Data sharing platforms: How value is created from agricultural data

Highlights

• When you create value from data, you operate a data sharing platform.

• Creating value from data requires a community of stakeholders, a faciliatory system, and data on, and for, the community.

• Agricultural datanomics is the interdisciplinary study of data as an economic good created by, and for, agriculture.

• Data sharing platforms enable managers to answer the question, ‘how do I increase the value of Smart Farming data?’

• This paper evaluates common data management techniques and proposes strategies that increase the value created from data.

Abstract

CONTEXT

Across agriculture, data is produced, enriched, and consumed through the centuries-old practices of producing food and fibre. The adoption of Smart Farming and its connected services and techniques accelerates agriculture's dependence on data, yet the process of creating value from data is not well understood.

OBJECTIVE

What assets and management decisions comprise the process of creating value from data? What are the properties of this process, and where should resources be invested to increase the value created from agricultural data?

METHODS

We extend platform economics theory with the results from a recent systematic literature review of Big Data in Smart Farming to show the creation of value from data occurs in Data Sharing Platforms.

RESULTS AND CONCLUSIONS

Data sharing platforms are systems that connect the layers of the ‘platform stack’ with pertinent management tasks to create value from data. We illustrate this arrangement as a sectioned, three-circle Venn Diagram and evaluate the efficacy of common data management techniques in the creation of value from data. This paper concludes that value is created from data only when each of the components of data sharing platforms are present and that the operation of a data sharing platform describes the process that takes data as an input and produces value as an output. Further conclusions relate to commercial and institutional aspects of the creation of value from Smart Farming data.

SIGNIFICANCE

The proposed model is useful for evaluating production processes that create value from data. This paper details several avenues for extension. Productive models of systems that rely on data as a core asset may now be assembled. Policies that trade off technical characteristics of data with social impacts of data may now be approached. Questions surrounding data ownership may be considered with greater clarity.

Introduction

Agriculture 4.0 has signalled a proliferation of connected sensors across farms and throughout value chains whose streams of data offer the alure of value to those who possess the requisite skills or acumen. Staple farm equipment such as rain gauges and soil moisture probes are transformed into Internet of Things (IoT) devices that document the farming system in constant and ever-changing data flows (Friess, 2016). Meanwhile, traditional farm machinery such as tractors or livestock weigh stations are connected to non-agricultural products like smartphone apps to leverage on- and off-farm data (Barrett, 2021). These devices digitise farms and create new opportunities for agricultural value creation through the data-driven “optimization of agricultural production systems, value chains and food systems” (Klerkx et al., 2019, p. 2). In turn, data-based opportunities present farms with operational models that are radically different from previous generations. Farming machinery is subsumed into the sharing-economy (Daum et al., 2021; Venkataraman, 2016), robotic workforces operate alongside human labourers (Christiaensen et al., 2021) and provide insurance against labour shortages (Barrett, 2021), while a single farm's access to finance can be hedged against whole-of-chain performance (AgriDigital, 2021; Jakku et al., 2019). In each case, a common belief in the value of agricultural data connects traditional agricultural problems with solutions from – often – non-agricultural communities. Where value creation in agriculture was once inseparable from labour- and capital-intensive activities, value creation in Smart Farming is now unavoidably, data intensive.

