recommendations and Implications .......................................................................................................................................10
Introduction...........................................................................................................................................................................................13
Theory of Change................................................................................................................................................................................15
Study Design.........................................................................................................................................................................................19
Study Area ........................................................................................................................................................................................19
Sampling Approach...................................................................................................................................................................... 20
Embedded Research Translation Process..............................................................................................................................21
Impact Evaluation...........................................................................................................................................................................21
Impact Evaluation Outcomes ......................................................................................................................................................... 26
Hypothesis 1: The blockchain application improves the functionality of the value chain by improving 
transaction efficiency and the flow of information between actors. ........................................................................... 26
Hypothesis 2: Improved functionality of the value chain will increase the incomes of value chain actors, 
especially women and youth......................................................................................................................................................37
Hypothesis 3: Increased incomes of VC actors lead to improved food security at the household level. ...... 45
Hypothesis 4: In retail outlets, the blockchain application will improve the marketability of AIV varieties by 
addressing gendered factors that influence food purchase, preparation, and consumption............................ 48
Hypothesis 5: Increased marketability of AIVs will translate into increased purchases of AIVs by low-income 
households. ..................................................................................................................................................................................... 52
Hypothesis 6: Value chain actors will be willing to pay a small fee for the blockchain application for the 
economic and nutritional benefits derived from its use.................................................................................................. 54
Gender Implications ......................................................................................................................................................................57
Youth Engagement....................................................................................................................................................................... 60
Other Findings................................................................................................................................................................................ 63
Recommendations ..............................................................................................................................................................................67
Conclusions .......................................................................................................................................................................................... 69
References.............................................................................................................................................................................................. 71
Appendix 1 – AIV Grading Tool ......................................................................................................................................................73
4 | Page
Table of Figures
Figure 1. Digital technology adoption continuum...............................................................................................14
Figure 2. AIV value chain in Kakamega County ...................................................................................................15
Figure 3. Kakamega County........................................................................................................................................ 19
Figure 4. Study value chain arrangements ........................................................................................................... 20
Figure 5. AgUnity V3 SuperApp interface.............................................................................................................. 21
Figure 6. Farmers, traders, and retailers participate in an endline focus group....................................... 23
Figure 7. Transaction variability................................................................................................................................ 26
Figure 8. Improved cooperation between groups ............................................................................................. 29
Figure 9. Average trust scores by value chain actor.......................................................................................... 30
Figure 10. Information desired at the point of sale............................................................................................. 31
Figure 11. Grading of AIVs for sale........................................................................................................................... 32
Figure 12. Retailers differentiating grades of vegetables ................................................................................. 32
Figure 13. Producers and study personnel pictured with the market stall poster ................................... 33
Figure 14. Screenshot of AgUnity app wallet....................................................................................................... 33
Figure 15. Changes in the use of records for decision making...................................................................... 34
Figure 16. Proportion of retailers with difficulty accessing demanded AIV varieties .............................. 34
Figure 17. Flow of information from producer to consumer - baseline...................................................... 35
Figure 18. Flow of information from producer to consumer - endline........................................................ 35
Figure 19. Quick Mart in-store display ................................................................................................................... 36
Figure 20. Average income by quarter and activity............................................................................................37
Figure 21. Changes in average income by gender, April – June (Quarter 2)............................................. 38
Figure 22. Changes in average income by gender, July – September (Quarter 3) ................................. 38
Figure 23. Changes in average income among youth, April – June (Quarter 2)...................................... 39
Figure 24. Changes in average income among youth, July – September (Quarter 3)........................... 40
Figure 25. Average increase in production by variety ....................................................................................... 41
Figure 26. Average price per kilogram by variety, quarterly........................................................................... 43
Figure 27. Amount received in shillings and quantity sold in kgs, quarterly............................................. 43
Figure 28. Respondents to FIES items by gender............................................................................................... 45
Figure 29. Factors that could increase AIV purchase ........................................................................................ 48
Figure 30. Consumer preference for information about AIVs........................................................................ 49
Figure 31. Reasons for increased purchase of AIVs ........................................................................................... 52
Figure 32. Perceptions of using the AgUnity app .............................................................................................. 55
Figure 33. Change in gender roles of AIV production...................................................................................... 58
Figure 34. Participants with a youth from the community who assisted during training ..................... 60
Figure 35. Ideathon workbook image..................................................................................................................... 61
Figure 36. Egerton student participants in the hackathon.............................................................................. 62
Figure 38. Trader and producer negotiate prices for bundles of vegetables ........................................... 64
Figure 38. Price transmission along the value chain ......................................................................................... 64
Figure 39. Changes in financial literacy by gender............................................................................................ 66
5 | Page
Table of Boxes and Tables
Box 1. Overview of study findings............................................................................................................................ 25
Table 1. Questionnaire respondents by value chain role and gender......................................................... 22
Table 2. Indicator Estimation..................................................................................................................................... 24
Table 3. Reasons for reduced variability and uncertainty in transactions ...................................................27
Table 4. Participant preference for selling to AgUnity versus non-AgUnity users. ................................. 29
Table 5. Why has income increased from transacting in AIVs?...................................................................... 41
Table 6. Perceived reasons for improved prices................................................................................................. 42
Table 7. Average Food Insecurity Experience Scale (FIES) score................................................................... 45
Table 8. Average DDS by value chain actor and gender................................................................................. 46
Table 9. Foods consumed in the previous 24 hours ......................................................................................... 46
Table 10. Foods purchased......................................................................................................................................... 46
Table 11. Ranking of AIV quality ............................................................................................................................... 49
Table 12. Availability of preferred AIVs in the market....................................................................................... 50
Table 13. Access to AIV information ....................................................................................................................... 50
Table 14. Cultural and nutritional perceptions of AIVs ...................................................................................... 51
Table 15. Increase in AIV purchase over the previous 6-month period...................................................... 52
Table 16. Average weekly expenditure on AIVs (KES) ....................................................................................... 53
Table 17. Interest of participants in complementary services......................................................................... 54
Table 18. Changes in household financial decision making............................................................................ 59
6 | Page
Abbreviations
AIV African Indigenous Vegetables
BCT Blockchain Technology
CBO Community-based Organization
DAT Digital Agricultural Technologies
DDS Diet Diversity Score
FAO Food and Agriculture Organization
FIES Food Insecurity Experience Scale
GOK Government of Kenya
ICT Information Communication Technology
KES Kenyan Shilling
LASER PULSE Long-term Assistance and Services for Research (LASER) Partners for 
University-Led Solutions Engine (PULSE)
LMIC Low- and Middle-Income Country
MoALF Ministry of Agriculture, Livestock, and Fisheries
PVCA Participatory Value Chain Analysis 
SSA Sub-Saharan Africa
VC Value Chain
7 | Page
Executive Summary
This study is one of the first to explore how blockchain technology (BCT) could be used to improve 
food security in communities that are reliant on agriculture but are the last to receive services or 
access to markets, known as the ‘last-mile’. The goal was to determine how BCT could contribute to 
improving the income of African indigenous vegetable (AIV) value chain actors (e.g., producers, 
traders, and retailers) and to the affordability, availability, and accessibility of nutritious foods like 
AIVs for consumers. It finds that BCT can simultaneously strengthen the functionality of an entire
agri-food value chain by increasing the efficiency of transactions among value chain actors, 
improving cooperation along the value chain, and enhancing access to information. A decrease in 
post-harvest loss, reduction in negotiation and search costs, and traceability of Grade A vegetables 
were facilitated by the blockchain functionality of the AgUnity V3 SuperApp. Producer income was 
improved by better meeting market demand, time savings on AIV activities, increasing the supply of 
Grade A vegetables, and making information on the vegetables more available to consumers. 
Increased incomes led to improved food security among producers by facilitating their ability to 
procure more food, especially higher quality proteins and fruits. Participants and consumers reported 
an increase in the consumption of AIVs over the study period because of increased quality, 
availability, and awareness of their nutritional importance.
AIV production in Kenya and other East 
African countries declined several 
decades ago due to increased pressure 
on small-scale producers to grow cash 
crops such as maize and sugarcane and 
the popularity of ‘exotic’ vegetables (i.e., 
cabbages, spinach, tomatoes, and 
onions). Fortunately, in the last decade, 
there has been a resurgence in focus on 
AIVs among researchers, nutrition 
practitioners, and consumers given their 
cultural and nutritional significance and 
economic potential. AIVs’ superior 
nutritional profile also makes them ideal 
candidates for addressing malnutrition 
in Kenya. 
AIVs’ contribution to food and nutrition 
security will only be realized when their 
value chain functionality is improved. According to an initial value chain analysis and previous 
research, the main value chain inefficiencies that constrain the contributions that AIVs make to 
food and nutrition security are an insufficient flow of information through the value chain, 
1 - Blockchain technology can improve trust and 
transparency within agricultural value chains.
2 - Food security can be positively impacted through 
value chain strengthening by improving 
incomes, reducing waste, and increasing the 
availability and desirability of nutritious 
foods like African indigenous vegetables. 
3 - The BCT app significantly reduced the time 
producers spent trying to coordinate the 
sale of AIVs, which led to the diversification of 
income sources for producers, especially women. 
4 - Corruption and exploitation by dishonest 
brokers can be significantly reduced by creating a 
network of trusted actors and providing access to 
reliable market information.
Key Takeaways 
8 | Page
mismatched markets, poor vertical coordination between actors, and a lack of trust and 
transparency. 
This study investigates how BCT can be deployed in AIV value chains with the express purpose of 
improving food and nutrition security among all AIV value chain actors, including consumers. The 
study also focuses on understanding how digital platforms using BCT will secure the place of women 
and youth in the agricultural value chain as upgrading activities usually attract men to participate. 
The study was undertaken through a 
collaboration between Virginia Tech, 
Egerton University, and the tech start-up 
AgUnity. An embedded research 
translation model was used to adapt 
AgUnity’s blockchain-based smartphone 
application (app) to the AIV value chain in 
Kakamega county. A non-experimental 
longitudinal study design was used to 
conduct an impact evaluation based on six 
core hypotheses.
The AgUnity BCT-based V3 SuperApp was 
selected as the test case for this study since 
it uses the distributed ledger functionality 
of blockchain to establish a secure record￾keeping system for all transactions that 
occur along the value chain. Each time a 
buyer and seller agree on the cash price for 
a certain amount of a vegetable, they both 
capture and approve this transaction in 
their app, which creates a block of 
information that is added to the blockchain. This information stored on the blockchain is immutable, 
giving users confidence that their records of transactions are accurate and reliable. It is also able to 
create traceability of information along the value chain, communicating characteristics of the 
vegetables from the producer through to the consumer.
Over six months, 56 smartphones containing the AgUnity BCT-based app were distributed to 
selected producers, traders, and retailers to explore how the platform could improve food security
and address the challenges faced by last-mile agricultural communities. The study participants were 
selected using purposive and snowball sampling methods to ensure actors across each node of four
different value chain configurations were included (depicted below). Each participant was provided 
with a free smartphone as a way of compensating them for time spent in trainings, working with the 
research team to co-design the app, and for participating in baseline, midline, and endline 
evaluations. A subset of participants also engaged in focus group discussions during the endline 
HYPOTHESIS 1 – The blockchain application improves 
functionality of the value chain by improving transaction 
efficiency and the flow of information between actors.
HYPOTHESIS 2 – Improved functionality of the value 
chain will increase incomes of value chain actors, 
especially women and youth.
HYPOTHESIS 3 – Increased incomes of VC actors lead 
to improved food security at the household level.
HYPOTHESIS 4 – In retail outlets, the blockchain 
application will improve the marketability of AIV varieties 
by addressing gendered factors that influence food 
purchase, preparation, and consumption.
HYPOTHESIS 5 – Increased marketability will translate 
into increased purchases by low-income households for 
AIVs.
HYPOTHESIS 6 – Value chain actors will be willing to 
pay a small fee for the blockchain application for the 
economic and nutritional benefits derived from its use.
Hypotheses 
9 | Page
evaluation in each study location. Finally, baseline, midline, and endline questionnaires were also 
conducted with consumers in the markets where participating retailers sold AIVs. 
Blockchain technology has the potential to transform the entire AIV value chain. 
The blockchain functionality of the AgUnity app was found to improve the functionality of the entire
value chain over the six-month period of evaluation (Hypothesis 1). An important factor behind the 
improved value chain functionality was the strong connectivity between producers and traders, which 
significantly reduced the time spent trying to coordinate the sale of AIVs. Since transactions in the 
AgUnity BCT-based app can only occur between actors registered in the network, coordination and 
cooperation were easier since each user had access to a known group of individuals they can transact 
with. Some 92% of producers and 87% of traders and retailers reported that using the AgUnity app 
improved their relationships with the people they transact with. There was also a strong preference 
for transacting with those in the AgUnity network versus those who were not, especially among 
traders. AIV producers used their time savings to engage in additional productive or social activities. 
The transaction data captured by the AgUnity app also provided producers with invaluable 
information on the revenues obtained from different vegetables. Prior to participating in this project, 
only 30% of participants kept paper-based records and none used mobile technology for record￾keeping. These records assisted value chain actors in better matching supply to demand and reduced 
the difficulty for retailers in finding demanded AIV varieties. Overall, transaction costs in the 
AIV value chain were significantly reduced by the BCT-based AgUnity V3 SuperApp.
10 | Page
Improvements in value chain functionality due to the BCT app led to an improvement of producer 
incomes, especially for women, from on-farm activities (Hypothesis 2); however, traders’ income 
decreased across all sources of income, while retailers’ income increased from on-farm activities and 
employment. The increased value chain functionality allowed producers to diversify their 
income, increasing their off-farm and employment-based income over the 6-month 
study period.
