Reducing food loss through sustainable business models and agricultural innovation systems

Purpose – This study aims to explore how sustainable business models (SBM) can be developed within agri-innovation systems (AIS) and emphasize an integration of the two with a systemic understanding for reducing food loss and value loss in postharvest agri-food supply chain. Design/methodology/approach – This study conducted longitudinal qualitative research in a developing country with food loss challenges in the postharvest supply chain. This study collected data through multiple rounds of ﬁ eldwork, interviews and focus groups over four years. Thematic analysis and “ sensemaking ” were used for inductive data analysis to generate rich contextual knowledge by drawing upon the lived realities of the agri-food supply chain actors. Findings – First, this study ﬁ nds that the value losses are varied in the supply chain, encompassing production value, intrinsic value, extrinsic value, market value, institutional value and future food value. This happens through two cumulative effects including multiplier losses, where losses in one model cascade into others, amplifying their impact and stacking losses, where the absence of data stacks or infrastructure pools hampers the realisation of food value. Thereafter, this study proposes four strategies for moving from the loss-incurring current business model to a networked SBM for mitigating losses. This emphasises the need to rede ﬁ ne ownership as stewardship, enable formal and informal bene ﬁ ciary identi ﬁ cation, strengthen value addition and build capacities for empowering communities to bene ﬁ t from networked SBM with AIS initiatives. Finally, this study puts forth ten propositions for future research in aligning AIS with networked SBM. Originality/value – This study contributes to understanding the interplay between AIS and SBM; emphasising the integration of the two to effectively address food loss challenges in the early stages of agri-food supply chains. The identi ﬁ ed strategies and research propositions provide implications for researchers and practitioners seeking to accelerate sustainable practices for reducing food loss and waste in agri-food supply chains.


Introduction
Food loss and waste in value chains present multi-faceted challenges associated with food security (Hamann, 2020), access to smart and resilient agriculture (Kim et al., 2021), reducing greenhouse gases emissions (Winans et al., 2020), enabling circular economy solutions (Dora et al., 2020), bringing transparency in food supply chains (Vasanthraj et al., 2023) and identifying ways to reduce postharvest losses (Mohan et al., 2023).Agricultural innovation systems (AIS) are relevant in this context, as it is about facilitating linkages and synergies among several actors for innovations within a food system.AIS refers to the network of actors and institutions in agricultural and related sectors that bring purposeful innovations into social and economic use (Klerkx and Begemann, 2020).However, operationalising the AIS needs a business model understanding that is networked and sustainability oriented (Neumeyer and Santos, 2018).There is an increasing need for actors at multiple levels in food supply chains to move towards sustainable business models (SBM) with food loss and waste reducing solutions and activities (Luo et al., 2022).Moreover, a networked SBM goes beyond the traditional product-and firm-centric functions and innovation, to a multi-actor configuration of systems-level innovation and action for creating economic, social and environmental value in the bottom-of-pyramid markets (Lüdeke-Freund and Dembek, 2017;Schoneveld, 2020).
The "business model" concept has been traditionally understood as a set of firm-level activities for value proposition, creation, capture and delivery (Zott et al., 2011).Recent research articulates the concept of "business model" as a boundary spanning from being firm-centric to systemic (Klang et al., 2014).The systemic approach, which examines the whole in terms of its contextual complexity in contrast to the linear approach, attempts at creating synergies between actors while implementing innovation systems for systemic change (Midgley and Lindhult, 2021).In a networked understanding of SBMs with sustainable innovation, this would mean enabling multiple actors to re-organise and re-align themselves with increased interconnectedness in dealing with sustainability challenges (Zucchella and Previtali, 2019).The networked SBM moves beyond product-centric solutions to a systemic understanding of implementing innovation for sustainability (Schaltegger et al., 2016;Lüdeke-Freund and Dembek, 2017).This requires re-organising the business model through technical innovation, organisational innovation, institutional innovation and social innovation, with a sustainability agenda (Klerkx and Begemann, 2020).This posits the need for exploring how to move from a non-sustainable business model that creates food loss and value loss at early stages of the supply chain, to a networked SBM operationalised with AIS for mitigating the losses.
The Indian agri-food value chain presents us with a distinct context to study the links between AIS and networked SBM for reducing food loss along supply chains, particularly postharvest losses as emphasised in Sustainable Development Goal (SDG) 12.3.India is the world's second-largest producer of agri-food and has the largest bases of small-holder producers operating with other informal and formal actors, allowing us to study their food loss business models as these are underrepresented in AIS literature.The Indian Government has started aggregating smallholders as farmer producer organisations (FPOs), which allows us to compare and contrast a network perspective of business models to individual actor-specific business models with respect to implementing AIS for food loss reduction.By examining this context, we aim to contribute to the broader discourse on reducing food loss and waste through sustainable innovation, by unearthing the conceptual elements that link AIS and a networked SBM.For that, we explore empirically how the agri-food value chain actors in our research setting of a low-and middle-income country, interpret the role of business models and innovative practices in reducing the postharvest loss of fresh produce at early stages in the supply chain.
We address the overarching research aim with three "how" research questions (RQs) related to the agri-produce supply chain at the postharvest level: RQ1.How does the current business model lead to postharvest food loss and value loss?
RQ2.How do the experiences of the value chain actors reveal the strategies needed for transitioning to a networked SBM?
RQ3.How can these strategies contribute to creating a networked SBM within AIS?
A qualitative research study was conducted over four and half years through focus groups, fieldwork and interviews, for inductively generating in-depth knowledge from the lived experiences of the study participants from agri-food supply chain in an emerging economy context.The subsequent sections first highlight the key literature background of this study.Then the methodology section elaborates the qualitative research design, followed by the findings.The findings are organised in such a way that Sections 4.1-4.3provide answers to RQs 1-3 consecutively.The article ends with the limitations of this study and directions for future research.

Agricultural innovation systems
One-third of all the food grown globally is lost within value chains.While most research focuses on retail-and householdlevel food waste in developed countries, a significant portion of food losses occur between farmgate and retail (Dora et al., 2020)."Food loss" refers to the gradual loss in the value of food produce (nutritional and economic value) as it moves from upstream to downstream, losses due to lack of second life value chains for food surpluses and losses due to lack of circular systems and end of life management, rather than an absolute physical loss of the product (Centobelli et al., 2021).To address these losses in an informally and formally networked, institutionally complex and uncertainty-prone (economic, climatic, political) environment, an AIS view appreciates the interconnectedness of multi-actor action and innovation for capturing value that can be translated to waste or food loss reduction (Klerkx and Begemann, 2020).AIS is "a network of organisations, enterprises, and individuals focused on bringing new products, new processes, and new forms of organisations into social and economic use, together with the institutions and policies that affect their innovative behaviour and performance" (Hall et al., 2006).From the AIS perspective, innovation could be technical, such as smart agriculture; organisational, such as creating communities of practice; institutional, such as new policies and business models; social, such as revising livelihood models and strengthening self-help groups (Watkins et al., 2015).For open innovations, such as AIS, to work within supply chains, recent research suggests that ambidextrous capabilities (purpose, span and orientation) are required to co-exist within the same framework to cater to values of a variety of stakeholders (Solaimani and van der Veen, 2022).However, most of the deployment of AIS in the literature would focus on one capability or the other.For instance, in terms of purpose, some focus on exploiting knowledge regarding AIS scalability (Cronin et al., 2022) while others explored IoT deployment for operationalising AIS (Parmar and Kumar, 2022) depending on whether they are working on top-down or bottom-up research.
Recent literature on the application of AIS to deliver values for multiple actors found that there are two major forms of failures that can occur in complex innovation systems: Multiplier failurewhere the failure of innovations in one business model cascades into failures in other interconnected business models (Cronin et al., 2022;Fieldsend et al., 2022) and Stacking failurewhere failures in infrastructure or data innovations in a business model within a value chain undermines the entire AIS (Cronin et al., 2022;Parmar and Kumar, 2022).Although multiplier failures cascade through innovation systems from a critical failure point to the other, stacking failures undermine the entire system because of failures in critical nodes related to data or infrastructure requirements.Previous research has examined how these failures occur in AIS (Brewster et al., 2017;Blash et al., 2020); however, the implications on SBM are underexplored.Given that AIS goes beyond digital transformation within a single part of the agricultural system and instead involves coordinated innovations across various actors in the system, it becomes clear that failures in AIS and their impact on networked SBMs are closely intertwined and cannot be separated (Lerman et al., 2022).
We argue that actualizing this requires a transition from the loss-incurring current business model to a sustainable business model (SBM) with renewed actor relationships, resource constitution, exchange of knowledge artefacts and value addition through co-creation of innovation.While SBM research has grown by developing typologies (e.g.green supply chain, circular supply chain), it has scarcely examined the context of post-harvest food loss and how it is intertwined with AIS for farm-to-retail supply chain actors to reduce food loss and create additional value.