The data produced by Smart Farming intertwines agriculture and business in new ways. Where Smart Farming is the application of information technology to optimise the farming system, and agribusiness, the science of making value-adding decisions regarding the agricultural assets in that system, we propose agricultural datanomics as the interdisciplinary study focused on the treatment of data as a valuable resource created by, and for, agriculture. Accordingly, agricultural datanomics includes consideration of the sources of value creation (Kieti et al., 2021) through to market design (Agyekumhene et al., 2020) where stakeholders act as both producers and consumers of the value produced by Smart Farming data. Transparency is vital for embryonic digital markets (Jakku et al., 2019) as buyers and sellers often collaborate around data-products to extend value created by their exchange (Faulkner et al., 2014). Stakeholders multi-home (Rochet and Tirole, 2003) with digital goods (Barrett, 2021) as their products or services participate in multiple markets simultaneously. This creates highly fragmented markets as ad hoc, agricultural platforms (Klerkx et al., 2019), whose primary advantage is access to crowd-sourced resources (Parker et al., 2016), compete with the ‘pipeline’ businesses (Choudary, 2015) that traditionally dominated agriculture. There is widespread agreement in the centrality of digitisation in agriculture's future (Jakku et al., 2019; Klerkx et al., 2019), and considerable research into the science of managing the value produced by those processes (Birner et al., 2021; Wiseman et al., 2019) but the issue of how value is created from Smart Farming data is only beginning to be addressed (Darnell et al., 2018; Rojo-Gimeno et al., 2019; Wiseman and Sanderson, 2019). The study of this issue requires consideration of both technical and economic properties of data (Birner et al., 2021; Fleming et al., 2018b; Nikander et al., 2020), in conjunction with agribusiness assets and associated management responses (Chen et al., 2014; Wolfert et al., 2017). In this context, this paper aims to address two of the fundamental questions for scholars of agricultural datanomics: (i) what assets and managerial tasks are required to support the creation of value from Smart Farming data, and (ii) what properties must the process of creating value from Smart Farming data possess? We approach both questions in parallel, initially collating extant academic literature to establish a framework of seven components required to create value from Smart Farming data, before examining how their arrangement dictates that value creation process.

Value created from data, $\mathbb{V}$, may be understood as the difference between the benefits enabled by the data, $\mathbb{B}$, less the costs incurred to realise those benefits, ℂ, or, $\mathbb{V}=\mathbb{B}-ℂ$. This approach is widely adopted (Chesbrough, 2003) and is well-established in agriculture. Value created from cropping or livestock production is the difference between the sum of market and non-market benefits less the associated costs of bringing those goods to market. Benefits enabled through the adoption of new technology or management techniques increase the value of the underlying agricultural assets through more effective observations, measurements or analysis. Likewise, a reduction in the costs of optimising the farming system through the application of Smart Farming increases the value of farming operations by improving access to, and application of, data across the farm. Investment in key assets or accompanying management tasks increases value created from Smart Farming data as benefits are increased, or the cost of achieving those benefits is reduced.

Data produced by Smart Farming is the core resource that enables value, is often the good exchanged, and even the currency that finances interactions throughout agricultural value webs (Darnell et al., 2018). While agricultural data is at least an asset and “must be managed like any other [asset]” (Wiseman and Sanderson, 2019, p. 3) unlike other classes of assets data possesses near-infinite economies of scale (Arrow, 1996). However, this property of data does not confer corresponding economies to the resulting value. Additional factors such as stakeholder's decision rules, social intentions, and the technology used to transform data into information, set bounds on the value created from Smart Farming data (Rojo-Gimeno et al., 2019). Value that is created from Smart Farming data is contingent on adoption and utilisation of that data by a heterogeneous community of stakeholders who assemble according to a congruous goal (Wysel, 2019). Bearing some resemblance to integrated value systems (Papazoglou et al., 2000), stakeholders voluntarily exchange data with one another provided their individual benefits exceed both their search costs and transaction costs (Kieti et al., 2021). In this way, the farming system acts as a platform upon which a community meets to exchange data. However, stakeholder's participation in data sharing platforms is more than transactional (Agyekumhene et al., 2020); value can be created from Smart Farming data as stakeholders collaborate around data (Birner et al., 2021), conduct arbitrage with data (Jensen, 2007), or compete for data (Jakku et al., 2019). Additionally, the data and value created on a data sharing platform can be treated as a private good (Wiseman et al., 2019), club good (Fleming et al., 2018b) or public good (Sanderson et al., 2017).

Therefore, the creation of value from Smart Farming data requires more than individuals or agricultural collectives investing in digital assets or technology management. An understanding of both the complimentary assets – that must exist beside data – and the management tasks required to develop these assets is an important first step in investigating the process of creating value from Smart Farming data. The remainder of this paper is organised as follows: Section 2 situates this work among the current state-of-art academic and management literature on the creation of value from data and the results from Wolfert et al. (2017)’s recent systematic literature review on Big Data in Smart Farming. Section 3 presents an intuitive framework placing each of the seven components required to create value from Smart Farming data. Section 4 examines the interplay between each component and assembles a model that describes their effect on value creation. Section 5 discusses the implications of the model for scholars and practitioners as they analyse and manage value created from Smart Farming data. Section 6 concludes with several avenues for extension.