There was clear evidence that food security did increase over the 6-month study period (Hypothesis 
3). Households demonstrated an ability to purchase higher quality proteins and fruits when 
compared to the baseline. In addition to increased income, the focus of the project on the nutritional 
quality of AIVs and how to use these characteristics to market products to consumers also helped to 
increase the demand of participants for the leafy greens.
Vegetable quality and information about how the vegetables were produced were found to be 
important factors shaping purchasing decisions (Hypothesis 4). BCT created an incentive for
producers to increase the marketability of AIVs. By the endline study, 100% of traders and the 
majority of producers were using a formal grading process introduced as part of this study to trace 
Grade A vegetables from farm to point of sale. Consumers perceived a significant increase in 
the quality and availability of AIVs from the baseline to endline, but still had limited access 
to information on the inputs used in the production of AIVs, production practices, water quality, and
how the vegetables were transported. 
Over the 6-month study, retailers reported an improvement in the marketability of AIVs, including a 
willingness among consumers to pay a price premium – estimated to be as much as a 10 to 30 KES
– for Grade A vegetables (Hypothesis 5). There was also a self-reported increase in the volume of 
vegetables study participants purchased. For all participants, better quality was the leading factor 
driving the increase, followed by better prices and more awareness of the nutritional importance of 
AIVs. All value chain actors, including consumers, self-reported an increase in 
consumption of AIVs, due to increased quality, availability, and awareness of the 
nutritional value.
Since the study participants were provided a smartphone and were not asked to pay a subscription 
fee to use the AgUnity app, the study explored potential pricing models for the BCT service. Finally, 
approximately 75% of producers, traders, and retailers indicated they were willing to pay 150 KES per 
month to continue to use the AgUnity app (Hypothesis 6); however, to be willing to pay this price, 
the app would also need to offer complementary services and features. In addition to 
participants’ willingness to pay, members of their communities were also willing to pay 
both $75 and the monthly service fee of 150 KES in order to join the AgUnity network.
Recommendations and Implications
This impact evaluation identifies the potential value that BCT can offer last-mile agricultural systems 
to improve value chain functionality and livelihoods. BCT cannot offer solutions for all the challenges 
faced by AIV value chain actors; however, there are clear linkages between the technology and food 
11 | Page
security outcomes. The following recommendations outline several areas where technology 
companies (i.e., AgUnity) and enthusiasts, policymakers, and practitioners can facilitate the research 
and development needed to fully capture the transformative potential of BCT.
Recommendation 1. Developing low-tech traceability solutions should be a core priority 
for increasing demand for nutritious foods and creating economic returns for the 
producers of those foods.
The demand for high-quality, organic AIVs is growing in Kenya. By integrating traceability into AIV 
value chains, producers using traceability solutions will earn a first-mover advantage by meeting 
unmet consumer desires for information. This will lead to improvements in income, food security, 
and livelihood resilience. 
Recommendation 2. Developing assistive technologies and monitoring mechanisms 
that minimize human interference will allow the full potential of BCT to be realized in 
agri-food systems.
There is a need to develop monitoring mechanisms to provide consumers with external assurance 
of the quality of vegetables and if they were produced organically. There is also a need to develop 
assistive technologies, such as internet of things sensors or weigh scales, that will minimize human 
interference in the metadata of the vegetables. If these solutions can be identified, national 
stakeholders would be able to leverage BCT to scale up compliance to standards across sectors. This 
in turn would develop consumer confidence in locally-produced horticulture products and facilitate 
access to export markets.
Recommendation 3. CBOs and other producer associations should be engaged to 
achieve network scalability and acceleration of transaction cost reduction.
The CBO New Vision was imperative to the success of this project. CBOs are adept at community 
transformation, encouraging behavior change, and harmonizing the community response. Working 
with CBOs, cooperatives, and farmer/agribusiness associations going forward will help to achieve 
scale in blockchain-based transacting networks. This will also lead to accelerated reductions in 
transaction costs in the agri-food system, leading to both increased income and a lower cost of food, 
both of which will contribute to improved national food security.
Recommendation 4. BCT should be deployed alongside complementary digital services.
The initial value chain analysis and outcomes observed in the impact evaluation reveal that BCT alone 
cannot address the pain points constraining the functionality of AIV value chains. Producers, traders, 
and retailers need access to advisory and extension services, financial services, weather information, 
and markets for their produce that will continue to minimize transaction costs and ensure they are 
fairly compensated for their respective activities in the value chain. One of the main findings of this 
12 | Page
project is that it is not BCT, but BCT Plus Services that has the most potential to alleviate the pain 
points of last-mile agricultural systems.
Recommendation 5. Digital solutions should not be a replacement for in-person 
technical training and capacity building.
Digital solutions are not a substitute for in-person interaction and engagement to build technical 
skills in producing or retailing agricultural products will continue to be invaluable. While some of the 
training that is needed to address production and retailing challenges of AIVs could be offered 
digitally, in-person engagement to problem solve and examine the existing practices used will be 
essential for improving productivity sustainably.
 
13 | Page
Introduction
In sub-Saharan Africa (SSA), 70% of livelihoods depend on agriculture, especially smallholder crop 
production, animal husbandry, and other small agri-businesses. These agricultural workers, roughly 
half of whom are women, produce 60 to 80% of food in the continent. According to Mellor (2017), a 
1% increase in agricultural GDP results in a 2.6% reduction in poverty, compared to a 0.25% reduction
in poverty with a 1% increase in overall GDP. The impact of smallholder productivity on poverty in 
particular places agriculture as a key priority for improving livelihoods, resilience, and food and 
nutrition security (Arndt et al. 2010). 
Kenya has demonstrated leadership in the SSA context for driving agriculture sector growth and 
improving food security through policy frameworks (i.e., including food security in the country’s ‘big 
four’ priorities), demonstrating openness to innovation, and creating favorable investment 
environments. However, one-third of the population remains under the international poverty line, 
2.1 million are facing acute food insecurity (up from 1.3 million in 2020), and women and other socially 
excluded populations (i.e., the elderly, persons with disabilities) have been disproportionately 
excluded from accessing the increased wealth in the country (USAID 2020; IPC 2021; WFP). 
Furthermore, the COVID-19 pandemic has exacerbated poverty, food insecurity, and malnutrition. 
Continued innovation is required to advance agriculture sector transformation to achieve 100% food 
and nutrition security in the country (FAO et al. 2020; GoK 2017). 
With deepening access to mobile phones, the internet, and other information and communication 
technologies (ICTs) in SSA over the last two decades, the private sector, technologists, agriculturalists, 
development agencies, government, and researchers, have turned towards digital agricultural 
technologies (DATs) as potential remedies for the multitude of challenges afflicting the food and 
fiber farming sectors. There is a growing consensus that DATs can contribute to widespread 
agricultural sector transformation, productivity growth, and improved access to services, factors, and 
output markets (De Ruyter de Wildt et al. 2019). 
Blockchain technology (BCT) has emerged as a prominent DAT that offers compelling value chain
development solutions for difficult challenges such as high transaction costs, the opacity of value 
chains, lack of trust between transacting parties, and lack of consumer confidence in final products
(FAO and Zhejiang University 2021). BCTs’ distributed immutable ledger functionality has the 
potential to more accurately value food, address market failures, capture and secure data along the 
chain, create an ecosystem for data-sharing, reduce value chain corruption, and improve trust, 
traceability, transparency, and efficiency (De Ruyter de Wildt et al. 2019; Tripoli and Schmidhuber 
2020). While the number of private sector companies that offer BCT-based traceability in food supply 
chains is increasing, particularly in a high-income country context, there is little evidence to 
substantiate the benefits that BCT can generate for last-mile farmers1 in a low- and middle-income 
country context. There are also no studies that investigate if BCT can create returns to food and 
 
1 Last-mile farmers are farmers who typically are last to receive services or access to markets.
14 | Page
nutrition security by improving the functionality of value chains for nutritious foods and improving 
value chain actor income.
According to a World Bank report 
on DATs in Africa, Kenya is ranked 
as the leading regional innovation 
hub with the largest number of 
scalable DATs, followed by South 
Africa and Nigeria. As internet and 
mobile phone access grows, users 
move along a continuum of 
technology adoption (Figure 1) 
(Kim et al. 2020). In the low internet 
penetration and low mobile access 
environments, SMS-based 
applications and mobile payment 
services dominate. However, as 
internet connectivity and mobile 
phone access reaches a critical mass, advanced big data analytics and farm robotics applications 
have increased in usage. Thus, Kenya is at an appropriate stage of the digital technology adoption 
continuum to test the potential for an advanced DAT like blockchain.
Virginia Tech (USA) and Egerton University (Kenya) partnered with the Australian tech start-up, 
AgUnity, to conduct the study Exploring the Use of Blockchain Technology to Improve Food Security 
Through African Indigenous Vegetables (AIVs) in Western Kenya. AIVs were selected as the focus of 
this study because they are an important food security crop in Kenya given their short growing 
season, ownership of production and marketing by women, and their nutritional value. However, the 
full economic and nutrition potential of AIVs has not yet been realized due to value chain 
inefficiencies and constraints. The research started with a formative phase that included an in-depth 
gender-sensitive value chain analysis of AIVs to inform adaptations of AgUnity’s blockchain-based 
application to the Western Kenyan AIV context (Agnew et al. 2021). The second phase consisted of 
introducing AgUnity’s blockchain-based application into the AIV value chain to assess if food security 
can be improved by increasing incomes and the availability and desirability of AIVs by improving the 
functionality of the AIV value chain. 
The findings of this study will assist policymakers, development practitioners, agricultural sector 
stakeholders, and technology enthusiasts and companies alike to understand the complexities of 
introducing an emerging technology like BCT into last-mile agricultural value chains, the impacts and 
unintended consequences of introducing the technology, and the extent of evidence still needed to
leverage digital solutions to transform agriculture in Kenya and other SSA countries.
Figure 1. Digital technology adoption continuum
Source: Kim et al. 2020
15 | Page
Theory of Change
The evidence of blockchain technology’s (BCT) contribution to strengthening agri-food value chains
has been steadily mounting since 2020 (Liu et al. 2021). Information sharing, data integrity, value 
chain efficiency, traceability, improved trust between actors, and data-driven decision-making have 
been a few of the investigated impacts to date. However, research has yet to investigate if BCT can 
be leveraged to improve food security and nutritional outcomes by improving the functionality of 
agri-food value chains that produce nutritious foods. 
AIVs were selected as the focus crop for this study since they support food security by generating 
income and providing important micronutrients to the diet. However, the preliminary AIV value chain 
analysis conducted for this study and other research in western Kenya demonstrates that the food 
security benefits of the leafy greens have not yet been fully captured due to value chain inefficiencies. 
While demand for AIVs is growing in Kenya, there is an insufficient flow of information through the 
value chain, markets are mismatched, there is poor vertical coordination between actors, and value 
chain actors’ relationships are characterized by a lack of trust and transparency (Gido, Ayuya, Owuor, 
and Bokelmann, 2016; Kansiime et al. 2018). This results in inconsistent supply and quality of AIVs in 
markets throughout Kenya. The characteristics of BCT are well suited to address these value chain 
inefficiencies (see Agnew et al. 2021). Figure 2 outlines the types of market arrangements that exist 
in AIV value chains in Western Kenya.
 
Figure 2. AIV value chain in Kakamega County
16 | Page
DATs like BCT can contribute to improvements in food security both by increasing income, allowing 
households to purchase more expensive but often nutrient-rich foods (e.g., animal source foods), 
and by increasing the availability and desirability of nutritious foods, through strengthening 
inefficient value chains of nutritious foods. However, this requires that digitization addresses the key 
pain points that value chain actors face in carrying out their respective activities and transactions and 
that they are able to sustain use of the technology and services over the long term. For value chain 
actors to use BCT in their transactions and continue to do so over the long-term, there are four key 
antecedents: 
1 - Value chain actors (i.e., producer, trader, retailer) who transact with one another are willing 
to participate in the system that is intended to be more open and transparent. 
2 - Value chain actors can own and retain a smartphone powerful enough to host a 
blockchain-based application.
3 - Value chain actors can use the smartphone and are able to understand how to execute 
the double handshake scanning process.
4 - Users have access to technical support and training.
Underpinned by these assumptions and based on existing evidence that demonstrates BCT’s ability 
to strengthen value chain inefficiencies, this study investigated the following theory of change and 
hypotheses.
IF the AgUnity blockchain-based smartphone application can improve the functionality of 
the AIV value chain in western Kenya, and IF improved functionality contributes to an 
improvement in the income of value chain actors AND an increase in the availability and 
desirability of AIVs in local markets, THEN food security will be improved among value 
chain actors, including end consumers.
Hypothesis 1 - The blockchain application improves the functionality of the value chain 
by improving the efficiency of transactions and the flow of information between actors.
Value chain functionality is the state where transaction costs (i.e., financial, labor, search and 
information, and bargaining costs) are minimized, value chain actors (i.e., producer, trader, and 
retailer) cooperate to deliver affordable, safe, and high-quality products to the final market, value 
chain actors are fairly compensated for their respective activities, and supply meets demand.
It was expected that the AgUnity BCT-based smartphone application would reduce transaction costs 
by improving transaction efficiency, which is defined as the minimization of time, effort, and money 
involved in the completion of a transaction between two parties (Transaction Commons 2019). 
Improved transaction efficiency was anticipated because BCT can improve the reliability, accuracy, 
and timeliness of transactions by locking in agreed-upon quantities and purchase prices, creating 
17 | Page
verified and secured transaction records, improving coordination and cooperation between actors, 
and enhancing trust and transparency.
Improved flow of information through the AIV value chain would allow value chain actors to 
cooperate to deliver affordable, safe, and high-quality AIVs to the market and for each actor to be 
fairly compensated for their respective activities (Agnew et al. 2021). Increased transparency 
regarding the vegetables was expected to incentivize improved production and quality. Creating a 
more cooperative value chain and improving access to information about the consumers’ demands
was expected to enable producers to consistently meet demand in terms of quantity, variety, and 
quality of the AIVs. Providing access to transaction history was expected to increase the negotiating 
power of producers and reduce the opportunistic behavior of buyers (both traders and retailers). 