Sustainable business model
Business models are understood differently in different industries.However, in general, a "business model" consists of actors and actions that propose, create, deliver and capture value (Preghenella and Battistella, 2021).At the firm level, a business model can be a set of attributes and inter-related activities, an operational design or schema or a formal blueprint of doing business (Massa et al., 2017).The firm has value exchanges (e.g.resource streams, revenue streams and cost structures), value-adding activities (e.g. operations, processes, product/service offerings) and value objects (e.g.delivery channels, communication channels, logistical streams, infrastructure) (M endez-Le on et al., 2022).In sustainability transitions, particularly for the bottom-of-pyramid markets, a "sustainable business model" is a business model that goes beyond traditional product-and firm-centric innovation to a multi-actor configuration encompassing interconnected relationships and resources for creation and embeddedness of systems innovation that contributes to environmental, social and economic value creation (Lüdeke-Freund and Dembek, 2017;Schoneveld, 2020).
The networked understanding of SBMs makes a shift from a hub and spoke model (single actor with multiple suppliers and buyers) to a multi-actor model (innovation network of small to large actors cooperating and coordinating on a common goal) (Hansen and Schmitt, 2021).In the context of agri-food value chains, it means that actor groups (such as farmers, transporters, distributors, processors and retailers) organise input flows, agri-business and supply chain capabilities and information systems to create value-added outputs (Barth et al., 2017).Operationalising this is challenging because of reasons such as complex contextual dynamics, different value creation activities by different actors, resource efficiency needs in physical systems and compliance practices as per institutional directives (Best et al., 2022).
Studies on SBM in agri-food value chains have focused on sustainable innovation through operational optimisation, organisational transformation or system building (Nosratabadi et al., 2019), but these mean differently to different actors.For instance, operational optimisation means resource efficiency to some and energy efficiency to others.Organisational transformation means functional improvements to some and changing of ownership structures to others (Ulvenblad et al., 2019).Some actors view system building as creating new value streams from waste while others see it as harnessing new value from alternative supply chains (Bocken et al., 2014).Recent research reviewed 15 food supply chain models and identified similar gaps in literature with regard to some value streams not being captured within the core business model by the key actors in agri-food supply chains (Corallo et al., 2023).These value streams (e.g.food loss/waste, disposal) are not owned by any one stakeholder.They proposed future studies should look into the roles of external stakeholders in enabling SBMs to capture the values.These external stakeholders form the links between the core business models in a network.
The networked perspective of SBM emphasises shared value creation to include different actor interpretations of agri-food supply chains owing to socio-economic conditions (e.g.livelihood, ownership, beneficiaries, stewardship) and institutional arrangements (e.g.regulations, finance, insurance), thereby enabling resilience and sustainability outcomes (Acquier et al., 2019).Hence, research on networked agri-food SBMs needs to explore the "value intention" of the actors, as contextual dynamics tacitly affect actor-led actions in a value network (Barth et al., 2017).Therefore, certain strategies and mechanisms are needed as per context to act as enablers for a networked agri-food SBM, capturing the value intentions of the local actors and making a shift from the norm of top-down business model principles (Nosratabadi et al., 2019).

Research gap
Configuring a networked SBM within AIS in farm to retail agrifood supply chain needs to draw upon actor-interdependencies in roles and responsibilities, actor interpretations of value chain and institutional practices and actor needs for collaborative arrangements.This is particularly crucial when the operating context is a low-income setting with formal and informal economies, resource scarcity, institutional complexity, market and climatic uncertainties and several regional and local supply chain operators interpreting value in many different ways within the network.Hence, our findings make a major contribution to how to make a transition to a networked SBM with AIS for reducing postharvest loss of fresh produce in a developing economy context (Corallo et al., 2023).Despite the significance of the problem in vast and emerging economies such as India, it has received insufficient attention not only in AIS, SBM and food supply chain literature but also in management and business research in general.Among the developing economies, India offers one of the most complex contextual settings for this problem (Guha, 2016).Contextual understanding through exploratory work matters in food loss and waste research when the context is characterised by populations and places that are difficult to reach and suffer from harsh socio-economic and climatic conditions (Filimonau and Ermolaev, 2021).
Loss of food and its value at post-harvest level is a major challenge in low-and middle-income countries, and India offers a significant research setting to understand that problem.Globally, 25% of arable lands are smallholder farms in countries which produce nearly 30% of the food consumed worldwide (Ritchie, 2021).Smallholder farmers and other upstream supply chain actors in India are affected by postharvest food loss on an everyday basis (Kaza et al., 2018).Artiuch and Kornstein (2012) reported that 20% to 40% of the food grown in India was spoiled before it reached customers.Nearly a decade later, vegetables comprised 25% of the total weight of food loss and waste worldwide, with India losing 31% of its vegetable produce (Chen et al., 2020).WWF-UK (2021) reported that in South and Southeast Asian countries, the economic loss because of food loss at the farm stage was estimated to be $15bn per year, in which India was among those with the highest losses.Challenges include long-distance transportations in a variety of climatic conditions, need for more affordable temperature and humidity-controlled logistics, replacing open-air dumping or impoverished storage infrastructure with affordable and replicable cold chain infrastructure for vegetables and greens.This is alarming for a country which is the second-largest global producer of fresh agri-food (Sagi and Gokarn, 2022;Mohan et al., 2023).However, very little research has been produced on preventing or reducing food loss and value loss across the supply chain in developing countries with thriving agrarian production, but with institutional and infrastructural complexities that are different from more industrialised countries (Dora et al., 2020).

Theory
We use "sensemaking" as the theoretical lens to draw upon the experiences and practices of individual actors in the agri-food supply chain.Sensemaking is an intentional act of comprehending and explaining past and present practices, while also anticipating potential futures (Gephart et al., 2010, Vanderlinden et al., 2020;Moqaddamerad and Tapinos, 2022).This activity enables comprehending an ongoing complexity, explicitly in the form of how people interpret and give meaning to their action and their context, thereby generating shared meanings that create grounds for new action (Weick, 1995;Ancona, 2012).Epistemologically, this perspective views actors as knowledgeable agents (Gioia et al., 1994;Gehman et al., 2018) capable of retrospective sensemaking (meaningfully constructing everyday activities, gains and challenges based on past and present experiences) and prospective sensemaking (meaningfully constructing a potential future that is persuasive and necessary).The actors implicate the challenging and complex context that they experience (Brown et al., 2015;Weber and Glynn, 2006) through this retrospective and future-oriented sensemaking (Moqaddamerad and Tapinos, 2022), which contributes to building in-depth knowledge on actor-level mechanisms that would trigger new patterns, linkages or opportunities, which would stimulate a systemic change for addressing a sustainability challenge (Köhler et al., 2019).
This constructionist approach (Gehman et al., 2018) views actors as agential (capacity to bring change) through collaboration and cooperation (Emirbayer, 1997;Vanderlinden et al., 2020), as the context is shaped through their interpretations and actions (Sengupta and Lehtimäki, 2022).They are not powerless beings but individuals with the capacity and propensity to create a sustainable future by playing an active role (Teerikangas et al., 2021).As an agri-food system has multiple actors deploying different business models, there is a need to understand how these actors relate to each other in the networked business model.Incorporating the shared sensemaking of different actors in the supply chain around particular issues or challenges brings forth different constructions of the business model, rather than taking to them the idea of a business model that comes from the most dominant actors (Hayden et al., 2021).When this is overlooked, top-down solutions for sustainability risk failing in implementation.Thus, sustainable futures can emerge through strategies constructed from the retrospective and prospective sensemaking of a range of actors who are expected to work together in a coordinated way, even though they pursue different interests within the same supply chain (Markard et al., 2012).

