Importance of Data Strategy for the Intellectualization of Agriculture. Business Recommendations Exemplified by the Chinese Agriculture Market

Table of Content

List of Figures

List of Tables

List of Appendices

List of Abbreviations

1 Introduction
1.1 Problem Statement
1.2 Objectives
1.3 Structure of the Paper

2 Data Strategy and Intelligent Agriculture
2.1 Data Strategy
2.1.1 Defensive Data Strategy
2.1.2 Offensive Data Strategy
2.2 Intelligent Agriculture – Agriculture 4.0

3 Analysis of Opportunities and Requirements

4 Lessons Learned

5 Summary and Conclusion

Bibliography

List of Figures

Figure 1: Data Strategy Framework.

Figure 2: Two-sided Prerequisites for the Development of Intelligence.

List of Tables

Table 1: Elements of a Defensive Data Strategy Approach.

Table 2: Elements of an Offensive Data Strategy Approach.

List of Appendices

Appendix I Data Strategy Framework: Defensive

Appendix II Data Strategy Framework: Offensive

Appendix III Development of Chinese agriculture (I)

Appendix IV Smart Agriculture: Applications, Services & Sensors

Appendix V Development of Chinese agriculture (II)

Appendix VI TPYN’s Roadmap to Success

List of Abbreviations

AI Artificial Intelligence

CAGR Compound Annual Growth Rate

CDO Chief Data Officer

CIO Chief Information Officer

Covid-19 Coronavirus SARS-CoV-2

ICD International Data Corporation

IFPRI International Food Policy Research Institute

IoT Internet of Things

IWD Informationsdienst des Instituts der deutschen Wirtschaft

MMVO Multiple Versions of the Truth

SITM Strategic IT Management

SSOT Single Source of Truth

TPYN Top Cloud-Agri Technology Co., Ltd.

ZB Zettabytes (Unit of information equal to one sextillion (10²¹) or, strictly, 2⁷⁰ bytes)

1 Introduction

“More than ever, the ability to manage torrents of data is critical to a company’s success.”¹ From a Chief Information Officers (CIO² ) perspective, this statement could not be any timelier. However, it is several years old. Prior to the current Covid-19 pandemic, a number of technology trends around Big Data, Internet of Things (IoT), and other emerging technologies have dramatically boosted the amounts of data being generated. Despite the impression one might get from the media, the pandemic has not only affected the companies for the worse. Among evident difficulties, tremendous opportunities have emerged. Latest data shows, that the economy have leaped “five years in digital adoption […] in about eight weeks”³. In figures, what does this mean? Looking back to 2018, about 90% of the data available was generated in the prior two years.⁴ This year, the Covid-19 pandemic is a true data booster⁵, which will result in a total annual data volume of more than 59 zettabytes (ZB).⁶ Hence, there is no sign of this trend stopping. By 2025, the volume of data worldwide will increase to 175 ZB – a CAGR of 27%.⁷

Alongside the data volume growth, a second, inexorable growing trend is evident: the world's population. Although the relative growth ratio is not exponential, according to the United Nations (UN), the population will increase from the current level of just under 8 bn. to nearly 11 bn. by the end of the century.⁸ In order to meet the future food demand, food production needs to double by 2050.⁹ Given the combination of climate change and expanding urbanization and the destruction of agricultural soil, farming needs to be more efficient and intelligent in order to prevent world hunger.¹⁰ In the transformation from analogue to digitized agriculture, aforementioned data growth is equally applicable to agriculture business. Developing “more efficient farming practices”¹¹ is one promising way to tackle this issue. This means to integrate artificial intelligence (AI) into agricultural production.

1.1 Problem Statement

It is generally agreed today that a proper data strategy serves as a decisive competitive advantage. However, empirical evidence shows a cross-industry lack of (optimal) data utilization.¹² As digitization does not leave any industry unaffected, even traditional ones – that have been operating in analogue environments – are now forced to enrich their (digital) business models with data while simultaneously developing appropriate strategies for dealing with the growing data volumes. A prominent example of such a traditional industry is agriculture. Consequently, in today's world, the management has to answer itself the question: What is our data strategy ? This question has prospered and occupied in high-level management. Many board executives assume that it is feasible to leverage (customer) data to gain an invincible competitive advantage.¹³ Nonetheless, the reality differs. Cross-industry studies have found, that less than half of structured corporate data is actively used for decision-making.¹⁴ It is important to emphasize that agriculture is a large heterogeneous industry, that results from the subsumption of several sectors.¹⁵ This paper will not cover the agriculture industry as a whole, but rather focus on the so-called crop sector. Chinese agriculture is increasingly becoming a pioneer in digitalization. With this in mind, the paper provides an analysis of success factors derived from best practices in Chinese agriculture.