Hypothesis 2 - Improved functionality of the value chain will increase the incomes of 
value chain actors, especially women and youth.
According to the initial value chain analysis, income from AIVs, especially for producers, was 
constrained due to the limited functionality of the value chain. A consistent mismatch with supply 
and demand, lack of insight into characteristics demanded by the consumer, lack of reliable 
transactions, and post-harvest loss due to lack of buyers, were contributors to not fully realizing the 
economic potential of growing and selling AIVs. By reducing transaction costs and improving the 
flow of information, it was expected AIV supply would better match demand (which is ample 
throughout the country), and the incomes of producers (the majority of whom are women) would 
increase.
Hypothesis 3 - Increased incomes of value chain actors and increased availability and 
desirability of AIVs, lead to improved food security among value chain actors.
An increase in income and an increase in the availability and desirability of nutritious foods are known 
drivers of increased food security (Henson and Humphrey, 2015). While the main outcome of interest 
was an increase in producer income, since they currently receive the lowest share of the final price 
paid by the consumer (Agnew et al. 2021), it was anticipated that strengthening the value chain 
functionality would lead to an increase in income for traders and retailers. Since all actors in the value 
chain have experienced food insecurity (Agnew et al. 2021), it is important to emphasize the role of 
improved value chain functionality in the livelihoods of everyone participating in the production and 
sale of AIVs.
Hypothesis 4 - In retail outlets, the blockchain application will improve the 
marketability of AIV varieties by addressing gendered factors that influence food 
purchase, preparation, and consumption.
Improved value chain functionality was expected to improve the marketability of AIVs to the end 
consumer by increasing the flow of information between value chain actors of the demanded AIV 
characteristics (i.e., grown without pesticides), varieties, and quantity. It was also expected to improve 
marketability by providing traceability of these characteristics. While eventually monitoring 
mechanisms and a consumer-facing application would be needed to fully capture the benefits of 
18 | Page
traceability offered by BCT, it was expected that an initial introduction of the traceability concept 
would capture consumer attention in the market where value chains are almost completely opaque. 
Hypothesis 5 - Increased marketability (i.e., affordability, types of varieties available, 
accessibility, knowledge of preparation of vegetables) will translate into increased 
purchases by low-income households for AIVs.
Increased desirability of nutritious foods is known to contribute to improved consumption (Henson
and Humphrey, 2015). It was expected that improving the functionality of AIVs in Kakamega County 
would also contribute to improvements in food security in consumer households by increasing the 
consumption of the nutritious leafy greens.
Hypothesis 6 - Value chain actors will be willing to pay a small fee for the blockchain 
application for the economic and nutritional benefits derived from its use. 
It was hypothesized that the records generated by AgUnity’s blockchain-based smartphone 
application would generate enough value among users that they would be willing to pay a small 
subscription fee to continue using the app past the life of the project. 
19 | Page
Study Design 
This study used a non-experimental impact evaluation approach to investigate if blockchain 
technology could improve food security through higher incomes and increased consumption of 
vegetables by strengthening African indigenous vegetable (AIV) value chains in western Kenya. 2 This 
study also employed the LASER PULSE embedded research translation (ERT) model, which is “an 
iterative co-design process among academics, practitioners, and other stakeholders in which 
research is intentionally applied to a development challenge” (LASER PULSE). 
In three sub-counties and one town in Kakamega County, 56 value chain actors (i.e., producers, 
traders, and retailers) participated in this study. Each received a phone that was powerful enough to 
host the AgUnity blockchain-based application, known as the V3 SuperApp. Participants first received
their phone in May 2021 and the final evaluation of the project was conducted in November 2021. 
AgUnity staff assisted by faculty from Egerton University, verified participants’ identities, conducted 
the onboarding process for the application,3 trained them on the basics of smartphone usage, and 
trained them how to use the app. To use the blockchain functionality, a double handshake takes 
place between buyers and sellers by scanning a QR code.4 AgUnity hired a local field officer who 
conducted repeated training on the scanning process. 
Study Area
Kakamega County is located in western Kenya, near the 
border of Uganda (Figure 3). In 2019, the population was 
approximately 1.8 million (GoK 2019). As of 2016, the 
absolute poverty rate was approximately 36%, compared 
to the national average of 37%, and the rate of food 
poverty was approximately 33%, compared to the national 
average of 32% (GoK 2017). Kakamega is one of four 
counties that has the highest rate of multi-dimensional 
and monetary poverty in absolute numbers in the country 
(KNBS 2020). 
Kakamega County is one of the country’s major AIV 
producers as horticultural production can take place most 
of the year (Laibuni, Losenge, and Bokelmann 2020). Most 
vegetable production is carried out by female small-scale 
producers (MoALF 2017). Accordingly, irrigation and horticultural production are core poverty 
 
2 The research protocol and all data collection instruments were approved by the Biomedical Research Alliance 
of New York (Protocol No. 20-059-568) and the Egerton University Research Ethics Committee (Approval No. 
EUREC/APP/113/2021).
3 AgUnity. 2022. Trainer Guide (Bali, Indonesia).
4 See this video for a demonstration of transactions on the AgUnity V3 SuperApp.
Source: https://kenya.opendataforafrica.org/
Figure 3. Kakamega County
20 | Page
reduction strategies in the region (County Government of Kakamega, 2018). Other crops produced
in Kakamega include maize, sugarcane, bean, cassava, finger millet, and sorghum (MoALF 2017). 
This study was carried out in the sub-counties of Butere, Mumias, Ekero/Shianda, and the town
Kakamega. These study locations were selected as they each produce and sell significant volumes of 
AIVs. They also represent different vertical coordination arrangements between producers, traders, 
and retailers. Four markets were selected in each of the study locations where participating retailers 
were located. These markets were selected because they primarily serve consumers in the third- and 
fourth-income quintiles. Since preparations for the elections were starting, there was sensitivity 
around collecting income data in public locations. Therefore, market selection was used to identify 
‘low-income’ consumers rather than asking participants about their income. 
Sampling Approach
The number of participants was determined by the project budget, which covered the cost of the 
smartphones. This approach was used as a way of compensating the participants for time spent in 
training, working with the research team to co-design the app, and for participating in baseline, 
midline, and endline evaluations.
Four types of value chain arrangements were found in the study areas (Figure 4), illustrating the 
informal and unstructured nature of AIV transactions. These sub-chains were used to guide the 
selection of participants, starting from retailers and tracing through the value chain to identify the 
producers and other actors that supply them. Purposive and snowball sampling methods were used 
to identify the participants for this study. A pre-recruitment checklist was conducted to ensure that 
they were qualified to participate (i.e., entering of their own free will, over the age of 18, and farmed 
or traded AIVs) and to meet our target participation rate of women, men, and youth (at least 50% 
men and women and at least 30% youth). 
Figure 4. Study value chain arrangements
21 | Page
Embedded Research Translation Process
Research partners Virginia Tech and Egerton University
selected the Australian tech startup, AgUnity, as the research 
translation partner. AgUnity’s V3 SuperApp is an out-of-the￾box smartphone application built on BCT that can be 
adapted to various last-mile agricultural contexts and value 
chains (Figure 5). Since its inception, AgUnity has focused 
primarily on working in value chains for cash crops that tend 
to be more formalized or straightforward (i.e., coffee, 
cocoa). The ERT model provided a valuable process for
adapting the app to a more informal value chain.
The iterative ERT co-design process conducted with project 
participants led to the adaptation of AgUnity’s BCT 
smartphone application to the context of AIV value chains.
The app functionality and the business processes in which 
the app is used (i.e., transactions) were designed to address 
the pain points of AIV value chain actors including a lack of 
coordination in the value chain, a lack of assurance of 
vegetable safety, the poor transmission of information, a 
lack of standardized grading and pricing, the weak market power of women, and the need for 
technical assistance (Agnew et al. 2021). The app’s user interface is designed for low literacy and lack 
of familiarity with using smartphone applications. It was also translated into Swahili for this project. 
Over the life of the project, the ERT process involved repeated interaction with the project 
participants to create opportunities for them to provide feedback and present challenges or insights 
that could be reflected in the app. AgUnity hired a field officer to ensure that this continuous 
engagement was possible. Virginia Tech and Egerton University also conducted meetings with 
participants 3 months after deployment to obtain additional feedback. Details on the app's
functionality and implementation can be found in the deployment report (Kristofikova, Muskoke, and 
Agnew, 2021).
Impact Evaluation
Value chain participants were followed longitudinally between the initial deployment of the phones 
(May 2021) and the end of the project period (November 2021). The impact evaluation included: (1) 
a baseline and endline structured questionnaire; and (2) voluntary participation in an endline focus 
group. A total of 4 focus groups were conducted – 1 in each of the sub-county locations. A baseline, 
midline, and endline questionnaire were also conducted with consumers in the markets where 
Figure 5. AgUnity V3 SuperApp interface
22 | Page
participating retailers were located to identify changes in market availability of AIVs, consumer 
demand trends for AIVs, awareness of vegetables sold using blockchain, etc.
Baseline and Endline Questionnaire
Baseline and endline questionnaires were conducted by students from Egerton University.5 The 
baseline questionnaire took approximately 1.5 hours to complete and preceded the AgUnity phone 
onboarding process. The endline survey took approximately 1 hour to complete. At both baseline 
and endline, surveys were collected at a common meeting point. Participants received compensation 
for their transportation costs to travel to the designated location. Responses were recorded on a 
tablet using the Qualtrics Offline App. Responses were checked daily for accuracy. 
The questionnaires covered topic areas such as (anticipated) effects of technology adoption, activities 
related to AIV production or sale, activities related to the participant’s respective role in the value 
chain, food security, gendered dimensions of value chain participation and household dynamics, 
financial literacy, and quarterly income-based on income-generating activity.
Some participants missed the original onboarding day. There were also challenges in getting 
enumerators and trainers back to the field because of COVID-19 restrictions. As such, there were 3 
participants at both baseline and endline that were not surveyed. Three participants were not 
surveyed at the endline because they were out of town. Three producers withdrew from the study 
but were replaced by a family or community member of their choosing. Table 1 breaks down the 
number of participants surveyed at baseline and endline by female and male participants.
Table 1. Questionnaire respondents by value chain role and gender
Focus Groups
Focus groups were jointly held by AgUnity’s Program Director, Nurvitria Kristofikova, and Virginia 
Tech’s co-PI, Jessica Agnew (Figure 6). The purpose of the focus groups was to understand the overall 
experience with the phones, the participants' perception of its continued use, ideas to build on and 
scale, and the challenges and benefits of using the phone in their respective value chain activities. 
The purpose of the focus groups was to hear the perspectives of the participants in their own words, 
rather than to elicit the impacts through the baseline and endline surveys alone. These focus groups 
later helped to further understand the observed changes (or lack thereof) between baseline and 
endline questionnaires.
 5 Questionnaires can be shared upon request.
23 | Page
Focus groups were recorded with the permission of the participants. Flip charts and stickers were 
also used to record participant responses and the guides used were designed for low literacy. 
Extensive notes were also taken by the facilitator. The focus groups were held in Swahili and 
translated by the AgUnity field officer. 
Figure 6. Farmers, traders, and retailers participate in an endline focus group
Consumer Baseline, Midline, and Endline
Structured questionnaires were conducted at baseline, midline, and endline with consumers in five 
participating markets in each of the sub-counties of Butere, Shianda/Ekero, Mumias, and Kakamega. 
The questionnaires were conducted by Egerton students and took place in the markets. The baseline 
and endline survey took approximately 15 to 20 minutes to complete and included questions on 
consumer perceptions of AIVs, the varieties they purchase, their availability and quality, nutritional 
awareness, and concerns about the AIVs. A total of 50 and 62 consumers were surveyed in the 
baseline and endline, respectively. 
The midline questionnaire took approximately 5 minutes to complete and did a deeper dive on 
consumer willingness to pay for various characteristics of AIVs, such as being grown organically. The 
midline survey was used to align activities between value chain actors and consumers, i.e., making 
Grade A vegetables more widely available and having information on the vegetables provided at the 
retail outlet. A total of 438 consumers were surveyed at the midline.
24 | Page
Data Analysis 
Quantitative, descriptive data were analyzed using Tableau. The Exact McNemar test for paired 
binomial data was used to test the statistical difference in proportions pre-and post-intervention, 
meaning, that this change would likely be observed in another simple random sample. The exact p￾value binomial test was used given the small sample size. A paired t-test was used to validate the 
significance of mean changes observed between baseline and endline. Table 2 describes the 
statistical methods of estimation for the indicators used in this report. 
Table 2. Indicator Estimation
Indicator Method of Estimation Source 
Food Insecurity 
Experience Scale
(FIES)
There are three methods for estimating FIES in a 
population. First, a respondent’s raw score aggregates the 
affirmative responses to the eight FIES questions. The raw 
score is ordinal, with higher scores representing more 
severe experiences of food insecurity. Second, given 
respondents' raw scores, the probability that a respondent 
experiences certain thresholds of food insecurity is 
calculated. Third, the prevalence of food insecurity in a 
population is estimated based on the weighted sum of 
the raw-score probabilities. This report uses the raw score 
to measure changes in participants’ food insecurity 
experience. Publications for peer review will assess the 
FIES probability.
FAO FIES
Diet Diversity 
Score
Household diet diversity was calculated based on a 24-hr 
recall. The number of 12 food groups consumed in the 
recall period is aggregated to create the score. The HDDS 
is a proxy for the household’s ability to access food and 
socio-economic status.
Swindale & 
Bilinsky, 2004
Qualitative data were analyzed by coding transcripts and notes conducted from the focus groups. 