Data collection and analysis
We adopted a multi-method qualitative research design suitable for creating new knowledge on nascent topics (Ketokivi and Choi, 2014).We used multiple sources of primary data through focus groups, fieldwork and interviews (Table 1), mostly in the natural settings of the study participants (Lincoln and Guba, 1985).Secondary data sources included sectoral reports.Primary data was periodically collected between June 2018 and December 2022 as part of a larger project, through 18 focus groups, 43 unstructured and 15 semi-structured interviews along with 98 days of fieldwork to gather observation data.The interviews lasted between 45 and 90 mins and focus group discussions lasted between 90 and 120 mins.The research setting being institutionally complex with a range of small to medium-scale actors with different roles and challenges in formal-informal settings, elaborate data collection enabled data sufficiency with inclusion of diverse voices, triangulation through different data collection methods and theoretical saturation with no additional strategies/ mechanisms and propositions emerging during the analysis (Andrade, 2009).
A purposive selection (Patton, 1990) was conducted in this study: identifying value chain actors with long experiences of working in the agri-food supply chain with multiple stakeholders (both public and private) at the local and regional levels (see Table 1).Some were identified during fieldwork and others were reached through snowballing."What" questions were used to initiate conversations while interviewing, such as: What, in your experience, are the pain points of various stakeholders in the current post-harvest value chain of fresh produce?How do you manage your own operations?What Focus group questions enabled more information on the value chain activities, exploring challenges leading to food losses in current business models and identifying the root causes of those challenges by asking: Who are the actors involved just after production and during distribution, infrastructural and logistical challenges and complexities, the roles of supporting organisations: government and non-governmental organisations (NGOs), food loss and value loss at each node in different seasons and who bears the loss and identifying root causes of all challenges and possible solutions.A second round of interviews was conducted with a few participants, as and when required.
The interviews and focus group discussions were transcribed verbatim.The inductive analysis of the transcriptions followed a coding technique for identifying thematic patterns and categories (Table 2) (Braun and Clarke, 2006).The first-level themes were broadly informant-centric, which were aggregated for thematic similarities to construct the second-order themes.The third-level synthesis created conceptual categories through a constant comparison between data, first-level and secondlevel themes.The interviewing and thematic analysis stopped on reaching theoretical saturation (Eisenhardt and Graebner, 2007).Further, a back-and-forth analysis spotting the interplays between data, themes and literature to arrive at a set of propositions, enabled us to achieve more rigour in the qualitative analysis (Grodal et al., 2021).
The observation and field interaction data also included information on key supply chain actors and their activities, their experiences with food and value loss at different nodes, storage and transportation conditions, grading and sorting practices, activities and everyday records at the collection and distribution centres and information on key resources and governmental support.Copious notes were made from observations and interactions.Close to a hundred photographs were taken as visual data to depict the context which aided researchers' discussions.Reflexive note-making and consolidation through discussions within research teams happened at the end of each day of the fieldwork.

Current business model
Our findings on the current business model for food loss and how this relates to AIS failure are detailed in Table 3.This section discusses the key internal and external actors, their roles and responsibilities, the challenges leading to food loss and the different forms of food value loss as a result.

Production function and actors in the current business model for food loss management
Farmers are key decision-makers in food production and their role in the current business for food loss management pertains to generating and utilising relevant data for planning and aligning food production to market requirements.The farmers, who were interviewed, explained that the loss in food value in the current business model includes pre-production losses (e.g.weather, seasonality), production losses (e.g.input, pest) and post-harvest losses (e.g.distress selling of food surpluses and non-market grade produce).Payment delay coupled with poor management of pre-production and production losses lead farmers "to depend on borrowing from CAs (Commission Agents), money lenders and local cooperatives and incur heavy losses in produce value to maintain the cash-to-cash cycles", FN2.As smallholders require collective handholding to develop standard operating procedures (SOPs) and share best practices for food loss management, farmers lose food value through distress selling and getting locked-in to unfavourable forward contracts with market agents to secure loans and hedge against market price fluctuations.
The focal farmgate actors that play an aggregation role in the current food loss business model are FPOs.FPOs comprise farmers, landowners, producer groups and other beneficiary collectives tasked with pooling production capacity to improve farmers' price bargaining power and access to markets as well as facilitate handholding to link farmer beneficiary groups to finance, infrastructure, capacity building and other AIS innovation.FN1 explained that: Many large cold storage investments are concentrated in some regions [Uttar Pradesh, Gujarat, West Bengal and Punjab] and are designed to the temperature and humidity requirements of specific products [potatoes]. . . the same applies to investment in cold chain transport.
FN2 further explained that most smallholders are not represented by FPOs; hence, food losses in the current business model between harvest and the time to markets: arise mainly from poor access to market and price information. . .lack of ownership of simple cold chain solutions needed to preserve rural production. . .reduce distress selling and justify large or multiple mobile small cold chain infrastructure investment by FPOs.

Logistics and processing function in the current business model for food loss management
From focus group discussions with private sector actors in the current business model, we identified logistics providers and processors (comprising collection centres and distribution centres) as the linking nodes whose primary functions are to minimise losses arising from "extrinsic" operations associated with product logistics scheduling, planning and routing as well as losses linked to "intrinsic" operations of sorting, grading and handling in collection and distribution centres.Their role in the current model is extrinsic, implying that "food losses occurring during logistics are not owned by the first party logistics [local truck owners] that carry farmers to local markets and collection centres" [FU2].As a result: The time value of food lost in the process is reflected in the amount for food that is downgraded from A to B and C grades due to logistics delays and poor scheduling and route planning to markets [CC7].
In addition, farmers carry out their own grading and sorting in line with market requirements.CC2 noted that "the loss in (extrinsic) value must be reassessed through additional grading and sorting" to re-evaluate the quality of deliveries.Moreover, the intrinsic operations required to sort and grade food produce are unstandardised, hence, the repetitive function results in losses associated with repetitive intrinsic quality assurance processes carried out on farms, collection centres and

Conceptual categories
There is always a debate whether we should take care of requirement of farmers, which require higher prices for their produce or we see the worries of consumer who requires procuring at very low cost, so the government has to play in between these two.I think lots of initiatives are being taken by government specifically, and the farmers are leveraging.[.
. .]The ease of doing business that government is promoting, I think that this is a very good approach to the ease of doing business so that farmers can easily avail land where they can put necessary infrastructure, resources, and they can easily avail credit (TP 4)

Increasing governmental role in multi-actor cooperation
Redefining ownership towards stewardship

Ownership configuration
Government should be consulting us, our community for where we want the cold storages [.
. .] whether for this produce we want at panchayat level, for that produce we want at district level, what size, capacity do we need [.
. .] and then provide us with a solution that reduces our loss (FU3) Farmer intake is more in a government facility.That is major advantage.[.
. .]If it is government, they feel it is theirs.The farmer intake will also be more.[. ..]They feel it is govt project (DC 6) Organizations like FPOs can become the last mile connectors for delivery of solutions and gateway to the market.These FPOs can play the necessary role.But it needs investments into human resources, infrastructure or financial access (FPO 7)

FPOs as frontline for institutional push
FPOs aggregate smallholders by beneficiary groups . ... to link smallholders to different markets (FF1) The capacity should be built at the FPO level [.
. .] making it a mass solution or offering for wider public (TP 4) But some kind of aggregation is needed so that cost of production can be reduced and maybe a model cluster should emerge, so that they can absorb these costs.The Federation based model can employ some good professionals they are involved in the infrastructure creation and aggregating FPOs, because all these require infrastructure and investment done collectively as per policy directions, thus making a different kind of business model (TP 4)

Collectivisation at local level through multi-actor aggregation
Mobilising stewardship Government has aggregated farmers and other small processors as FPOs [.
. .]We have been actively involved in establishing market linkages for these FPOs, facilitating access to credit and finance, and promoting technology adoption for efficient farming practices [.

. .]
empowering farmers, improving their livelihoods, and enhancing overall agricultural productivity (FN2) Let's say certain set of farmers come and they put up a farmer market cluster on certain days in certain slots in certain societies and we use them.Right now, if you see such things are happening and that is good.If we say there is a solution which is now made available for a local community or a set of farmers in one particular cluster, that makes a lot of sense because they can actually build up or begin put up a capacity out there, and all of the supply chain can be aligned with that (R 11) The organized players are certainly in a better position.They can come together and put up capacities to manage new solutions, then there would be more takers.[.
. .]Whoever has a bigger stake in this can certainly be the one who can be considered as the first stakeholder where we should provide the solutions (TP 2) Organised players sharing responsibility for solutions Retailers have a lot of potential to be powerful advocates for loss prevention and sustainable food systems.They can source responsibly, manage their inventory, educate consumer, educate farmers how their farming practices impact on food value (FN1) A model would be the involvement of different multi stakeholder players to come together within policies promoted by government.Some state governments are very active, they are engaging various stakeholders, including private players, financials organizations, have number of state level schemes, specifically with horticulture department.They can leverage these ones and subsidise for farmer production clusters, so that the necessary investment is done (FPO 7)

(continued)
Sustainable business models Subhanjan Sengupta et al.