1.2 Objectives

The objective of this paper is, firstly, to emphasize the relevance of data and the impact of applying an appropriate corporate data strategy. Secondly, to raise awareness for the intellectualization of productive industries (e.g. agriculture) in order to become more effective. Both objectives are partly covered in the introduction chapter and will be further illuminated in the course of the paper by elaborating on the theoretical background. The aspects Data Strategy and Intelligent Agriculture have received much attention in the past decade. Consequently, they became a central issue in numerous research papers. Previous work has only focused on one of the subjects at a time and therefore failed to analyze their interplay. Few researchers have addressed both topics simultaneously. This paper seeks to address how both subjects influence and depend on each other. Thirdly, the derivation of concrete guidelines for companies how data strategy and the development of intelligence must be synchronized. Using best practices based on the analysis of the agriculture industry, recommendations will be formulated. The ability to manage data wisely is becoming increasingly important for businesses both today and in the future, and it is therefore a decisive key competence for their success.¹⁶ One central question is to be discussed by this paper: When implementing a data strategy, should an agricultural company develop and continuously adapt the strategy to changing requirements to remain competitive in fast-moving markets? To answer this question, we look at a company that has successfully made the transition from a traditional agricultural business to a cloud service provider: Top Cloud-Agri Technology Co., Ltd. (TPYN).¹⁷ This methodology ensures proximity to practical application, which in turn facilitates the development of practice-oriented use cases. Later, business recommendations (Chapter 3) will be derived from the company's strategic decisions based on the theoretically knowledge about data strategy gained in Chapter 2.

Based on the subject, the paper can be assigned to the discipline of strategic IT management (SITM). Furthermore, a reference to the cross-disciplinary field of digital transformation can be made. Eventually, emerging technologies are required to enable and support the growing intelligence.

1.3 Structure of the Paper

This paper is structured in five chapters. The first chapter (Chapter 1) elaborates the background and the motivation behind the topic. Furthermore, the general approach and the objectives of the research paper are defined. In the second chapter (Chapter 2), the two main aspects are discussed: Data strategy (Chapter 2.1) and intelligent agriculture (Chapter 2.2). The data strategy is explained considering the current academic approach. This is followed by a narrative of the evolution from analogue agriculture to intelligent agriculture, which necessitates the utilization of digital solutions. Both subjects build the foundation for the third chapter (Chapter 3). Herein, the current and future data strategy requirements of enterprises are analysed by considering the advancement in the Chinese agricultural market. This analysis embodies the essence of the paper. With the preceding analysis, tangible business recommendations will be derived in the fourth chapter (Chapter 4). Finally, a summary and critical reflection of the results (Chapter 5) concludes the paper.

2 Data Strategy and Intelligent Agriculture

Against this background, the central aspects that determine this paper are data strategy (Chapter 2.1) and intelligent agriculture (Chapter 2.2). It is important to be clear about their definitions and relevance. Hence, in the following two chapters, these two aspects are discussed, and the theoretical background elaborated.

2.1 Data Strategy

To truly assess the relevance of data strategy, one first has to become aware of the impact data has on a company's operations. For years, data has been enabling top management to derive strategic decision-making based on models. Recently, the aforementioned growth in the amount of data is forcing companies to think strategically about handling their data. Data therefore takes on a dichotomous character in an organization. One is the input to management decisions and strategy development. However, data itself requires a strategy for its generation, storage, evaluation, use and deletion. For the purposes of this essay, the term data strategy therefore will be taken to mean the latter.

When it comes to data strategy two key approaches to design an appropriate data strategy arises. Both will be concretized in this paper: Defensive (Chapter 2.1.1) and offensive data strategy (Chapter 2.1.2).¹⁸ For each approach, there are ideal circumstances and a corresponding collection of technologies supporting each approach. They can be differentiated “by distinct business objectives and the activities designed to address them”¹⁹. However, in practice, there should always be a weighting and a constant realignment based on changing conditions. This balancing act is crucial for the success of a company.