Recording of codes and alignment with study hypotheses was conducted using Excel. The list of 
coded themes was expanded during analysis until themes reached saturation. 
Qualitative and quantitative data were used to triangulate the findings in the impact evaluation and 
provide rich insight into the observed outcomes. Consumer data further enriched the findings in 
addition to understanding changes in their behavior.
25 | Page
Box 1. Overview of study findings
IMPACT EVALUATION FINDINGS SNAPSHOT
HYPOTHESIS 1 – The blockchain application improves functionality of the value chain by 
improving transaction efficiency and the flow of information between actors.
The nature of BCT creates an approach to strengthening an entire value chain simultaneously as the app requires all 
nodes of actors to be engaged in transacting, compared to other apps that are used by single nodes alone. 
Transaction costs, cooperation, and access to information improved over the six-month period of evaluation. 
HYPOTHESIS 2 – Improved functionality of the value chain will increase incomes of value 
chain actors, especially women and youth.
Improvements in value chain functionality due to the BCT app led to an improvement of producer incomes, 
especially for women, from on-farm activities; however, traders’ income decreased across all sources of income, 
while retailers’ income increased from on-farm activities and employment.
HYPOTHESIS 3 – Increased incomes of value chain actors and increased availability and 
desirability of AIVs, lead to improved food security among value chain actors.
There was clear evidence that food security did increase over the 6-month study period. Households demonstrated 
an ability to purchase higher quality proteins and fruits when compared with the baseline. In addition to increased 
income, the focus of the project on the nutritional quality of AIVs and how to use these characteristics to market the 
product to consumers also helped to increase the demand of participants for the leafy greens.
HYPOTHESIS 4 – In retail outlets, the blockchain application will improve the marketability 
of AIV varieties.
This study revealed that vegetable quality and information about how the vegetables were produced were important 
factors shaping purchasing decisions. BCT created an incentive for producers to increase the marketability of AIVs.
Consumers perceived a significant increase in the quality and availability of AIVs from the baseline to endline, but still 
had limited access to information on the inputs used in the production of AIVs, production practices, water quality, 
and the transportation methods of the vegetables.
HYPOTHESIS 5 – In retail outlets, the blockchain application will improve the marketability 
of AIV varieties by addressing gendered factors that influence food purchase, preparation, 
and consumption. 
Over the 6-month study, retailers reported an improvement in the marketability of AIVs, including a willingness 
among consumers to pay a price premium – estimated to be as much as a 10 to 30 KES – for Grade A vegetables. 
There was also a self-reported increase in the volume of vegetables study participants purchased. For all participants, 
better quality was the leading factor driving the increase, followed by better prices and more awareness of the 
nutritional importance of AIVs.
HYPOTHESIS 6 – Value chain actors will be willing to pay a small fee for the blockchain 
application for the economic and nutritional benefits derived from its use.
Approximately 75% of producer, traders, and retailers indicated they were willing to pay 150 KES per month to 
continue to use the AgUnity app; however, to be willing to pay this price, the app would also need to offer 
complementary services and features, such as providing producers with access to micro or crop insurance, 
information on integrated pest control techniques, trainings, access to new markets for AIVs, etc. 
26 | Page
Impact Evaluation Outcomes
Hypothesis 1: The blockchain application improves the functionality of the value chain by 
improving transaction efficiency and the flow of information between actors. 
Improvements in the functionality of the AIV value chain were observed between baseline and 
endline, with qualitative insights provided by the focus group discussions. Transaction efficiency 
increased as price and quantity variability was reduced, and coordination and cooperation between 
value chain actors increased as a result of enhanced trust between transacting parties. Flows of 
information improved in terms of transmitting the grade of the vegetable from producer through to 
end consumer; however, no other significant improvements in information transmission were 
observed.
Improved Transaction Efficiency
Price variability reduced 
For transactions to be efficient, they must be reliable; meaning there is minimal variability and 
uncertainty in the agreed transaction quantity and price. The AIV value chain analysis conducted in 
the first phase of this study revealed transactions frequently changed after the initial agreement 
(Agnew et al. 2021). Figure 7 reveals that there was a decrease in the proportion of producers and 
traders that experience such variability; however, there was a slight increase in the proportion of 
retailers reporting variability. 6
 Participants indicated that changes in the market and changes in 
demand were the main reasons that transactions varied. Traders also cited changes in supply as 
reasons for changes in the final transaction. There was no change in the underlying cause of 
variability between baseline and endline. The majority of producers (65%), traders (100%), and 
retailers (86%) confirmed that the AgUnity app was the reason that variability and uncertainty in 
transactions had been reduced in the last six months. 
Figure 7. Transaction variability
The AgUnity app contributed to improving the reliability of transactions by increasing their precision, 
generating accurate and verifiable records, and creating a certainty of being paid (Table 3). Increased 
 
6 Producer p-value: 0.288; Trader p-value: 0.6875; Retailer p-value: 0.5075.
27 | Page
precision and quality of records are facilitated through the requirements to input an entry onto the 
blockchain, using the AgUnity app interface. When initiating a transaction, the buyer enters the 
quantity of the vegetable variety in kilograms along with its grade and total sale price. This is 
approved by the seller who then scans the buyer's unique QR code. Requiring the sale quantity to 
be specified in kilograms introduced standardization into the value chain; previously, vegetables 
were weighed by hand measure (see Agnew et al. 2021 for additional details). The record generated 
from the transaction is secure and immutable, meaning neither party can change it once it has been 
agreed upon. The precise and immutable transactions led to reduced disputes, increased purchase 
quantity, greater trust and transparency, and records available for informed decision-making. 
Table 3. Reasons for reduced variability and uncertainty in transactions
28 | Page
At baseline, only 30% of participants kept paper-based records and none used mobile technology
for record-keeping. One producer shared, “we never recorded our production or sales in whatsoever 
form before AgUnity.” This limited the efficiency of transactions because, as one farmer put it, they 
were “forgetting how [they] were 
transacting, [such as the] price and 
quantity.” 
All participants, except for two, shared 
that the AgUnity app was extremely 
beneficial for record-keeping. The 
records created by the AgUnity app 
contributed to transactional efficiency 
by allowing sellers to easily track sales made on credit and reducing the time for repayment. 
Approximately 27% of producers, 63% of traders, and 29% of retailers indicated that the AgUnity 
app had helped them to collect their debts more efficiently than before. One trader said, “now my 
retailer never fails to pay me as she always knows what needs to be paid, and I can ask anytime with 
certainty [of what she owes].” Records also improved transactional efficiency by providing a record 
of previously agreed prices. This reduced negotiation and, according to traders and retailers, helped 
them set prices more fairly and consistently. By the end of the project, approximately 50% of 
participants listed the record-keeping functionality as one of the top three features of the AgUnity 
app.
Improved connectivity and coordination between actors 
The AgUnity BCT-based app has demonstrated the potential to strengthen the entire value chain
simultaneously by improving coordination between actors. For example, since transactions in the 
AgUnity BCT-based app can only occur between actors registered in the network, coordination and 
cooperation are required in order to transact. Furthermore, each user has access to a known group 
of individuals they can transact with, reducing the transaction cost of finding buyers. According to 
one producer, “normally I had to hawk around the market for the whole day to find buyers. Now it is 
so swift. I know already who to trade veggies to and the buyer comes to fetch the goods.” One retailer 
shared during the focus group, “it is beneficial to know a complete chain. One always knows who 
she/he should trade to, and can check down the stream.” This improved coordination ensures that 
high-quality, safe, and affordable products are delivered to local markets while minimizing 
transaction costs.
Beyond BCT, smartphone ownership also contributed to more effective coordination and efficient 
transactions. For example, one producer shared, “I easily reach out to traders by phone and confirm 
[the] price and quantity, so I spend less time looking for buyers.” Another shared that they use 
WhatsApp to take pictures of the vegetables, which makes it easier to find buyers. 
During the study’s training and troubleshooting sessions, a shift in attitudes between the groups of 
actors was observed. Previously, there were feelings of frustration and resentment between actors 
PRIOR TO PARTICIPATING IN THIS 
PROJECT, ONLY 30% OF PARTICIPANTS 
KEPT RECORDS. BY THE END OF THE 
PROJECT, ALL PARTICIPANTS WERE 
USING THE AGUNITY APP TO GENERATE 
TRANSACTION RECORDS.
29 | Page
for not being treated fairly. As the study progressed, participants started to view themselves as a 
team working towards a common goal. One participant noticed this change. She shared during the 
focus group, “everybody is now connected and active.” Another shared, “everybody in the system is 
motivated as we are in a group.”
While there was no measurable change 
between baseline and endline in the number of 
producers who believed their buyers treated
them fairly, 92% of producers and 87% of 
traders and retailers reported that using the 
AgUnity app improved their transacting
relationships. These improved relationships and coordination between value chain actors 
contributed to a general preference for participants to transact with other AgUnity users (Table 4). 
Table 4. Participant preference for selling to AgUnity versus non-AgUnity users.
When asked if the AgUnity app made it easier to cooperate with other value chain actors, and if so, 
who, at least 80% of each VC actor group confirmed it was easier to cooperate by using the app
(Figure 8). The most commonly cited reasons for this increased ease of cooperation include improved 
coordination between producers (19%), increased connectivity between all value chain actors (17%), 
improved information sharing (17%), and improved communication (15%).
Figure 8. Improved cooperation between groups
THE AGUNITY APP CONTRIBUTED 
TO SIGNIFICANT IMPROVEMENTS 
IN RELATIONSHIPS BETWEEN 
BUYERS AND SELLERS.
30 | Page
Increased cooperation led to a reduction in AIV post-harvest loss and contributed to improved 
delivery of high-quality, safe, and affordable produce to local markets. Previously, producers used to 
harvest their vegetables and find sellers. Now producers find buyers for their vegetables before 
harvesting, so they only harvest what they can sell that day. One producer shared, “we have now [an] 
assured market and we minimize our loss and wastage.” This efficiency has led to a reduction in profit 
loss, ensures producers are fairly compensated, and helps supply to better meet demand. 
In addition to increased cooperation between groups of actors, there was also an observed increase 
in cooperation within groups of actors, particularly producers. In each study location, farmers 
organized themselves into legally registered self-help groups to strengthen their skills in using the 
AgUnity app, coordinate their production so they can more effectively meet demand in the market, 
and start a saving and loans fund. Producers also coordinated the varieties of AIVs produced and 
when they were sold so that the market is not flooded with one type of vegetable. 
Enhanced trust between actors 
The AgUnity app sought to improve trust between actors by verifying the identity of users and 
creating an immutable ledger of transactions stored on a public blockchain. Anecdotally, trust was 
strengthened as a result of using the AgUnity app, particularly between producers and their buyers.
For example, one retailer shared that “as the suppliers’ and buyers' names are on the phone, you 
develop trust and consistency with the connection.” Another shared that they trust the records of the 
people they are transacting with because they know they cannot be altered or changed. Overall, 
there is “a lot of trust between the players, we [work to] accommodate each other” one producer said. 
When examining the 
trust that value chain 
actors have towards 
various groups of 
stakeholders (Figure 
9), there was no 
statistically significant 
change between 
baseline and endline, 
except for traders' 
reduced trust in 
creditors and an 
increase in retailers' 
trust in other retailers 
(all p-values exceed 
0.1).
Figure 9. Average trust scores by value chain actor
Note: Grey circles represent the average trust score for on a 5-pt Likert Scale, where 1 is do not trust at all 
and 5 is completely trust. The width of the band in the figure represents the proportion of participants who 
selected the ordinal Likert response.
31 | Page
Improved Flow of Information
Information is integral to the functionality of agri-food value chains. It contributes to consumers’ 
willingness to pay a price that fairly compensates value chain actors for their respective activities and 
ensures that demanded varieties of AIVs are consistently available in the market. The AgUnity BCT￾based app facilitated the improved transmission of information on the grade of the vegetable which 
increased the prices producers received. Price signaling led producers to more effectively meet the 
demand for high-quality, popular varieties of AIVs. However, while BCT-generated records increased 
access to information for individual user groups, there was little improvement in the transfer of 
information from producer to consumer and consumer to producer throughout the study. This 
suggests there may be economic and efficiency gains still to be captured in AIV value chains. 
Adoption of vegetable grades & standardization 
According to the consumer midline survey responses, information on the date of harvest, use of 
pesticides, nutrition information, and recipes for the vegetables were important to consumers when 
deciding to purchase AIVs (Figure 10). While only 2 respondents were interested in grading 
information at the point of sale, 70% of consumers indicated they would pay a price premium of 5 
to 50% over the current price of 100 KES per bundle of vegetables for Grade A produce, signaling 
their demand for high-quality produce.
Figure 10. Information desired at the point of sale
Given BCT’s ability to track information on bundles of products from farm to fork and the findings
from the initial value chain analysis and consumer midline responses (Figure 10), it was collectively 
32 | Page
decided with participants to grade the vegetables. With guidance from Dr. Joseph Mafura, an 
Egerton University professor in horticulture, and collaboration with all participants, a tool was 
developed to grade the AIVs into Grade A and Grade B (Appendix 1 ). Figure 11 shows the adoption 
of the grading process by producers and retailers between baseline and endline.
7 While the figure 
suggests that a majority of producers and traders were grading their vegetables prior to the study, 
the initial study trainings revealed that there was no official grading system in use.
Figure 11. Grading of AIVs for sale
By the end of the study, 100% of traders and the majority of producers were using the 
formal grading process designed in this study.
Using the grading system and the BCT-based app 
to transmit this information along the value chain 
was highly accepted by project participants. For 
example, one producer noted, “with the standard 
quality and clear communication about Grade A, it 
is really easy to sell. [Its] very efficient.” By the end 
of the study, approximately 60% of retailers were 
marketing different grades of vegetables at their 
stalls (Figure 12). 