Supply Chain Management: An International Journal
Table 2 Representative data

First-level abstraction
Second-level abstraction

Conceptual categories
In business model, each node does not need to generate revenue.But what data does the node generate and how that can be used in the revenue model is important.There are certain things where public investment should come in, such as for data that will go for government decision making.[.
. .]You are not selling the data.The data which shows.
. .this intervention was used and this is the result.
That kind of data is valuable data.If you can show a trend, there are certain things where public investment should come (FPO 7)

Collaborative planning for information feedback
Improving farm-level planning

Value Addition
Access to both demand side and supply side information is crucial in our efforts to facilitate food loss reduction for farmers.By understanding market demand trends and aligning them with how and when farmers should sow and harvest [.
. .] we can bridge the gap, minimise waste and maximise value (FN2) I believe how much needs to be produced if that data and dashboard information is sent back to the farmer, then it will be more meaningful so that the storage can be planned well ahead and waste is less.
. .so the feedback is missing now (TP 2) If you are producing something, you should be aware of what is going out of your farm.If there is a rotten material, you got to throw it away, or you're going to compost it to make up some organic farmyard out of it, the rest will go.So doing that sorting at the farm is the first necessity, which is based on the precondition that these farmer co-operatives and FPOs are skilled, adept and good in doing that (NGO 3) Sorting and storing at farm level [.
. .] absence of storage facilities near production clusters is problematic as this leads to losses during transportation, especially for perishables.There is an urgent need for better planning and investment to ensure that storage and sorting infrastructure is appropriately positioned to serve the needs of farmers, collection centres, distributors, retailers, etc.
There are a few farmers who got convinced with grading and sorting at the farm level and then having different grades of the produce.But these particular changes are still limited to those who are convinced.[.
. .]More awareness is needed on this particular activity, as far as the supply chain is concerned, because it actually carries a garbage or it carries a waste which actually can be consumed at the farm level (TP 1) For certain product, shelf life maybe 4 to7 hours, but then there is transportation loss.We can may be reduce by putting those commodities in the storage for maybe half a day or full day so it gets a very good valuation, because we are reducing the wastage.
Large wastage happens in transportation (TP 4) Improving preand post-harvest procedures Aggregating quality management in upstream we believe that seed selection is also linked with the losses farmers have during and after harvest [.
. .]Each seed variety interacts uniquely with the soil, climate and other environmental factors.When we manage these linkages effectively, we can minimize crop failures, reduce pest and disease incidences and improve overall farm productivity (FPO 5) A value addition will be how quickly you can transport, without a delay and if it can be sold at retailer point or directly to home.[.
. .]The loss of transportation is another thing.The time taken in transportation that can also be reduced.
So what I'm saying is what is it that needs to be done, as far as the the price volatility is concerned.Now there is a farmer who had harvested, or who is about to harvest.[. ..]Suddenly there is a change in the market prices and things are not going right.He should have an option of basically storing it for a while, or something like that.From that perspective, I was talking about short term storage (FPO 7)

Building local level capacities for maintaining food value
We do not have access to infrastructure like shades, louvers on farm [.
. .] we do not get loans on these so we cannot invest in such infrastructure by ourselves but this would be life changing for us [.
. .] because generally the products lie in the farms for 2-4 hours after harvesting [. ..] sometimes there are issues when the labour is available and sometimes when the tempo comes to pick these up [.

. .]
and if this is peak summer when the temperature is 40-45 C then we know that our produce would have started deterioting and have started losing significant value before even it leaves our farm (FU 4) There is always a scenario where there is no zero sum game.So you are left with some quantities.And those left out quantities get sold over a period.Probably a day or two and the chances of that which is not sold in time is the thing which is going to get stored.So, it is going to be a residual type of produce, which is left out with you and that needs to be now taken care of in the best possible way.
Otherwise, it increases the cost, it also increases the inefficiencies and I would say it also multiplies the food waste (TP 1)

(continued)
Sustainable business models Subhanjan Sengupta et al.

Supply Chain Management: An International Journal
Table 2 Representative data

First-level abstraction
Second-level abstraction

Conceptual categories
So, if there a staggered production, so the entire production is not coming on time.Based on this stagger in the sowing period, the harvest is coming in different time periods, so they need some buffer time in hand to market the produce (TP 4) Reducing gap between harvesting time and pick-up time

Strengthening producermarket and market-market connections
We are encouraging staggered sowing and staggered harvesting through our SHGs so that not all of them reach the market with the same produce as that causes a dip in price and then everybody losses (FG2) In order to ensure that there is limited loss of quality and perishability, the order needs to be taken in not much in advance because of the necessary logistics, which again, is a big challenge to quantify (TP 1) So right from the farm to enterprises, there are a lot of aggregators and there are a lot of people who have come up with different type of solutions, right from the whole chain solutions to even the storage solutions.There are some village level agents, or some farmers, who are slightly entrepreneurial.What they do is that they pick up material from their own farm and some fellow farmers and drive it down to the urban place and they open up a shop which is a mobile shop right on the vehicle only and they sell from the vehicle only (R 11)

Entrepreneurial innovation in linking producers to nearest markets
Beside strengthening producer-market relationships which is really important, we should also look into strengthening market-market linkages.
. .. there are many formal organised retailers whose grade B and C can be used by other informal retailers, push carts or even local roadside hotels, dhabas (FN1) Care has to be taken for connecting both farmers and buyers together.[.
. .]We worked on a concept called veggie cart.That was focused on direct B2B and B2C.We would basically take them out and send it to low-income parts of the city.We would basically organize the low-income vendors in those areas so that they can actually be more aligned to sell those kinds of vegetables (NGO 3) There is organized retail market, and then there is unorganized retail market -push carts, weekly local mandis, petty shops, these three together.What are the different things impacting them?The losses in the mandi is huge.At the retail end there is a loss.So what can help all of them?Till date, no improvement has been done for the push carts (NGO 1)

Identifying informal sector beneficiary Beneficiary Identification
In our region, around 60-70% produce is sold by push carts or local petty shops [.
. .] you cannot understand loss without understanding that segment.These have to be identified and mapped to decide which interventions are best for them to reduce their losses, else the best interventions would be missing out of the largest segment of the market actors responsible for value that is lost (FN2) So, retailers, I can see that, who certainly benefit from them as retailers who are typically the smaller size or the midsize one because the bigger one already has pending and design their retail outlets with air conditioning and temperature control and all those things (TP 1) It is mainly aggregation.Beyond the grading and sorting there is no value addition.Mostly it is fresh veggies that are sold.That case maximum of the price should go to the farmer.The product cycle is also very low -2 to 3 days.Maximum share of consumer price should go to the farmer.But that is not happening (NGO 1)

Farmers as beneficiary
Ideally farmers should be part of the value chain and gain most of the premium if they are doing farming practices that benefits retailers and their labelling, say organic or natural products (FN 3) Immediate gain actually happens for the people who are the organizations, or they are middle men, or traders, or people controlling the mandis.If we come to policy making part of it, many state governments are trying to remove the middleman and mandi concept.It is very tough.Farmer and market yards will have benefit in long term.In short term it is those people who are investing in it (TP 2) Then there should be some commission of some part of that cake should go to meet the cost of the infrastructure (TP 3) Sustainable business models Subhanjan Sengupta et al.

Supply Chain Management: An International Journal
Table 2 Representative data

First-level abstraction
Second-level abstraction

Conceptual categories
Infrastructure cost bearing entities

Identifying formal sector beneficiary
Our company (retailer) has invested heavily in grading, sorting facilities here and it makes our life easier to quickly complete these tasks which use to take a lot of time and people (DC1) We should be able to bring sprinkled temperature controls under one funding infrastructure so that it can be run using a service oriented model where maybe farmers who needs access to that can do it through the farmer producer cooperative (NGO 3) It is integral to install something which was allows a buyer who is willing to pay a premium price or a better price, which cuts off the intermediaries so that there could be a direct transaction between the farmer and the farm gate and the off-taker (TP 3)

Upstream functionary as beneficiary
Aggregating farmers and producers as FPOs using right criteria such as farmers livelihood is crucial for capturing right values that contribute to food loss reduction [.
. .]FPOs are the first formal level of small holders so including them as beneficiaries allows us to capture and standardise local knowledge about food that can contribute to reducing food losses [.
. .] to dessiminate local innovations around food loss reduction (FPO 6) The person who is buying the produce and another guy who is quality input supplier, these are the two critical ones other than the farmers, who are basically investing into the whole supply chain.
. .Right?The farmer is a seller trying to sell it is as fast as possible as they don't want to keep it into custody.Assuming that the intervention is going to increase the shelf life right, then it is the aggregator and the retail side.
. .there's basically a responsibility of the stakeholders who are there in the supply chain (TP 1) Strengthening the people, community institutions, so that they can handle their businesses and their problems at grass root level, is significantly very important.[.
. .]There is another way to look at it.That is how we improvise unorganized sector.Building their skills, organizing them in a better way; 70-80 percent of the market is unorganized.How we organize that also is important rather than displace that and replacing it with current modern retail (NGO 1)

Community capacitybuilding
Small holders, whether they are farmers or processors, they all need practical assistance and guidance tailored to their specific businesses [.
. .] training programs, workshops, mentorship initiatives are all needed for empowering them with the necessary skills and knowledge for organic farming, processing, packaging for getting premium value from that market [.
. .] can empower farmers to effectively engage with various organisations and access the resources they require for sustainable agricultural practices (FPO 4) To understand the requirement on the ground level, another element is to build the capacity and handling mentoring, but more from a block level kind of a incubation centre approach [. ..] that could be a game changer because we then demonstrate a few things and others can of course replicate (NGO 2) The only way farmers will start doing this is if you keep on providing them necessary tools, bear the cost of certification, which is for three years, and provide them the access to key markets and a premium every other time.[.
. .]So doing that at the farm gate is the first necessity, which is based on the precondition that these farmer co-operatives and FPOs that are being formed should be skilled, adept and good in doing these things (NGO 3)

Handholding for accessing institutional support
Currently, farmers are expected to connect with over 20-25 different organizations to access the information they need [.
. .] assuming they even know where to find them and what services they offer.This fragmented approach creates immense challenges for farmers and hampers their ability to make informed decisions [.
. .]As NGOs, it is our responsibility to provide handholding capacity building to farmers, particularly in areas such as insurance, finance and access to government programs [.
. .] but then government should also recognise the importance of this handholding responsibility and enable us with resources, policies and linkages (FN 3) When we are ambitious enough to let's say put up cold storages at micro level and do certain innovations there.But people at that level need proper hand holding or support to come to the nearest bank and submit a proposal which would be processed.[.Digital platform is enabled with three very critical things.One is the sowing date, the planting date [.
. .] forecasting a planting date for a farmer, depending upon multiple variables that may affect the right sowing and seeding of the plant.Second is we know the seed rate and the germination rate, depending upon the recommended seed variety and the seed brand, the farmer purchases from a marketplace only.And that allows us to determine what will be the plausible harvest dates (NGO 1)

Digitalising access to farm data
Farmers are blind folded when it comes to market demand, I mean future market demand, 3-6 months down the line [.
. .] when they know this, they would plan accordingly, we can then aggregate this and plan for capacity and shift to different markets [.
. .] some of the bigger FPOs who are working with 2-3 products, they have done this successfully but with low cost vegetables and staples it is difficult (FPO 5) I believe how much is to be produced if that data and dashboard information is sent back to the farmer, then it will be more meaningful so that the storage can be planned a little well and waste is less.
. .so the feedback is missing now (DC 6) Best practice is early hours or late hours.But farmers are doing after 9 to 10 o'clock because labour available from 9.30.Obviously, the shelf life will lose 30% because of the temperature.When you are harvesting the tomato, you are harvesting at 35 degrees.From then onward degradation starts.They will heap in one place, that is also not in shade.So inside temperature will increase.In evening hours they will collect the material and send to collection centre or the market.From there we need to collect the requirements.From there to here it takes 4 hours if it in near, if far, it takes 12 hours (DC 7)

Predicting sowing and harvesting time
There's material flowing from farm to market [. ..] wish there was information flow from market to farm telling us when to sow, when to harvest so that we can plan production scheduling and aggregating products accordingly for different markets and different seasons (FPO 6)  Representative data

First-level abstraction
Second-level abstraction

Conceptual categories
We need (community level needs assessment for infrastructure) storages for staples, fisheries, high value products at district level and for others with lower shelf life at panchayat level (FU 4) So, one that's essentially the concept that I'm trying to build with them is how might we be able to build a decentralised network of offgrid modular, structured, uh, called cold.And it was very controlled systems so that the farmers would have an opportunity to look at high value crops, such as tamarind or turmeric, or, uh, more, uh, finger millets and others, which requires not only cooling, but also certain kinds of moisture control (NGO 3) Also when you look at when processing steps in, storage need not be only about fresh produce, but also pulp.The kind of intrinsic value and storage value if you calculate for that one, where pulp can be used (TP 4)

Increasing value with processing opportunities
We need more information and training on packaging [.
. .] because of weather, logistics, multiple grading processes, packaging is really important for preserving the grade and nutritional value of our products [.
. .] gap at our community level (FU 5) Very few FPOs at dealing with processed product [.There is a lot of market food losses because farmers have to make uninformed decisions to sell their products to commission agents of APMC through open auctions or through informal auctions organised by market yard traders' associations during a short time window in early mornings when prices are set by auctioneers and bidders with informal market rules.
Upon arriving at local markets or regulated market yards, farmers must initiate sales by offering their produce for sorting, grading and packaging and take whatever price is offered by the agents in good faith, because they depend on these agents for advance finance.Moreover, FN3 explained that farmers are forced: to sell mixtures of product grades and qualities that have been poorly aggregated at minimum acceptable prices [. ..] to avoid lengthy and complicated bargaining and pricing processes but they incur significant market value losses in the process.
Regarding formal wholesale markets dealing with the bulk purchases through formal channels such as retail and hospitality are more amenable to regulations, but a majority of transactions are still conducted among market intermediaries and middlemen (agents and brokers).Furthermore, we found that for most formal actors, relying on informal market prices that are prone to fluctuations results in purchase contracts to hedge against fluctuations in wholesale market prices, which leads to surplus purchases.In this regard, the government has set up a National Agriculture Market Electronic Trading (e-Nam) platform to provide market access and improve price discovery for farmers, but FN3, cautioned that: such systems can only work if the FPOs themselves are grouped into collectives to capture the needs of most underrepresented beneficiary groups that bear most of the losses in food value from both production processing and market inefficiencies.

Government function in the current business model for food loss management
The government is the primary custodian of the whole food chain losses in the current business model.However, most actors' food loss value is derived from the ability to retain as much of the value lost from production, extrinsic logistics, intrinsic processing and market access, government returns on investment in the food loss business model according to FG3 is "about meeting centrally set targets such as local consumption, food security, food availability and sustainability and improved livelihoods".FU5 explained that in the current model, farmers are unable to realise better prices and premiums due to a "lack of transparency in the trading process, price cartelisation, and payment delays".The lack of government oversight of the aggregation function for marketing infrastructure and data coupled with the fragmented landholding and differentiated production landscape upstream has: prevented a lot of private players, landowners and other entities from investing in developing specialist markets, and e-markets because they are unable to guarantee returns to shareholders on such investments outside of government support, FN3.
In the current food loss model, cluster-based business organisations (e.g.NCDC) serve as resource institutes working with central, state and local governments to ensure that FPOs can reduce food losses to farmers.However, "'smallholder beneficiaries' have complex livelihood, socio-economic and socio-ecological challenges that are difficult to effectively represent through cluster-based organisations pursuing the government agenda" FN3.The key challenge with instituting a centrally managed aggregation function is that the current cluster-based approach for aggregating FPOs has the unintended consequence of excluding smallholders.

Non-governmental organisations functions and current business model for food loss
In the current business model, the fragmented nature of NGOs' role at national and state levels has led to inefficiencies in facilitating the data, infrastructure and capability aggregation functions, for which NGOs are better positioned to link to the private actors.Furthermore, with the devolution of FPO aggregation function to several agencies and sub-department of the government, the relevant data is managed in silos, thus making it challenging for NGOs to perform a defined aggregation role.The challenge with the undefined role (through policy) for NGOs is lack of data sharing which is:

Strategies for transitioning to a networked sustainable business model
The following emerged as potential strategies (Table 4) for transitioning to a networked SBM for reducing loss of fresh produce at early stages in the agri-food supply chain: ownership configuration, value addition, beneficiary identification and community capacity building.The following subsections explain these in relation to the context.