The independent evaluation of consumers in the 
study markets found that of the consumers who 
have started purchasing more AIVs in the last six 
months (n=35, 56%), 20% did so because Grade A vegetables were now available in the market. Of 
those same consumers, 92% were purchasing more because the quality of the vegetables had 
improved. This improvement was an outcome of producers realizing the improved prices they can 
obtain from high-quality, Grade A produce, which is reinforced by the records stored on their 
phones.
 7 Producer p-value: 0.0006; Trader p-value: 0.0309.
Figure 12. Retailers differentiating grades of vegetables
33 | Page
The initial value chain analysis and 
consumer baseline survey revealed that 
consumers also want to know more 
about the producers growing the 
vegetables and their farming practices. 
While BCT enables this traceability, 
smartphone ownership is not prevalent 
among consumers, limiting their access 
to information stored on the blockchain. 
While options are explored on how to 
provide consumers without a 
smartphone with access to blockchain￾based data, we created posters that 
provide pictures of their suppliers and the production practices used in growing the vegetables (e.g., 
grown without the use of pesticides, organic fertilizers used, etc.) for retailers to attach to their market 
stalls (Figure 13). These posters introduce the idea of traceability into informal, open-air markets in 
peri-urban areas of Kenya. In markets where consumers own smartphones, this information could 
be accessed directly from the blockchain by scanning a QR code or entering a lot number. In addition 
to providing information demanded by consumers, this data is valuable to producers as well. One 
producer shared, “It is exciting to be known. The struggle is worth it to keep on selling the vegetables 
and we are proud of it.” Another shared, “we feel happy and that we are known for our AIV produces.”
Increased access to information 
For popular varieties of AIVs to be consistently available in the 
market, producers must view the production of the leafy vegetables 
as worthwhile. The AgUnity app and the records it generates help 
producers track revenues and quantities of each AIV variety sold
(Figure 14). One producer shared that she now understands how 
much she is making from the sale of their vegetables because of the 
records stored in the app’s wallet. This information helped her realize 
it is worth her time to continue vegetable production. 
Before participating in this study, only a portion of those keeping 
records used them in decision making. By November 2021, the 
majority of participants, particularly producers, were using records in 
decision making (Figure 15) on varieties and quantities to plant, when 
to harvest the vegetables, traders and retailers to sell to, and means 
to reduce post-harvest loss, etc.8 In conjunction with improved 
 
8 Producer p-value: 0.237; Trader p-value: 0.5078; Retailer p-value: 0.7539
Figure 13. Producers and study personnel pictured with the market stall poster
Figure 14. Screenshot of AgUnity app 
wallet
34 | Page
coordination and cooperation between value chain actors, these changes resulted in an improved 
supply of quality AIVs to retailers. 
Figure 15. Changes in the use of records for decision making
Figure 16 shows the difference between retailers’ access to an adequate supply of demanded AIVs 
between baseline and endline. African black nightshade, spider plant, cowpeas, and amaranth are 
the most popular varieties of AIVs among participants. There was a reduction in the number of 
retailers who face difficulty in sourcing these varieties. 9 One retailer shared that before participating 
in the study, she had regular difficulty in accessing the vegetables that consumers demanded; now
she does not and has the produce she needs to stock her outlet.
Figure 16. Proportion of retailers with difficulty accessing demanded AIV varieties
 9 African black nightshade p-value: 0.6072; Spider plant p-value: 0.3877; Amaranth p-value: 1.0.
35 | Page
While there has been a notable increase in access to information on the quality of vegetables and 
practices used in their production as a result of the BCT record generation and traceability, it is not 
apparent that the flow of detailed information regarding AIV production and demanded 
characteristics have improved between actors. Figure 17 and Figure 18 show the flow of information 
from producer to consumer and consumer to producer. The width of the bands represents the 
number of individuals who indicated they share the information type with people they are transacting 
with. 
Figure 17. Flow of information from producer to consumer - baseline
 
 
Figure 18. Flow of information from producer to consumer - endline
By the endline, producers were sharing more information on production practices with the trader. At 
both baseline and endline, the flow of information is impeded as it is passed through traders and 
retailers, though the reduction in band size should be interpreted with caution as there are fewer
traders and retailers that participated in the study than producers. Regardless, only two to three 
36 | Page
traders share information that they receive with retailers and information they receive from retailers 
with consumers. Responses from consumers in the market match the information retailers indicated
they pass on to their buyers. The most common information shared with consumers includes the 
date of harvest (100%), production methods (70%), grade (60%), and harvest methods (30%), but the 
majority of consumers still do not have access to information about the AIVs.
Access to new markets 
Offering vegetables with information
rather than vegetables coming from 
opaque value chains has also opened 
opportunities to reach new markets 
that are currently undersupplied with 
AIVs in Kakamega and surrounding 
counties. The New Vision CBO will be 
able to supply the supermarket chain, 
Quick Mart, in Kakamega and Kisumu 
with their vegetables. Quick Mart was 
very interested in the traceability and transparency that BCT facilitated in the AIV value chains. An 
in-store display was developed (Figure 19) that will allow shoppers to scan a QR code that will take 
them to a website with information about the farm-to-store journey of the vegetables.
Six-month status update
Since the end of the LASER PULSE funding, the CBO New Vision continues to scan each 
transaction with their farmers. In addition to offering AIVs to institutional buyers (such as 
hotels) out of their office location, they have started a formal relationship to sell AIVs to 
SNV’s Veggies for People and Planet Project at a profit of 20 KES per kilogram. Earning 
a profit will allow New Vision to provide additional services to their buyers, including acting 
as an off-taker and buying their vegetables in bulk each week.
Figure 19. Quick Mart in-store display
37 | Page
Hypothesis 2: Improved functionality of the value chain will increase the incomes of value 
chain actors, especially women and youth.
The improved functionality of the AIV value chain led to an improvement in income for producers, 
especially women; however, the incomes of traders and retailers decreased compared to the same 
months the previous year. Only three youth (one producer, one trader, and one retailer) participated 
in this project due to enrollment constraints. In both quarters, youth reported an increase in income.
Improvement in Producer Incomes
Improvements in value chain functionality resulting from the AgUnity blockchain-based application 
led to an improvement in producer incomes from on-farm activities; however, traders’ income 
decreased across all sources of income, while retailers’ income increased from on-farm activities and 
employment. Figure 20 compares the average income reported by each actor type from April to 
June (Quarter 2) and July to September (Quarter 3) in a typical year (baseline) and after using the 
AgUnity app (endline). During Quarter 2, producer average on-farm income increased by 32% (p￾value: 0.184), and by 25% in Quarter 3 (p-value: 0.286). Their average total income increased by 26% 
and 5%, respectively (p-value 0.356; p-value: 0.879). Contrastingly, traders’ income decreased across 
all sources of income, while retailers' income increased by 80% from on-farm activities and 86% from
employment between April and June (p-value: 0.559) and 86% between July and September; 
however, average total income decreased by about 20% during both quarters (p-value 0.584). 
Figure 20. Average income by quarter and activity
Female producers experienced a larger increase in average total income in Quarter 2 (Figure 21) and 
Quarter 3 (Figure 22) than male producers, who experienced a decrease in average total income in
38 | Page
both quarters. While male producers had a larger increase in on-farm income, female producers 
brought in more income from off-farm and employment activities, contributing to a larger total 
increase in income. This difference may have resulted from the time savings created by using the 
AgUnity app. For example, one producer shared, ”I can use my time to work on other crops, to be [a] 
better mother, and serve the community.” The increase in on-farm income among men may have
resulted from increased participation in AIV production or the obtaining of livestock from increased 
income from the sale of AIVs.
Figure 21. Changes in average income by gender, April – June (Quarter 2)
Figure 22. Changes in average income by gender, July – September (Quarter 3)
TIME SAVINGS 
CREATED BY 
THE AGUNITY 
APP LED TO 
IMPROVED, 
DIVERSIFIED 
INCOME FOR 
WOMEN.
39 | Page
Part of this increase in income was derived from an increase in the sale of AIVs. Approximately 85% 
of producers, 75% of traders, and 43% of retailers indicated that their income from trading in AIVs 
had increased over the last six months. Approximately 67% of producers, 87% of traders, and 57% 
of retailers felt that the records generated from the AgUnity app contributed to an improved income.
Improvement in Youth Income
While only three youth participated as users in this project,10 each reported an increase in income in 
both quarters of the study period compared to the same months in the previous years. The 
producer’s on-farm income increased by 20% in Quarter 2, 40% in Quarter 3, and was able to 
diversify into off-farm income earning activities (Figure 23, Figure 24). The trader’s income increased 
by approximately 300% in Quarters 2 and 3 with no income reported at baseline. The retailer’s 
income increased by approximately 300% in Quarter 2 and 750% in Quarter 3, similarly with no 
income reported at baseline.
Figure 23. Changes in average income among youth, April – June (Quarter 2)
 10 The method of recruitment to ensure all parties transact with one another in each of the value chain 
arrangements constrained the number of youth that could be engaged in this project as users of the phone 
since youth participation in agriculture is low in Kakamega County.
40 | Page
Figure 24. Changes in average income among youth, July – September (Quarter 3)
Six-month status update
The producer youth was appointed by the producers in his value chain group to serve as 
an ‘AgChampion.’ This role was designed by AgUnity to ensure that farmers had support 
in learning how to use the app and to create functional feedback loops. Because of this 
youth’s leadership role in the project and use of technology in his operation, more than 
150 youth in his sub-county are now interested in participating in the AgUnity 
network and re-engaging in agriculture. Some of these youth had no interest in 
agriculture until they saw this AgChampion’s success with his operation and the fair market 
linkages that the app creates.
Contributors to Improved Income
The factors that value chain actors perceived as contributing to improved income are summarized 
in Table 5. Producers increased the volume of their production by organizing themselves into groups
and pursuing training to improve yields and the quality of vegetables produced. This allowed 
producers to transact higher volumes of Grade A vegetables which led to higher prices for AIVs. 
Traders also perceived that they were also able to get better prices for the AIVs from the retailers.
41 | Page
Table 5. Why has income increased from transacting in AIVs?
Increased production of quality AIVs 
Of the 70% of producers (n=26) that indicated that increased income resulted from increased 
production, 74% of female producers cited higher volumes of production (n=10) compared to 65% 
of male producers (n=7). Figure 25 summarizes the average increase in production by variety with 
the number of producers that cited an increase in production of that variety.
Figure 25. Average increase in production by variety
42 | Page
This increase in production was achieved by improved production practices since producers realized 
they could earn a price premium for higher quality vegetables. Improvements in coordination within 
and across groups of actors in the AIV value chain enabled groups to seek training to improve the 
quality of their AIV production, recognizing the price premium that Grade A vegetables offer. One 
producer emphasized the importance of this training stating, “the training about production, quality, 
grading and how to produce organic vegetables really help[ed] our productivity.” Some of this training 
was provided through the study of how to maximize the benefits of BCT. Value chain actors were 
taught about how to ensure traceability of the different grades and how to market these to the 
consumer. Increased income has also allowed producers and producer groups to invest in expanding 
their production. One producer said, “as a group now we look for land to lease on the riverbank to 
increase our production yield.” 
Higher prices for AIVs 
Roughly half of producers and traders, and a third of retailers, perceived they were getting better 
prices for AIVs as a result of using the AgUnity V3 SuperApp because they were able to trace the 
grade of the vegetable. For example, one trader said, “other retailers [in the market] already know 
about Grade A vegetables and are willing to pay more [as well].” When asked why the price increased, 
the majority of producers cited increased demand and better quality of vegetables (Table 6).
Table 6. Perceived reasons for improved prices
Figure 26 shows that prices increased for cowpeas and Sunn Hemp, while they decreased for 
amaranth and pumpkin leaves. However, as Figure 27 shows, the quantity and amount paid to the 
seller of the varieties transacted over the study period (cowpeas, Sunn Hemp, pumpkin leaves, and 
amaranth) increased. Thus, the sellers may perceive they are receiving better prices, even though the 
average price per kg decreased for amaranth and pumpkin leaves. These data were obtained from 
the transactions that were recorded on the AgUnity app.
43 | Page
Figure 26. Average price per kilogram by variety, quarterly
Figure 27. Amount received in shillings and quantity sold in kgs, quarterly
44 | Page
Diversification of income-earning opportunities 
Increased income from AIV production has created access to other types of economic opportunities. 
For example, one producer said that the additional income earned by her producer group enabled 
them to acquire land to start growing sweet potatoes. About 50% percent of producers have also 
indicated they have been able to acquire livestock such as local chickens (n=18), local cows (n=7), 
local goats (n=5), improved cows (n=2), and sheep (n=2).
Not all opportunities to diversify income had materialized by the end of the project; however, the 
improved coordination and cooperation within and across value chain actors have helped orient 
participants towards identifying diversification opportunities. For example, the CBO intends to 
explore other revenue-generating activities, such as producing and selling compost tea or selling 
agricultural inputs, which would allow them to provide training for their members and assist them in 
growing their production and market access.
45 | Page
Hypothesis 3: Increased incomes of value chain actors lead to improved food security at the 
household level. 
There was clear evidence that food security did increase over the 6-month study period. Value chain 
actors (i.e., producer, trader, retailer) demonstrated an ability to purchase higher quality proteins and 
fruits when compared with the baseline. In addition to increased income, the focus of the project on 
the nutritional quality of AIVs and how to use these characteristics to market the product to 
consumers also helped to increase the demand of participants for the leafy greens.
Reduction in Experiences of Food Insecurity
Table 7 shows that the average food insecurity experience scale (FIES) score for producers, traders, 
and retailers decreased over the study period, though the changes were not statistically significant. 
Scores closer to zero represent less severe food insecurity experiences.
Table 7. Average Food Insecurity Experience Scale (FIES) score
Figure 28 breaks down the responses to the FIES items based on gender. For all items, a larger 
proportion of women compared to men reported these experiences. Reported instances of these 
food insecurity experiences decreased for all items except for 4 additional women who reported that 
their household ran out of food. The greatest decrease occurred for female value chain actors who 
were hungry but did not eat, reducing from 42% to 11%. There was also a 20% decrease in the 
number of women who only ate a few types of food and a 17% decrease in the number of women 
who ate less than they thought they should.
Figure 28. Respondents to FIES items by gender
46 | Page
Diet Diversity Score (DDS)
As shown in Table 8, there was no significant increase in diet diversity. For female traders, the average 
diet diversity score (DDS) decreased by almost 1 point and male traders’ average score decreased by 
0.5 points. Male traders’ average DDS is higher than female traders, but there is no statistically 
significant difference between male and female producers (all p-values greater than 0.1).
Table 8. Average DDS by value chain actor and gender 
According to reports of consumption of 12 food groups in the previous 24 hours, the number of 
households reporting consumption of leafy green vegetables decreased between baseline and 
endline. In fact, there was a decrease in the number of households reporting the consumption of 
several food groups including grains, roots and tubers, dairy, flesh foods, shellfish or fish, vitamin A￾rich fruits and vegetables, other fruits, eggs, fat, and sugar (Table 9). However, as shown in Table 10, 
there has been an increase in the number of households who purchase dairy, legumes and nuts, and 
fruits (vitamin A rich and others).
Table 9. Foods consumed in the previous 24 hours Table 10. Foods purchased 
47 | Page
Increased Vegetable Consumption
Despite the fact that the dietary recall indicated that consumption of leafy greens decreased over 
the study period, 92% of producers, 88% of traders, and 100% of retailers indicated that the quantity
of indigenous vegetables consumed in the household increased during the previous six-month
period. The most common reasons cited for this were increased production (42%) and the fact they 
are now producing a surplus (14%), increased awareness of nutritional importance (21%), and 
increased preference for vegetables (11%). Increased awareness of the nutritional characteristics of 
indigenous vegetables was commonly shared by participants during the endline focus groups. 
According to one participant, “we consume AIV more than before as we understand its nutritious 
quality [now].” 
48 | Page
Hypothesis 4: In retail outlets, the blockchain application will improve the marketability of 
AIV varieties by addressing gendered factors that influence food purchase, preparation, 
and consumption. 
Of central concern to both women and men is the availability and quality of the AIV varieties they 
are looking for – both of which increased over the study period. There was also a small increase in 
access to information about the vegetables and how they are produced, which is particularly 
important to women. Over the study period, the awareness of the nutritional importance increased 
among both men and women; however, a relatively small percentage learned of the nutritional 
characteristics of AIVs from the retailer they purchase from.
Gendered Factors that Influence Food Purchase, Preparation, and Consumption
At baseline, consumers in the five study markets were asked what factors would be likely to 
contribute to increased AIV purchase, preparation, and consumption. Quality of the vegetables, 
availability of preferred varieties, knowledge on how to prepare the vegetables, quantity, and 
information on how the vegetables were produced were all important factors to consumers. Figure 
29 demonstrates that the male and female responses were similar. There are some small differences
though; for example, more men than women indicated that more money would be required to 
purchase AIVs, and increased quality and quantity of vegetables needed to be available. More 
women than men indicated that the varieties they prefer needed to be more available and that they 
wanted more information on how the vegetables were produced.
Figure 29. Factors that could increase AIV purchase
49 | Page
The consumer midline survey (n=366) confirmed that quality and information about how the 
vegetables were produced are important factors in the purchasing decision (Figure 30). 
Figure 30. Consumer preference for information about AIVs
Quality 
Consumers were asked during the baseline and endline to rank the quality of vegetables they were
able to purchase in local markets. As Table 11 shows, there was an approximately 60% increase in the 
proportion of men and women who perceive vegetable quality as excellent. By the endline, neither 
women nor men perceived the vegetable quality as poor.
Table 11. Ranking of AIV quality
When asked directly if vegetable quality had specifically improved over the last 6 months compared 
to other periods, 94% of respondents answered yes.
Availability 
AIV availability is largely determined by season. Leveling out the fluctuations in availability is essential 
for capturing economic and food and nutrition security benefits. Table 12 shows that the availability 
of AIVs increased during the six months of the study in the focus markets, despite a drought that 
was affecting agricultural production in the region. By the end of the study, 100% of both men and 
50 | Page
women agreed that the AIV varieties they were looking for were available in the market; however, 
approximately 6% of women indicated that when they were available, they were too expensive.
Table 12. Availability of preferred AIVs in the market
Food safety 
Guarantee of the safety of AIVs is central to the consumer’s decision to purchase and consume leafy 
greens. According to one retailer, “the consumer wants to know for sure what’s been … used [in the 
production of the vegetables] and it requires trust.” Table 13 shows that access to information about 
AIV production and food safety has only increased slightly among consumers since the beginning of 
the project. Access to information about the vegetable grade has increased, especially for female 
consumers. The proportion of consumers with no access to information about their vegetables 
decreased by about 25% for both men and women; however, consumers still have little access to 
information on the inputs used in the production of AIVs, production practices, water quality, and 
the transportation methods of the vegetables.
Table 13. Access to AIV information
Cultural and nutritional importance 
AIVs are also important cultural vegetables as they are part of the Kenyan cuisine, served alongside 
ugali (a thick porridge made from maize meal) and fish or chicken, particularly in rural areas. They 
also have gendered roles in their preparation. Consumers were asked the extent to which they 
agreed with the cultural importance and preparation of the vegetables at baseline and endline (Table 
14). Women originally held a much stronger belief than men around AIVs and their cultural 
51 | Page
significance; this changed significantly for men between baseline and endline. Women also more 
strongly believed that AIVs have positive nutritional benefits; however, this too significantly increased 
throughout the study among men. Generally speaking, there is a confidence that both the 
respondent and the respondent’s spouse can prepare AIVs; there is a substantial improvement in 
the confidence of women that their spouse can prepare the AIVs between baseline and endline. 
There is also a consensus among both men and women that AIV producers should be supported.
Table 14. Cultural and nutritional perceptions of AIVs
 
When asked if the awareness regarding the nutritional quality of AIVs increased over the previous 6-
month period, 84% of women and 67% of men indicated that it had; however, only 5% of 
respondents indicated that they had learned about the nutritional value of the AIVs from their 
retailer. When asked if their awareness of how to cook vegetables in ways that retain their nutritional 
value had increased, 97% of women and 67% of men said yes. Similarly, only 1 respondent had 
learned about this information from the retailer.
52 | Page
Hypothesis 5: Increased marketability of AIVs will translate into increased purchases of
AIVs by low-income households.
The markets that were selected for this study primarily serve consumers in the third- and fourth￾income quintiles. According to retailers, participating in the network of value chain actors using the 
AgUnity BCT-based app has improved the marketability of AIVs. This was primarily because the 
quality of the vegetables increased, and retailers were able to market vegetables from AgUnity 
producers as Grade A. One retailer stated, “knowledge about grading and quality is easily accepted 
by the consumer. There’s no challenges to communicate.” Another said that “most consumers 
appreciate the grade and are willing to accept the premium price [for Grade A vegetables].” This was 
supported by the consumer midline, which revealed that there could potentially be as much as a 10 
to 30 KES price premium for grade A vegetables.
As Table 15 shows, there has been a self-reported increase in the volume of vegetables purchased 
in the last 6 months. A slightly larger proportion of women compared to men reported increasing 
the quantity of vegetables purchased.
Table 15. Increase in AIV purchase over the previous 6-month period
Figure 31 outlines the reasons for this increase. For all participants, better quality was the leading
factor driving the increase, followed by better prices and more awareness of the nutritional 
importance of AIVs. 
Figure 31. Reasons for increased purchase of AIVs
53 | Page
In addition to an increased quantity of AIV purchases, consumers spent more on average per week 
between baseline and endline on cowpeas, amaranth, and jute mallow. This is consistent with the 
price information collected from the AgUnity record-keeping app that indicated the increased prices 
were passed on to producers. 
Table 16. Average weekly expenditure on AIVs (KES) 
 
54 | Page
Hypothesis 6: Value chain actors will be willing to pay a small fee for the blockchain 
application for the economic and nutritional benefits derived from its use.
The value of the records generated by the AgUnity BCT infrastructure was clearly demonstrated by 
all participants; however, in order to be willing to pay the 150 KES needed by AgUnity to sustain 
service provision, participants indicated they would need to have access to complementary services. 
In addition to the willingness to pay for the app, participants also were willing to pay the full cost of 
the phone (USD 75) if they could have access to affordable repayment terms. Anticipated benefits 
and intention to use the AgUnity V3 SuperApp were mostly unchanged between baseline and 
endline, though a slight cooling-off effect was observed.
Willingness to pay for the AgUnity app 
Approximately 75% of producers, traders, and retailers indicated they were willing to pay 150 KES
per month to continue to use the AgUnity app; however, to be willing to pay this price, the app 
would also need to offer complementary services and features. One participant shared, “I would pay 
more, about 50 KES, for additional services such as reports, or micro or crop insurance.” Table 17
summarizes the services that consumers were interested in. All participants in the final focus groups 
where this activity was conducted indicated that they were interested in complementary services. An 
applet providing weather information and the ability to market vegetables online were the most 
popular apps that could be offered in addition to the blockchain. 
Table 17. Interest of participants in complementary services
Service Number of Participants
Weather information 23
Online marketing 21
M-Pesa integration 16
Pest control consultation 8
Extension services 7
Transportation service 
location 7 
Information on nutrition and 
recipes 5 
Sentiments toward the AgUnity App
Prior to participating in the study, participants were asked about the expected value of the AgUnity 
app in their operations and the extent to which they were willing to incorporate its use into their 
operations. The questions were modeled after Michels, Bonke, and Mußhoff (2019) and the Unified 
Theory of Acceptance and Use of Technology (UTAUT). Initial (baseline) perceptions of the AgUnity 
app were favorable, and participants for the most part expressed a strong interest in its potential 
value, indicating they intended to use it in their operations. By the endline, there was a slight cooling 
55 | Page
effect, with more users tending towards a moderate agreement with the use and impact of the app
among the participants using the AgUnity phone (Figure 32). 
Figure 32. Perceptions of using the AgUnity app 
Note: A 5-pt Likert Scale was used to measure attitudes towards the AgUnity app.
Participants largely enjoyed using the AgUnity app for reasons related to BCT as well as the design 
interface. It was clear that users understood they could not alter the records and that the people 
they were transacting with could not alter theirs. This contributed to the trustworthiness of the 
records. Several participants shared that they knew they could trust the records because they had 
been verified by two people (themselves and the buyer/seller), meaning they were mostly free from 
mistakes. This differs from recording transactions on paper – which can be lost or have an error. 
Participants also appreciated that the app was in Swahili which helped increase usability.
Despite its many positive benefits, there are still some technical challenges that need to be addressed
to improve the usability of the app. For example, one participant shared that at times “scanning is 
quite challenging.” This occurs because of a reflection created in the sun that prevents the QR code 
from being scanned. One participant said that “we have to find shade in order to be able to scan.” 
Another challenge is the intuitiveness of the app for traders. Since AgUnity has historically worked 
with value chains centered around cooperatives, their system needed to be adapted to include 
traders. Thus, this new capability had not benefited from a prior deployment, so additional 
improvements may be needed to ensure traders have the same user experience as producers and 
retailers.
56 | Page
Participants had several ideas on how AgUnity could expand its network, which is critical for ensuring 
there are enough buyers and sellers in the system. One idea was for AgUnity phones to be available 
in the local shops or to have specific AgUnity retail outlets. Another idea was to work closely with 
community-based organizations whose members could be trained to provide support and onboard 
new users into the system. 
Community Interest in Joining the Network
In addition to participants recognizing the value of using the AgUnity app and smartphone in 
performing their respective activities in the value chain, members of the participants’ communities 
and social networks were also interested in joining the network. One of the youths participating in 
the project shared that his friends were getting interested in agriculture that used technologies such 
as the smartphone and AgUnity app; they recognized how much their friend’s agricultural business 
had improved since joining the study.
All participants in the study shared that if they had to pay for the smartphone to use the AgUnity 
app, they would be willing to pay 8,500 KES (USD 75) and believe that others in their community 
who wish to join the network would pay the same. They estimated that this could be paid for over 
six months, perhaps needing a month or two longer to repay if they had to simultaneously pay for 
the AgUnity subscription fees.
Six-month status update
A field visit conducted in May 2022, revealed that there is a group of at least 150 youth
that have learned about the AgUnity V3 SuperApp and the benefits of blockchain-based 
record-keeping and value chain coordination. They are currently saving to be able to 
purchase a phone in order to join the network. The entirety of New Vision CBO’s 120 
farmers is also interested in joining the network. AgUnity plans to pilot a kiosk operation 
in collaboration with New Vision to start incorporating these interested participants. 
57 | Page
Gender Implications
In addition to investigating the outlined six hypotheses, this study was concerned with understanding 
how ICTs such as smartphones and blockchain technology might help women secure their place in 
the value chain as upgrading activities are undertaken. It was also concerned with how introducing 
such technology into households might shift gendered roles, not only between spouses but also 
among children and adolescents in the household. Shifts in agricultural roles were reported in about 
one-third of participating households, while shifts in decision-making regarding financial decisions 
were reported more widely. 
Change in Roles
Approximately 20% of male producers and 30% of female producers reported that household 
responsibilities for producing AIVs had changed since introducing the AgUnity blockchain-based 
app into their respective production. At baseline, decision-making regarding the vegetables or 
phone management was not reported as a defined household role. Of the female farmers that 
reported a change in roles, 27% said they now had more decision-making regarding the vegetables 
and 54% said they had control over phone management (Figure 33) that they didn’t have before.
Reasons for a change in roles for female farmers: 
 “I saw the value in AIV production and marketing.”