Sustainable business models
Subhanjan Sengupta et al.
Supply Chain Management: An International Journal

Ownership configuration
There is a need for redefining ownership to stewardship through multi-actor participation for shared value creation and capture, while at the same time increasing governmental role and presence in taking some ownership of the losses and operationalising infrastructural solutions.Multi-actor participation, both public and private, in SBM refers to cooperative participation in the network for sharing value creation and costs in recognising and addressing sustainability challenges (Freudenreich et al., 2020).Without that, there is risk of innovation getting staggered due to ownership disputes as found by some studies (Mohan et al., 2023).Interestingly, we find that the existence and visibility of institutional coalition are interpreted by the supply chain actors as an indication of a potentially conducive environment in the ease of doing business and in access to subsidies, input resources, infrastructure and credit.Farmers tend to have more trust in government-promoted collection and distribution centres.In creating a hub-based aggregation of organised players to share responsibilities for developing feasible solutions, FPOs also need to bring the government's participation for legitimacy and outreach, while at the same time working on synergies in policymaking, planning and implementing.
Networked SBMs need value networks created through informal-formal collaboration in the value chain (Comin et al., 2019).Collectivization is essential to connect formal and informal actors (farms, supply chain operators and marketplaces) to absorb cost and achieve economies of scale with new interventions.According to TP4: Some kind of aggregation is needed so that cost of production can be reduced and maybe a model cluster should emerge, so that they can absorb these costs.The Federation based model can employ some good professionals who are involved in the infrastructure creation and aggregating of FPOs, because all these require infrastructure and investment done collectively as per policy directions, thus making a different kind of business model.
However, mobilising such stewardship has challenges at scale.For local-level collectivization, farmers and farmer groups need support to form their supply clusters and generate supply information for estimating demand for affordable and portable temperature and humidity-controlling solutions.However, this would need public-private partnerships, as the product eventually travels through them.Recent research finds that cooperation between upstream and downstream actors is crucial for managing food supply chains (Luo et al., 2022).We find that formal players owning collection and distribution centres need to come forward and invest in novel solutions at a network level, because they are the largest market-oriented beneficiaries of a streamlined supply chain with minimal loss.

Value addition
In this study, we find that solutions for reducing the loss and waste of fresh produce at the most local level need quick logistics, distribution and market infrastructure to make the solutions viable.Frontline operators lose value due to poor quality of transportation, logistical delays or unavailability of vehicles and manpower.Short-term cooling solutions are not available to increase the shelf-life of fresh and fast perishable produce, to tackle uncertainties found in this study as well as in past studies, such as uncertainties in transportation, market what needs to be done, as far as the price volatility is concerned!Now there is a farmer who has harvested, or who is about to harvest.[. ..]Suddenly there is a change in the market prices and things are not going right.He should have an option of basically storing it for a while, or something like that.From that perspective, I was talking about short term storage.
Past research signifies the role of digitalisation in overcoming sustainability challenges in agri-food value chains (Corallo et al., 2023).We find that business decisions that would prevent or reduce food loss at the post-harvest stages of the supply chain, need to be driven by information feedback among the actors in each node of the supply chain, most importantly with the farmers for planning sowing, harvesting and farm-level facilities.
I believe how much needs to be produced if that data and dashboard information is sent back to the farmer, then it will be more meaningful so that the storage can be planned well ahead and waste is less [. ..] so the feedback is missing now, says TP2.
Without a feedback loop between actors there is limited scope for new value addition to mitigate loss of value due to food loss."Value" is at multiple levels in a multi-actor networked SBM, thereby necessitating a systemic understanding of solving sustainability problems at the micro-level for creating win-win situations (Comin et al., 2019).With real-time information, a viable sorting and storage facility and market linkages at the very farm level, farmers can not only deal with uncertainties but also have better planning and can also reuse poor-quality produce in creating organic farmyards.This calls for solutions to ensure that low-quality food at the farm level is put through circular economy practises implementable for the benefit of the farmer, reinforcing the need of a circular mindset in agri-food (Corallo et al., 2023) starting from the grassroots value chain.Climatic and soil conditions, poor seed quality and staggered production, alter the sowing and harvesting times.This challenges decision-making on when to take market orders and order transportation.Estimating harvesting and pick-up becomes problematic, thereby needing a solution to cover for time and value lost.Further, the producers and their excess produce need to be brought closer to urban consumers, such as in the form of mobile shops, or to low-income markets with low rejection rate, to generate additional value through informal settings, over value creation through the formal economy value chain.

Beneficiary identification
Two types of beneficiaries emerged from the analysis: informal sector beneficiaries and formal sector beneficiaries.Informal markets are of enormous capacity, with no access to cold chains.They need low-scale, highly affordable and portable temperature and humidity-controlling solutions that can create value for them.Although the informal sector actors constitute an important part of the agri-food value chain in developing economies (Wegerif, 2020), they are undercapitalised, severely affected by rapid change and ineligible/unaware of support that may exist (Christiaensen et al., 2021).A networked SBM within AIS would require practices that make dominant actors higher up in the value chain include informal sector actors in designing and implementing systemic solutions.We find that while large retailers can invest in temperature-controlled infrastructure, solutions are desperately needed for small and informal retailers who face worse challenges with unsold produce.With Currently, the demand-supply match-making in our research setting was found to be based on premonition.In the current configuration of the business model, the majority income share goes to market agents and aggregators, rather than the farmer producers."The product cycle is also very low -2 to 3 days; maximum share of consumer price should go to the farmer, but that is not happening," says DC5. SBMs are inclusive as they engage income-constrained groups in the value chain for addressing long-neglected social and economic problems (Schoneveld, 2020).We find in our research setting that a lease and service-oriented model might allow multiple users (actors) access to a shared infrastructure of agri-tech solutions, for everyday decisions at the grassroots level.The cash and service model can be routed and operated through producer cooperatives for transparency and consolidation.
We should be able to bring sprinkled temperature controls under one funding infrastructure so that it can be run using a service-oriented model where maybe farmers who need access to that can do it through the farmer producer cooperative, says NGO3.
The cost-bearing entities can be producer groups/cooperatives and the off-takers who can give access to the infrastructure through micro-leasing.For the produce that has consumers paying premium price, direct transaction is desirable between the farmer producer groups and those offering farming contracts, to increase farmer income and enable win-win capital infusion.

Community capacity building
A networked SBM materialises through multi-actor participation for value creation, which is inclusive in ways that support the sustainability of participating groups/collectives instead of individual actors (Schoneveld, 2020).Capacity building is necessary for developing the organising skills of small-to medium-scale actors in the post-harvest supply chain so that they become beneficiaries of innovation rather than being displaced by it.Enabling resilience at the community level and creating community-level resources is crucial to becoming a sustainable agri-food supply chain, rather than focusing on the economic viability of an individual actor in the supply chain (Stone and Rahimifard, 2018).Incubation and hand-holding are needed for farmers and other front-line actors accessing institutional support and managing new technologies.Organised collectives such as farmer cooperatives and FPOs need training on professionally managing everyday operations, accessing finance and meeting certification requirements.This goes beyond the focus on the business community in the generic food supply chain (Van Beusekom-Thoolen et al., 2023), to recognise the social economy actors in the agri-food value chains of the developing world, who need support for addressing their sustainability challenges.
The three-month period between demand estimation, buyer identification and sowing seeds, can be used for forecasting demand for off-grid cold-chain equipment.This is also the time in between crop cycles where digital solutions are needed for real-time market linkages and for predicting harvesting time, pricing and demand.However, digitalizing through platform-

Sustainable business models
Subhanjan Sengupta et al.
Supply Chain Management: An International Journal based applications and making it accessible to local supply chain actors is a significant behavioural change in these locations.
Data digitisation from a register to an app is not really a real challenge as much as it is shown.So, they know it, but they don't want it.Digital transformation and change management is always a problem.It is more of a mindset issue than technology.There are a few people who will benefit from it but there are others who won't; they are looking at short term gains, says TP2.
It calls for extensive campaigning on digital awareness and training of the farmers and supply chain operators, alongside capacity development in infrastructure for access to internet and data.
Affordable off-grid cold-storage capacities at the farm level can prevent the loss of produce, respond to delay in market demand and even create opportunities for cultivating highvalue crops.Another possibility for achieving higher intrinsic value is to use these capacities for alternative value propositions, such as food processing.

Integrating AIS with networked sustainable business model through proposed strategies
We argue that proposed strategies would enable a networked SBM infused with AIS for reducing multiplier and stacking food value loss.We elaborate on the relations between the enabling strategies and two forms of food value losses in networked SBM using ten propositions.These propositions are developed through an interplay between existing literature, proposed strategies and causes of food value loss in the current business model (Cf.Sections 4.1, 4.2, Table 5).Further considerations and potential questions for future research are presented in Section 4.3.2 and captured in Table 6.