 “Since I got training from the AgUnity group, I find it easy raising indigenous vegetables hence
it takes me a shorter time to complete the farm work and do my side chores. Thus, I engage in 
so many more roles than before.”
 “Nowadays, it’s easy to sell online unlike before where I had to search for customers manually,
so roles had to change.”
 “This training has helped me know what it takes to produce vegetables.”
 “Impressed with income.”
Reasons for a change in roles from male farmers:
 “They change with weather conditions.”
 “Kids are in school.”
 “I have solely undertaken vegetable farming.”
Household roles for other members of participating families also changed. For example, the 
responsibilities of adolescent children were reduced for all activities except marketing, while the 
responsibilities of young girls and boys were significantly reduced for preparing the field, planting, 
irrigating and bundling, cleaning, and packaging. Young boys' participation in harvesting was 
reduced while young girls' participation remained unchanged.
58 | Page
Figure 33. Change in gender roles of AIV production
There was also a shift in gender roles as it pertained to money and financial decision-making in the 
household between baseline and endline (Figure 34). For female producers, there was a 5% reduction 
in the households where financial decisions are made by someone else. Among male producers, 
there was almost a 20% reduction in making bigger decisions on their own without the input of a 
spouse. Also, among female producers, there was a 15% increase in the ability to make day-to-day 
decisions even if their spouse disagreed; however, this increased for men as well by 35%. For female 
traders and retailers, sole decision-making ability decreased significantly over the study period.
59 | Page
Table 18. Changes in household financial decision making
Gender Risks
Only 15% of female producers and 12% of male producers indicated that at the time of the baseline 
they faced gender-based risks. For women, these risks included physical risks in transporting goods 
to market, having many people to support, and a lack of timely payment for products. For men, risks 
included the challenges of having many people relying on them for support and being responsible 
for understanding and predicting the weather. By the endline, only 5% of male and female producers 
reported gender-based risks. Men did not respond to the follow-up question on what these risks 
were, but the female producers indicated these risks were because of the exploitative behavior of 
their spouses who gained financially from the increased value of their vegetable production, even 
though women do all the work.
60 | Page
Youth Engagement
Only two participants in this study were under the age of 30.
11 One unanticipated constraint of 
engaging youth as the primary beneficiary in this study was that since all value chain actors must be 
already connected, finding youth engaged in farming was extremely difficult. As a result, we were 
not able to engage as many youths in the project as we anticipated. However, there were two 
important findings we did observe through the ERT processes. The first was that because the project 
used smartphones and technology, youth that became aware of the project started to become 
engaged with the phone users even though they were not direct beneficiaries. The second was that 
there are numerous opportunities to engage youth in ICT-based projects beyond agricultural 
production alone.
Youth Engagement through Farmer Groups 
As discussed further in the next section, 
one of the direct outcomes of this project 
was producers organizing themselves into 
farmer groups to create support in 
learning how to use the technology, 
marketing their AIVs, creating village 
savings and loans mechanisms, and 
pursuing training to increase their 
production of Grade A vegetables. These 
farmer groups attracted youth to the 
project. In addition to being interested in 
joining the AgUnity network, youth 
assisted the older beneficiaries of the 
project in learning how to use the phone.
In some instances, as pictured to the right, 
youth assisted their parents in receiving trainings and understanding how to use the phone.
Complementary Services and Opportunity Development to Engage in Agriculture
Youth in agriculture must also consider sectors other than agricultural production through which 
youth can be engaged. These opportunities must be interesting to students and align with their 
career aspirations and quality of life goals. This project created two avenues other than the direct 
AIV value chain to encourage youth participation in agriculture – research and tech development.
Research team 
A team of 10 Egerton University fourth-year students and 2 master's students from the agricultural 
faculty, most of whom were agribusiness or agricultural economics students, assisted in the 
 11 USAID defines a ‘youth’ as any one 29 years of age or younger.
Figure 34. Participants with a youth from the community who assisted during 
training
61 | Page
formation of data instruments, data collection, and data verification at every stage of this study. 
Drawing on students’ expertise and contextual knowledge of communities and ways of life in the 
study greatly enhanced the quality of the study design and implementation. Such opportunities also 
provide students access to global networks and provide important real-world experience that assists 
in accessing the job market upon graduation.
Ideathon & Hackathon 
In November 2021, AgUnity in collaboration with Egerton University and supported by Virginia Tech 
hosted a two-week ideathon and 3-day hackathon at the university’s Njoro campus. Forty-two 
students formed interdisciplinary teams of 5-6 individuals from the agricultural and computer science 
faculties. The purpose of the event was to further expose youth to the ethos of the LASER PULSE 
project, to deliver practical, research-driven solutions to global development challenges. Students were 
briefed on the outcomes from the initial value chain analysis and other challenges that producers 
and AIV actors shared with the research team that had not yet been addressed through the AgUnity 
V3 SuperApp. 
The Ideathon consisted of several 
synchronous trainings for students on 
the outcomes of Phases I and II of the 
research on AIV value chains and 
employing blockchain technology to 
address market inefficiencies. It also 
trained students on human-centered 
design, the user experience, gender 
considerations in tech development, 
and the protection of intellectual 
property. Each team was required to 
complete the Ideathon Workbook before they could participate in the hackathon. This workbook 
asked students to think about the value proposition of the applet they would code for the AgUnity 
Super App environment, how they would address the pain points of their target users, map the user’s 
journey, and create the task, wire, and user flows. AgUnity developers then met with students to 
discuss their workbooks and proposed an applet to simplify and strengthen their ideas so they would 
be prepared for the hackathon.
The hackathon took place over 3 days. Computer science students were responsible for coding the 
applets. AgUnity provided mentorship to the computer science students on how to code for the 
AgUnity environment. The hackathon concluded with students providing a pitch on their respective 
applets. The final applets developed were:
 AgFinders – connects value chain actors to improve the marketing of their vegetables by 
providing the location of other value chain actors to connect with. 
Figure 35. Ideathon workbook image
62 | Page
 Agribook – records the transactions and expenditures of farmers to help improve financial 
management. 
 AgForum – creates an online forum for farmers and experts to share solutions to production 
challenges. 
 AgGenius – helps farmers find solutions to the challenges they face. 
 My Wiki – provides an offline Wikipedia of information for farmers on how to improve their 
production of indigenous vegetables. 
 AgFuture – provides an e-commerce applet for marketing vegetables online. 
 DuduFree – provides information to farmers on how to manage pests and diseases. 
Applets were judged by a panel of experts based on: how well the students communicated the 
challenge they were addressing with their applet; if the applet addressed that problem effectively; 
the functionality and usability of the applet; the visual appeal of the user interface; the complexity of 
the code required to create the applet; and the ability of the team to address the challenges they 
faced during the design and coding process. The first-place prize was awarded to AgFuture, second 
place was awarded to AgFinders, and third place was awarded to My Wiki. The first-place prize 
consisted of $500 and ongoing mentoring from AgUnity to deploy the applet permanently in their 
environment. The second-place prize consisted of $300 and the third-place prize was $150. All 
students were able to enter their app idea into the Kenya ICT Authority Whitebox program, which 
provides incubation for youth with app concepts to get them ready to take to market.
Figure 36. Egerton student participants in the hackathon
63 | Page
Other Findings
This study produced rich findings on AIV production and marketing, the impact of introducing 
technology into an otherwise informal value chain, and research and development opportunities for 
AgUnity beyond our original six hypotheses.
AIV Production and Marketing
While the AgUnity BCT-based app helped address some of the pressing challenges in AIV value 
chains, particularly in terms of the nature of transactions between value chain actors, producers still 
need considerable production assistance for them to capture the potential economic benefits of 
producing highly nutritious crops like AIVs. In particular, AIV production is affected by pests, disease, 
and drought. These issues in turn affect the income-earning potential, consistency of supply, and 
prices in the market. 
Producers frequently cited the need for production assistance. For example, one producer stated 
that “we need to be supported with weighing scales, equipment, fertilizer, wheelbarrows, and a water 
and irrigation system so we have reliable production.” There are also challenges in accessing high￾quality inputs such as seeds and organic fertilizers and biological control for pests and disease. 
However, some of the participating producers have already taken the initiative to use skills, 
knowledge, and additional income gained from participating in this project to start investing in their 
production. For example, a member of the CBO shared that they “are looking for training on grading 
and integrated pest management.” Others shared that through forming a legally registered self-help 
group, they were able to open a joint bank account and start a savings and loan scheme that allows 
producers to invest in new equipment such as irrigation equipment or better-quality seeds.
Continuing to expand value chain actors participating in the AgUnity BCT network will increase the 
income-earning opportunities for producers in growing Grade A vegetables. One producer 
emphasized the importance of this since currently, “other buyers do not necessarily know about 
grading, and so we cannot sell as we cannot reduce the price.” Other producers echoed this sentiment. 
One stated, “it is challenging when producers are many but there is only one trader,” while another 
said, “we need to onboard more traders and retailers, so vegetables are always bought.” 
Retailers still need assistance in marketing AIVs, particularly on the nutritional value and the organic 
quality of the vegetables. One retailer stated, “we need still to be supported to promote better the 
knowledge about AIV, its nutrition and the organic process. Umbrellas or tents help [in attracting 
attention], especially in open markets.” Retailers also felt that other materials branded with the 
AgUnity name, or something related to vegetables would also assist in marketing. They suggested 
aprons or t-shirts as examples. As previously discussed, the increase in consumer knowledge 
regarding the nutritional quality of indigenous vegetables did not occur through the retailers. 
However, even among value chain actors, there is a recognition of the improved quality of the AIVs 
that is driving increased consumption. One participant shared, “the app makes me more aware of the 
quality of the AIV as the selection of goods in the app are readily differentiated between grade A and 
64 | Page
B.” However, there is still a need to communicate the nutritional value of AIVs and how to prepare 
them so they retain their nutritional content. One retailer shared, “we would like to understand better 
the nutritional value and how to process the AIV and to share the knowledge to our customers.” 
Potential for Standardizing Transactions
The AgUnity BCT-based app and improved 
transactional efficiency have the potential to 
contribute to higher revenues from AIV sales by 
introducing standardization into the value chain. 
Prior to the study, hand bundling was the most 
common way for AIVs to be priced. AIVs are 
commonly sold from producer to trader in 
bundles the size of two hands (an average of 
500g + 100g) for 10 to 15 KES. These are then split 
by the trader who sells to the retailer in bunches 
roughly the size of one hand (250g + 100g) and 
sold for 20 to 25 KES. Retailers then sell to 
consumers for roughly 20 KES per bundle of 100g + 150g. Figure 39 the splitting process and the 
average price in shillings for an average of 500g of a vegetable. The producer captures only 10-15% 
of the final price, while the trader and retailer capture between 40 and 50%. During a marketing 
training session, a demonstration was given on the variability of bundle weights and the revenue 
that all value chain actors were missing out on (Figure 38). The participants demonstrated an
emphatic consensus that standardization should be introduced into the value chain. 
Figure 38. Price transmission along the value chain
Improved Financial Literacy
Figure 39 shows changes in the financial situation and beliefs over the 6-month study period, which 
shows the percentage of respondents who agreed, to various degrees, with the statement on the 
left. There was about a 3% reduction in the number of respondents, including men, who believed 
that men should have the final decision regarding household finances. There was also a 10% 
Producer  Trader
~500g – 10-15 KES
Trader  Retailer
~500g – 40-50 KES
Retailer  Consumer
~500 g – 100 KES
Figure 37. Trader and producer negotiate prices for bundles of 
vegetables
65 | Page
reduction in the proportion of women who find it more satisfying to spend than to save. Another 
interesting change is the 7% reduction in both men and women who worry about paying for essential 
expenses. 
There was a slight decrease in the proportion of the participants who have money left over at the 
end of the week or month. We know from the voluntary focus groups that many producers organized 
themselves into registered groups and started a village savings and loans scheme. However, 
additional follow-up is needed to understand from traders and retailers if their available cash leftover 
at the end of the week has increased or decreased, and, if so, why. 
Of concern is the slight reduction in participants who are honest, even if it puts them at a financial 
disadvantage, a slight increase in the percentage of men who are unable to save due to their financial 
position, and the proportion of those who are paying attention to their financial affairs.
Food Traceability and Safety
AgUnity’s BCT-based app introduced different grades of vegetables to value chain actors and how 
to keep these grades distinguishable and traceable as they move from the farm to the retailer. As 
mentioned previously in this report, the grading system evaluated whether or not pesticides or 
inorganic fertilizers were used in the AIV production process. Important next steps will be to 
introduce an external AIV grading verification and monitoring mechanism, deepen consumer trust 
and create transparency for consumers who may not be located so close to the area of production 
(i.e., urban Nairobi consumers). Participants recognize the value of introducing an external 
verification mechanism into the system to continue to increase consumer confidence in their product. 
One producer shared that “we feel it is sensible to have the quality assurer come check our plot 
quarterly.” This was echoed by the other participants.
A non-BCT-related outcome of working with retailers to improve the marketability of AIVs was an 
increase in the recognition of the importance of food safety because of the focus on the issue in 
trainings related to grading. This led several retailers to improve the cleanliness of their stalls and 
how they presented their AIVs for sale. One retailer said, “as retailers, we are more aware of the 
importance of the cleanliness of the stalls and showing the quality of our vegetables.” 
66 | Page
Figure 39. Changes in financial literacy by gender
67 | Page
Recommendations
This impact evaluation identifies the potential value that BCT can offer last-mile agricultural systems
to improve value chain functionality and livelihoods. BCT cannot offer solutions for all the challenges 
faced by AIV value chain actors; however, there are clear linkages between the technology and food 
security outcomes. The following recommendations outline several areas for technology companies 
(i.e., AgUnity) and enthusiasts, policymakers, and practitioners that will facilitate the research and 
development needed to fully capture the transformative potential of BCT.
Recommendation 1. Developing low-tech traceability solutions should be a core priority 
for increasing demand for nutritious foods and creating economic returns for the 
producers of those foods.
The demand for high-quality, organic AIVs is growing in Kenya. By integrating traceability into AIV 
value chains, producers using traceability solutions will earn a first-mover advantage by meeting 
unmet consumer desires for information. This will lead to improvements in income, food security, 
and livelihood resilience. This evaluation reveals that conditions in the market are primed for 
traceability. BCT can facilitate improved transparency into otherwise opaque value chains. Tech 
companies like AgUnity need to consider how to make traceability functional for multiple types of 
consumers – i.e., those with a smartphone and those with a lower-tech phone but who still value 
information that will help keep her and her family safe.
Recommendation 2. Developing assistive technologies and monitoring mechanisms 
that minimize human interference will allow the full potential of BCT to be realized in 
agri-food systems.
Through the ERT process, we realized the need for developing monitoring mechanisms to provide 
consumers with external assurance of vegetable quality and if they were produced organically. We 
had initial discussions with the Ministry of Agriculture, Livestock, and Fisheries to collaborate on this 
monitoring and verification; however, further investment is needed to develop these systems with 
input from consumers to ensure that it is perceived as trustworthy. There is a need to develop 
assistive technologies, such as internet of things sensors or weigh scales, that will minimize human 
interference in the metadata of the vegetables. If these solutions can be identified, national 
stakeholders like the Kenya Bureau of Standards would be able to leverage BCT to scale up 
compliance to its standards across sectors. This in turn would develop consumer confidence in 
locally-produced horticulture products and facilitate access to export markets.
Recommendation 3. Community Based Organizations (CBOs) and other producer 
associations should be engaged to achieve network scalability and acceleration of 
transaction cost reduction. 
The CBO New Vision was imperative to the success of this project. CBOs are adept at community 
transformation, encouraging behavior change, and harmonizing the community response. In this 
68 | Page
case, we consider the community as the participating producers. The New Vision director was 
instrumental in coordinating activities among all participants, even those that do not belong to the 
New Vision CBO. AgUnity has worked with the director as a main point of contact to enable the field 
officer to easily carry out capacity-building exercises. Given the number of farmers in the CBO, it has 
also facilitated an easy pathway to scalability. Working with CBOs, cooperatives, and 
farmer/agribusiness associations going forward will help to achieve scale in blockchain-based 
transacting networks. This will also lead to accelerated reductions in transaction costs in the agri￾food system, leading to both increased income and a lower cost of food, both of which will contribute 
to improved national food security.
Recommendation 4. BCT should be deployed alongside complementary digital services.
The initial value chain analysis and outcomes observed in the impact evaluation reveal that BCT alone 
cannot address the pain points constraining the functionality of AIV value chains. Producers, traders, 
and retailers need access to advisory and extension services, financial services, weather information, 
and markets for their produce that will continue to minimize transaction costs and ensure they are 
fairly compensated for their respective activities in the value chain. BCT has natural 
complementarities to some of these services. For example, BCT-based transaction records create a 
low-cost, low-risk financial history that could theoretically open up access to credit. The use of good 
agricultural practices or organic production could be stored on the profile of the farmer and
communicated to the final consumer. One of the main findings of this project is that it is not BCT, 
but BCT Plus Services that has the most potential to alleviate the pain points of last-mile agricultural
systems.
Recommendation 5. Digital solutions should not be a replacement for in-person 
technical training and capacity building.
The potential for digital solutions is attracting increasing investment and attention for their potential 
to address deeply entrenched challenges in the agri-food system. However, in-person interaction 
and engagement to build technical skills in producing or retailing agricultural products will continue 
to be invaluable. For example, while the AgUnity BCT app helped address some of the pressing 
challenges in AIV value chains, particularly in terms of the nature of transactions between value chain 
actors, producers still need considerable production assistance for them to capture the potential 
economic benefits of producing highly nutritious crops like AIVs. Retailers also need assistance in 
marketing AIVs, particularly on the nutritional value and the organic quality of the vegetables. While 
some of this could be offered digitally, in-person engagement to problem solve and examine the 
existing practices used will be essential for improving productivity sustainably. 
69 | Page
Conclusions
This study is one of the first to explore how blockchain technology (BCT) can be used to enhance 
the functionality of last-mile agricultural AIV value chains in low- and middle-income countries like 
Kenya. It is also one of the first to investigate the potential impacts of the emerging technology on 
food and nutrition security. The results indicate the significant potential of BCT to improve 
transaction efficiency, create trust, connectivity, and coordination between value chain actors, and 
create better-matched markets for the highly nutritious AIVs. 
The main advantage of BCT is its ability to create an immutable and secure record of transactions. 
Providing value chain actors with access to easily interpretable and trustworthy transaction data was 
found to increase the functionality of the entire value chain, which in turn increased the quality of 
the final products and the price that consumers were willing to pay for AIVs. In addition, improving 
the connectivity between producers and traders significantly reduced the time spent trying to 
coordinate the sale of AIVs. These time savings provided actors in the value chain with opportunities 
to engage in additional productive or social activities. While BCT did enhance relationships between 
those actors using the AgUnity app, trust among value chain actors who were not using the app did 
not change. Thus, expanding the deployment of the app to additional actors along the AIV value 
chain has the potential to enhance the gains already realized. 
While producers accessed and leveraged the data provided by the AgUnity app, information sharing 
did not pick up among traders and retailers. This issue could be addressed by developing ways to 
share information in the app across the different groups of value chain actors – i.e., producers, 
traders, retailers, and consumers. Further, while the blockchain functionality did increase access to 
information for each user group, there was little improvement in the transfer of information from 
producers to consumers and consumers to producers throughout the study. This finding means the 
full gains from price premiums and matching supply and demand have yet to be realized.
Improving access to transaction data and introducing an AIV grading system enabled producers to 
focus on production practices that could generate higher incomes from Grade A vegetables. The
new grading system also encouraged groups in the value chain to seek training on how to improve 
the quality of their AIV production. This expressed demand for training presents an opportunity for 
extension agents and other agricultural service providers to create programs that can provide such 
support. An applet containing important production information could also be developed for the 
AgUnity app that producers could access when needed. Given that the AIV grading system has been 
adopted by actors along the value chain, an important next step is to introduce an external 
verification and monitoring mechanism, to provide consumers with confidence that the vegetables 
they purchase meet established quality and safety standards. 
The increased income from the sale of high-quality AIVs and improved production practices were
found to enhance food security among study participants by increasing the consumption of
vegetables and higher-quality proteins and fruits. However, the impacts were constrained since the 
70 | Page
AgUnity app currently does not provide value chain actors and consumers with information that 
could improve the marketability of AIVs. This study also identified several important gendered factors 
that should be considered when sharing information intended to influence the purchase, 
preparation, and consumption of AIVs. 
Finally, this study has identified a range are areas where additional research is needed to develop 
the transformation landscape for agriculture in Kenya and other SSA countries. First, since 
participants in this study were provided with a smartphone and were not asked to pay a subscription 
fee for the AgUnity app, an important next step will be to test different service and payment models. 
Study participants did express a willingness to pay a monthly fee to use the AgUnity app, but only if
additional complementary services and features were offered on the app. There is also a need to 
enhance several existing features of the app, such as improving its daylight scanning capabilities and
ensuring the app is more intuitive for traders. In addition to the AgUnity service fee, attention also
needs to be given to how value chain actors will purchase smartphones that can run the app.
Second, and in relation to the above point, new applets for the AgUnity app need to be developed 
that provide value chain actors with access to critical information and services and provide 
opportunities to expand the sale of AIVs. The hackathon undertaken as part of this study presents a 
unique model to engage youth in the development of these applets. Expanding the services provided 
by the AgUnity app may also create new opportunities to reengage youth in the agricultural value 
chain. 
Finally, future research should consider applying the AgUnity app to additional value chains, 
especially for those crops and products that have a direct connection to food security. 
71 | Page
References
Agnew, J., J. K. Mwangi, R.P. Hall, D. Sumner, and N. Kristofikova. 2021. Transaction and Information 
Pain Points in African Indigenous Vegetable Value Chains in Western Kenya: A Gender￾Responsive AIV Value Chain and Market Analysis Report. Virginia Tech (Blacksburg, VA).
Arndt, C., R. Benfica, F. Tarp, J. Thurlow, and R. Uaiene. 2010. "Biofuels, growth and poverty: a 
computable general equilibrium analysis for Mozambique." Environment and Development 
Economics 15 (1): 81-105.
De Ruyter de Wildt, M., M. Van Ginkel, K. Coppoolse, B. van Maarseveen, J. Walton, and G. 
Kruseman. 2019. Blockchain for Food: Making Sense of Technology and the Impact on 
Biofortified Seeds. CGIAR Platform for Big Data in Agriculture. 
https://cgspace.cgiar.org/bitstream/handle/10568/106615/%5b2%5d.CoP_SED_report_2019.
003_Blockchain_For_Food_version_1.pdf?sequence=1&isAllowed=y. 
FAO, IFAD, UNICEF, WFP, and WHO. 2020. The State of Food Security and Nutrition in the World 
2020: Transforming food systems for affordable healthy diets. FAO (Rome).
FAO, and Zhejiang University. 2021. "Digital Agriculture Forum: Highlights." Hangzhou, China, 5-6 
December 2020.
Gido, E.O., O.I. Ayuya, G. Owuor, and W. Bokelmann. 2016. "Consumer's choice of retail outlets for 
African indigenous vegetables: Empirical evidence among rural and urban households in 
Keny." Cogent Food & Agriculture 2: 1-14.
GoK. 2017. The Big Four: Immediate priorities and actions. Government of Kenya (GoK) (Nairobi). 
http://cn.invest.go.ke/wp-content/uploads/2018/12/Government-of-Kenya-Big-Four￾Plan.pdf. 
Henson, S., and J. Humphrey. 2015. Assessing the Effectiveness of Agri-food Value Chain 
Interventions Aimed at Enhancing Consumption of Nutritious Food by the Poor. Institute of 
Development Studies (Brighton, UK).
IPC. 2021. IPC Acute Food Insecurity and Acute Malnutrition Analysis: Kenya. FAO (Rome). 
https://reliefweb.int/report/kenya/kenya-ipc-acute-food-insecurity-analysis-and-acute￾malnutrition-analysis-july-2021. 
Kansiime, M.K., J. Ochieng, R. Kessy, D. Karanja, D. Romney, and V. Afari-Sefa. 2018. "Changing 
knowledge and perceptions of African indigenous vegetables: the role of community-based 
nutritional outreach." Development in Practice 28 (4): 480-493. https://doi.org/
https://doi.org/10.1080/09614524.2018.1449814. 
Kim, J., P. Shah, J.C. Gaskell, A. Prasann, and A. Luthra. 2020. Scaling Up Disruptive Agricultural 
Technologies in Africa. World Bank (Washington, DC).
Kristofikova, N., I. Muskoke, and J. Agnew. 2021. Exploring the Use of Blockchain Technology to 
Promote the Production and Consumption of African Indigenous Vegetables in Western 
Kenya: Deployment Report. AgUnity (Bali, Indonesia).
LASER PULSE. "Embedded Research Translation." Purdue University. 
Liu, W., X. Shao, C. Wu, and P. Qiao. 2021. "A systematic literature review on applications of 
information and communication technologies and blockchain technologies for precision 
agriculture development." Journal of Cleaner Production 298 (20). 
https://doi.org/https://doi.org/10.1016/j.jclepro.2021.126763. 
72 | Page
Mellor, J.W. 2017. Agricultural Development and Economic Transformation. Edited by C. 
Barrett.Palgrave Studies in Agricultural Economics and Food Policy. Ithaca, NY: Palgrave 
Macmillan.
Michels, M., V. Bonke, and O. Mußhoff. 2019. Understanding the adoption of crop protection 
smartphone apps: An application of the Unified Theory ofAcceptance and Use of Technology, 
. Georg-August-UniversitätGöttingen, Department für Agrarökonomie und Rurale 
Entwicklung (Göttingen).
Tripoli, M., and J. Schmidhuber. 2020. Emerging opportunities for the application of blockchain in the 
agri-food industry. FAO & ICTSD (Rome, Geneva).
USAID. 2020. Food Assistance Fact Sheet: Kenya. United States Agency for International 
Development (Washington, DC). 
https://www.usaid.gov/sites/default/files/documents/1866/FFP_Fact_Sheet_Kenya.pdf. 
WFP. "WFP: Kenya." WFP. https://www.wfp.org/countries/kenya. 
73 | Page
Appendix 1 – AIV Grading Tool
AIV Grading Tool
Criteria Assessment Levels Scores
How will you know if 
inorganic pesticides or 
organic pesticides have 
been used?
 
• Ask the 
producers
• Check for 
residues on the 
produce
No inorganic/organic pesticides 
used 
3 points
Some inorganic pesticides used 1 point
How will you know if 
inorganic fertilizers 
have been used?
• Ask producers
• Signs for fertilizer 
containers, etc. 
on the farm
No inorganic/organic fertilizers 
used
3 points
Some inorganic fertilizers used 1 point
Natural color (usually 
green for the leafy 
vegetables)
• Observe 
coloration
Natural green color 3 points
Light green color 2 points
Green-yellow color 1 point
Insect pest and disease 
injury
• Observe if free 
from damage
No pest/disease damage 3 points
10-30% pest/disease damage 2 points
Over 30% pest/disease damage 1 point
Mechanical 
damage/bruises
• Observe No mechanical damage / bruises 3 points
10-30% mechanical damage / 
bruises
2 points
Over 30% mechanical damage / 
bruises
1 point
Freshness • Observe Fresh/turgid 3 points
Partially wilted 2 points
Wilted 1 point
Grade A – 13 points or above (> 70%)
Grade B – Less than 13 points (< 70%)