From a siloed to a network view of sustainable business model
From our findings, we argue that to design an SBM within AIS, the value that is derived from siloed business models has to be networked and aligned.There are two broad categories of value creation in networked SBM infused with AIS, namely, multiplier (Cronin et al., 2022;Fieldsend et al., 2022) and stacking value (Cronin et al., 2022;Parmar and Kumar, 2022).We argue that to achieve inclusive and sustainable value creation within AIS with a networked SBM, the enabling strategies are crucial.
Firstly, in line with previous studies (Srai et al., 2021), we see that the business models used by the actors for food loss reduction in the Indian fresh food AIS are fragmented.Despite government investments in FPOs and digital initiatives, varying, competing or conflicting policies make different actors pursue different scopes of food value loss (Srai et al., 2021;Sagi and Gokarn, 2022).For instance, while farmers are more concerned with minimising production value losses, the FPO business model captures the institutional value of food through livelihood, food security and other broader sustainabilitydriven initiatives.Owing to a lack of linkages between formalised government top-down initiatives and informal business models for managing production (pre-production, onfarm and postharvest), intrinsic and extrinsic losses diminish the realisation of market, institutional and future food values for actors through government AIS initiatives, particularly small-holders (M endez-Le on et al., 2022).Furthermore, we found that supply chain actors and policymakers have competing values which are not aligned, leading to double marginalisation of farmers, i.e. farmers bear most of the losses in other actors' business models as the competing policies do not reward or penalise any other actors for food loss (Srai et al., 2021).For instance, through state FPOs, the government incentivised production of particular crops, such as tomatoes, but there were conflicting policies for accessing the markets, storage and pricing leading to massive food loss, all borne by farmers.In response to this, the Federal Indian Government has recently launched the "Tomato Grand Challenge" in 2023 (Department of Consumer Affairs, 2023).
Secondly, in line with previous literature on multiplier and stacking effects in AIS (Cronin et al., 2022), our study proposes two forms of interdependencies that enable or hinder value realisation for operationalising a networked SBM for reducing food loss.We argue that the ownership of the values generated in the current business models is fragmented and certain categories of values, particularly those generated by informal actors, have no formalised ownership and business models.Hence, those affected most by these value losses are farmers.We characterise this form of networked food loss as the multiplier food loss, which results when production value losses in pre-production, on-farm and postharvest food loss management has a multiplier effect in other scopes of food losses in the chain, including intrinsic value, extrinsic value and market value.
Thirdly, we found that there are gaps in community-level capacities (managerial, digital, infrastructure) that are required to realise production, intrinsic and extrinsic values despite government-led AIS deployments for market value realisation.In line with previous studies on stacking effects in AIS (Brewster et al., 2017;Blash et al., 2020), we characterise this form of food loss as stacking food value loss.From our findings, we argue that there are significantly important data stacks at the farm gate for managing pre-, on-farm and postharvest food losses, such as data related to weather, soil, pest, production planning, etc. and if any of these data stacks are missing because of lack of capacity to generate and use them, it leads to losses stacking up in other parts of the food chain (cf.Table 3).

Propositions for achieving networked sustainable business model through AIS
While previous studies have identified multiplier and stacking effects in AIS, our findings about the current business models show that multiplier losses arise from the lack of shared ownership of production value due to largely informal pre-, onfarm and post-harvest practices.These losses in production value cumulate in diminished market value.Further, our findings show that stacking losses in food value arise because of gaps in infrastructure and data stacks for managing food losses arising from inadequate community-level capacities (managerial, digital, infrastructure) for generating and sharing crucial production, processing and logistics data stacks (Fieldsend et al., 2022).These losses cumulate in diminished institutional and future food value because of the lost opportunity to discover new values (e.g.net zero, carbon trading, new connections to market, competitive pricing, etc.).We emphasise that these two forms of food value losses are not mutually exclusive, however, they are analytically distinct       as one is primarily based on coordination and relational dynamics of the actors while the other pertains to the flows and allocation of tangible (e.g.infrastructure) and intangible (e.g.data, finance) resources in AIS (Cronin et al., 2022;Elias and Marsh, 2020).Considering the interplay between these value losses due to siloed business models and the enabling strategies (see Section 4.2) identified by actors in the Indian agri-food system, we make the following ten propositions for developing networked SBM within AIS (also see P4. Aggregating quality management and assurance procedures in pre-and post-harvest operations can reduce market losses arising from mismatches between demand-supply specifications. The traditional top-down audits and assurances that characterise the current business model cannot work in a networked business model because the assurances required are bottom-up.To enhance compatibility in quality management and assurance procedures in a networked SBM, robust monitoring from a bottom-up (participatory) approach at the primary loss prevention stages (pre-production and postharvest) are necessary to provide the baseline data on which other assurances within a networked SBM can be accessed (Hansen and Schmitt, 2021;Planko and Cramer, 2021).
P5. Strengthening producer-market and market-market linkages within AIS prevents the stacking of market losses and enables value realisation.
Communication and data exchange between producers, markets (formal and informal) and other actors reduces the chances of stacking problems where relevant data or infrastructure from different parts of the agri system remains isolated or unused.Informal actors, such as small-scale farmers, local traders and informal market vendors, often lack access to finance, technology, training and market information, which hampers their productivity and market competitiveness.By recognising their needs and connecting them to relevant support systems, informal beneficiaries can enhance their capacity to reduce food losses and improve their overall performance (Fieldsend et al., 2022).As these actors possess traditional knowledge, skills and practices, integrating them into formal networks and platforms can lead to innovation, improved practices and reduced food losses within networked SBM.
P7. Identification of formal beneficiaries and their facilitation roles mitigates food losses from stacking onto each other.
Connecting formal actors for a networked SBM, including their needs, operations, resource requirements and expected payoffs, needs to take into account how they can facilitate financial support, technical assistance, infrastructure and market access.This is needed to ensure that the necessary support is provided to mitigate food losses and reduce the risk of stacking problems resulting from resource mismatches or inefficiencies because of poor data or infrastructure.This reduces the likelihood of food losses being hidden or passed on without accountability and enables the discovery of new value by discouraging stacking problems where losses accumulate unnoticed (Baygi et al., 2021).Identifying formal asset-bearing actors unlocks opportunities for sharing best practices, technical knowledge and innovative solutions.Capacitybuilding programs can then be designed to enhance the skills and capabilities of formal beneficiaries, empowering them to address food losses more effectively.Managerial capacity building fosters problem-solving capabilities among formal and informal beneficiaries by providing them with tools and techniques to analyse complex issues, identify root causes of losses and develop innovative solutions.This enables them to address challenges early on and prevent losses from accumulating and multiplying throughout the value chain.Managerial capacity building also cultivates a sense of responsibility and agency that enhances actors' understanding of the agricultural value chain, market dynamics and best practices (Schoneveld, 2020).This allows both formal and informal beneficiaries to make informed decisions for better-assessing risks, identifying opportunities for improvement and implementing appropriate strategies to mitigate losses at different stages of the value chain.
P9. Digital capacity building at community level enables formal and informal beneficiaries to identify and reduce stacking losses.
Digital capacity building equips formal and informal beneficiaries with the skills and tools to collect, store and analyse real-time information on market prices, supply and demand trends, weather conditions, their own operations, such as production quantities, quality parameters, transportation routes and storage conditions.For instance, beneficiaries can combine the use of accessible low-tech solutions, e.g.mobile applications for inventory management, temperature monitoring, pest detection and quality control to generate relevant inputs (data stacks) (Parmar and Kumar, 2022;Acquier et al., 2019).This enables beneficiaries to minimise stacking losses throughout the value chain by making data-driven decisions, anticipating potential losses, identifying patterns, detecting inefficiencies and adjusting their strategies accordingly.
P10. Creating capacity and access to storage and processing infrastructure that is not available at the community level can be enabled with the identification of opportunities offering new value propositions to local formal and informal actors.
Within a networked business model, access to facilities such as cold storage, drying units, milling equipment, packaging facilities and decisions to locate these at a right aggregation stage (e.g. at village level (micro) vs at a district level (mesoscale), can reduce stacking losses by extending the shelf life of agricultural products, improving product quality and diversifying local actors' value-added products (Best et al., 2022).Optimal facility location decisions based on local actors inputs (e.g.network configurations, livelihood needs) are necessary to reduce the reliance on intermediaries, increase direct market access for producers (local, regional, national distribution networks) and reduce losses incurred because of data and infrastructure gaps (cf.Propositions 5, 9) or through the involvement of multiple intermediaries.
Overall, these ten propositions contribute to the alignment of AIS and SBM literature which has been developing independently.In SBM literature, the food loss is predominantly examined based on different typologies or categories in silos.This kind of siloes view makes it difficult to incentivise networked SBM.Our research identifies networked level value or losses (multiplier and stacking) and strategies that actors can use to manage or mitigate them through the above ten propositions.We further provide future research directions, potential research questions and methodology that can be adopted to test these propositions in Table 5.In addition, our research advances the literature on AIS by introducing system-level networked SBM for food loss reduction and proposing strategies for facilitating such models which can be applied to other networked sustainable systems where actors may be running different business models in silos (e.g.energy, water, critical minerals, fashion, etc.).This analytical distinction between the two sources of networked business model value (and losses) is useful for facilitating future investigations on system-level dimensions of (food) value loss within a networked business model (see Table 6 for details).

Contributions, limitations and future directions
5.1 Contributions, limitations and future directions for theory and research This study contributes incrementally to the field of supply chain management (SCM) by presenting a structured framework that clarifies the shift from traditional isolated business models to collaborative, networked approaches, aligned with the innovation literature in SCM (Planko and Cramer, 2021).It extends the discourse on food losses in SCM by introducing an innovative categorization across different business models in the food chain, emphasising the importance of value creation over value capture, echoing previous research that takes a holistic view of value that incorporates social, economic and environmental considerations of various stakeholders (Schoneveld, 2020).In the complex context of AIS and SCM, this study distinguishes between two types of losses, multiplier and stacking losses, advancing SCM theory and highlighting the need for value alignment in collaborative supply chain models.These findings align with discussions on risk management and offer significant insights into mitigating food losses in supply chains (Cronin et al., 2022;Fieldsend et al., 2022).Additionally, the research emphasises the alignment between AIS and SBM, underscoring the importance of identifying value owners and beneficiaries within a network of AIS actors, identifying potential strategies that can enable moving from an unsustainable business model to a sustainable one and facilitating equitable and sustainable value distribution throughout the supply chain (Best et al., 2022;Acquier et al., 2019).This contribution showcases the potential for creating value from otherwise lost resources, ultimately reducing food losses in supply chains and advancing SCM towards more sustainable and efficient practices.
Although this study takes an essential first step in creating contextual knowledge about the phenomenon, the findings have some limitations which set the stage for future research.
Generated from lived experiences through induction, the results are naturalistically generalizable to other developing countries but lack statistical generalizability (Lincoln and Guba, 1985;Ketokivi and Choi, 2014).A potential quantitative study would be an impact measurement of the operationalizing the strategies and mechanisms while being implemented by a coalition of actors.Another objective could be to identify the variables of each mechanism and examine the implicit relationships and moderated mediations (see Table 6).
There is also a need for qualitative enquiries to generate a processual understanding of the project management in enabling an AIS with a networked SBM for addressing food value loss.With these actor-centric findings on potential actions, the next step in creating a process framework entails identifying the action flows, decision-making and resource allocation, to create a conceptual schema of practice loops in a relational, temporal and performative sense (Baygi et al., 2021;Garud et al., 2013) (see Table 5).

Contributions, limitations and future directions for practice and policy
The study demonstrated critical roles of external actors in AIS in facilitating the development of a networked sustainable business model for optimising supply chain management.This facilitates more effective operations and reduces food losses.The study also aligns AIS with SBM, enhancing actors' ability to deploy coordinated efforts to address food losses and promote sustainability.Furthermore, fostering mission-driven strategies that target collective goals such as reducing multiplier and stacking losses systematically empowers actors' collaborative interventions for improving resource allocation and risk management practices.Additionally, governments, NGOs and financiers can develop more targeted interventions to reduce food losses, improve value co-creation, enhance resilience by optimising resource utilisation across the agricultural value chain.These external actors have the know-how and potential to generate a variety of innovation models that are required to be managed holistically.However, as they are external actors their roles are often excluded from the development of sustainable agri-food business models.The study included key external actors, such as, government, financial institutions and NGOs.However, other external actors (e.g. research institutes, exporters, schools, etc.) whose roles might be important in understanding the value from reducing the food loss, were outside the scope of the study.These external actors play a vital role in facilitation and aggregation of values within an AIS.
The study identified four interconnected actions required to translate strategic AIS goals to specific mission-oriented business models e.g.food loss.Causal links (e.g. which action leads to which) among these actions to inform prioritisation by AIS actors need to be addressed in future studies.For instance, for more mature FPOs that have a well-structured beneficiary identification system, future studies should investigate how such actors should prioritise their actions for initiating and operationalising a networked SBM.
From the practice point of view, it is often observed that many of these innovations to business models occur outside the configuration of important actors and such innovations are integrated into the business models at a later stage.Future studies may expand on the findings of this study by exploring the efficiencies and impact of an integrated approach to operationalise an SBM while all important actors are placed within the AIS to create a supportive ecosystem.From the policy perspective, misalignment could be found between policy expectations and actors' requirements within a configuration.Future studies could explore the best possible direction for policy formulations that can encourage a bottomup approach and align the goals of different actors with the governmental interventions to create an SBM (see Table 6).

Conclusion
AIS can create SBMs if there are deliberate attempts to design a configuration of actors around a specific value chain (meaning, what is the value to different actors for reducing food loss).AIS remains like a top-down tool that informs policy development but such policies would often be misaligned with the business needs of the actors because AIS literature and SBM have evolved separately.This study contributes to linking these two fields and calls for future research on building this interdisciplinary research further with bottom-up knowledge for reducing food loss and value loss in food supply chains, particularly at postharvest level (SDG 12.3).By bridging the AIS and SBM literature, we identified two forms of networked business value (multiplier and stacking values) and four enabling strategies from the perspectives of supply chain actors required to operationalise a specific SBM for reducing food loss with broader AIS goals.
With the evolution of industrial-level technologies such as Industry 4.0 and system-level sustainability models such as circular economy, our study bridges the need for networked SBM literature to align with a system-level innovation concept such as AIS by highlighting how actors derive system-level values (multiplier, stacking) using a range of innovations.We offer ten propositions for future research and areas for further exploration on how such networked SBM can be operationalised within AIS or other innovation systems (e.g.technology, energy, water, critical minerals, fashion, etc.).These propositions aim to explore areas such as collaborative governance mechanisms, data-sharing protocols and impact assessment frameworks to further enhance the effectiveness and sustainability of networked SBM.Our study offers pathways for businesses and other AIS actors to map value streams associated with other broader sustainable goals such as circular economy, carbon accounting that share similar principles for value co-creation.
[. ..] is necessary to enable partnerships that are transparent, result-oriented and complementary [. ..] to establish clear targets for measuring and monitoring co-leadership and coordination of food loss management by various actors FN3.In this regard: NGOs clustering roles need to be clear and should ideally entail bridging the silos [. ..] data and infrastructure created by FPO clusters to establish co-lead roles among governments, FPO clusters and NGOs and identify activity areas [data management, infrastructure, capacity building] where actor's roles in food loss management complement [. ..] enhance [. ..] or substitute each other FF2.

Table 1
Interview and focus group participants with mnemonics

Table 2
Example of the inductive thematic analysis am very particular about harvest date and time.The moment they harvest, I want them to send me on the same day.Sometimes what they do is that, because this is all labour oriented, they will do some other order.They will harvest yesterday, and then pack and sell today (R 10) Today I am doing 100 kgs.If there is no cold storage, I will lose around 25% to 30%.If you can provide those interventions, the quality of the material will be good and the price will be high.Today it is 6 rupees tomato, tomorrow is holiday for the rest of the states.The local farmers know very well about the holidays.If those storage facilities are there, they will keep the material in the storage (NGO 1)

Table 3
Actors and associated aspects in current business model leading to food loss

Table 4
(Mohan et al., 2023)itioning to a networked SBM for reducing post-harvest food loss and value loss Supply Chain Management: An International Journal fluctuations, climatic conditions and fluctuating prices(Mohan et al., 2023).Easily accessible and affordable solutions are needed to preserve and assure food quality at pre-and postharvest stages, to wait out potential value loss and to manage surplus.TP1 says: Building digital capacity: Digital awareness and training for mindset change, digitalizing access to farm data and for predicting harvesting time Building infrastructural capacity: Installing cold solutions at local level and increasing value with processing opportunities Source: Authors' own creation

Table 5
From data and literature to propositions

Table 6
Future research questions

Table 5
(Cronin et al., 2022;Fieldsend et al., 2022) would usually execute their own business model in silo to achieve their innovations, with or without clarity on ownership.Where there is no clear ownership, any losses accrued from one actor cascade and multiply through the chain irrespective of the strategies in place(Cronin et al., 2022;Fieldsend et al., 2022).
Information flow also enables the discovery of new values and the realisation of institutional values (Best