Each approach involves both a strategical and a technical dimension. The strategic part tackles the question of which data source a company bases its operations on.²⁰ The technical part addresses the supporting IT system.²¹ The Data Strategy Framework in Figure 1 has been developed to picture a holistic synopsis of the most relevant elements within a data strategy. For a clear understanding of the two approaches and to ensure comparability among them, the following description is limited to four elements of the data strategy: key objectives, core activities, data management orientation and enabling architecture.²² Those are the result of the publication of DalleMule & Davenport in 2017 and are particularly appropriate to differentiate between them.

Figure 1: Data Strategy Framework.

Abbildung in dieser Leseprobe nicht enthalten

Source: Own representation based on DalleMule & Davenport (2017), p. 4–11.

2.1.1 Defensive Data Strategy

Let us turn the attention on the first of the two approaches – namely the defensive approach. The fundamental objective of data defense is “minimizing downside risk”²³ of business operation. In order to achieve this, multiple activities support risk minimization such as compliance functions with regulations (e.g. for rules governing data privacy, integrity of financial reports).²⁴ The defensive approach²⁵ can be distinguished from the offensive approach on the basis of four criteria. The specification of these criteria related to the defensive data strategy is shown in Table 1.

Table 1: Elements of a Defensive Data Strategy Approach.

Abbildung in dieser Leseprobe nicht enthalten

Source: Own representation based on DalleMule & Davenport (2017), p. 6.

To sum up, data defense is particularly about data security, privacy and integrity. These are the building principles to ensure governance to meet regulatory compliance. Defense is the first choice when management seeks the utmost control over internal and external data streams. The concept of a single source of truth (SSOT) supports the claim for control from an architectural point of view. SSOT addresses one of the two most important concerns when it comes to data: redundancy and consistency. This is achieved by storing and extracting all company data from one single source – often a cloud-based repository containing an authoritative copy of all crucial data.²⁶ A consequence of this is the fact that data defense is applied extensively where regulation imposes stringent requirements on companies (e.g. hospitals and health sector).

2.1.2 Offensive Data Strategy

Despite the approach just introduced, there is the offensive data approach. Reflecting back on the objective of the defensive approach - risk minimization - one can observe that the offensive approach is not only the counterpart by name. By implementing offensive data strategy²⁷ management is taking risk in order meet business objectives such as “increasing revenue, profitability, and customer satisfaction”²⁸. This can be achieved by performing activities such as data analytics, modelling, visualization and transformation. These activities require a high degree of flexibility in handling data. Thus, a key component is the deviation of multiple versions of the truth (MVOT) from the SSOT. In this process, data from the SSOT is processed and sectioned differently for certain business functions to achieve the best possible flexibility. Due to the high demands on new processing approaches like data analysis, emerging technologies are increasingly applied (e.g. data lake, digital twin²⁹ ).

The specification with respect to the offensive data strategy is shown in Table 2.

Table 2: Elements of an Offensive Data Strategy Approach.

Abbildung in dieser Leseprobe nicht enthalten

Source: Own representation based on DalleMule & Davenport (2017), p. 6.

2.2 Intelligent Agriculture – Agriculture 4.0

Despite the perception one could have, the reality is that today’s agriculture industry is data-centered, precise, and highly intelligent.³⁰ The first electronic control systems were used in agriculture more than two decades ago.³¹ The Chinese agriculture takes a leading role in the development towards intelligence and autonomy. Agriculture in China experienced four periods of development: traditional, mechanization, automation, and intellectualization (Appendix III).³² The degree of digitalization and intelligence has been increased in each period. In retrospect, the shift towards Agriculture 4.0 seems to be consistent. However, two factors had to come together to make this development possible (Figure 2). The high maturity level³³ can only be accomplished by combining the digitalization of the economy and the digitalization of individual sectors (e.g. agriculture). Digital agriculture is crucial to generate the data required (e.g. information on plant growth, pests or fertilizer) and to support the processes with real-time monitoring through sensors and integrated systems (Appendix IV). Integrating these digital processes and matching them with data from the digital economy (e.g. market developments, customer expectations, storage, transport, suppliers), provides the correlations to derive intelligence from the data.

The higher the share of intellectualization raises, the greater the amount of generated data climbs and hence the need for managing the data in a strategic manner.

[...]

---

¹ DalleMule & Davenport (2017), p. 4.

² Throughout this paper, all gender-specific terms are to be considered to refer to both the feminine and the masculine form – except when referring to a particular person.

³ Baig et al. (2020).

⁴ Cf. Marr (2018).

⁵ Caused by a surge in the number of remote workers and video communications, as well as a noticeable increase in the consumption of downloaded and streamed video. Cf. International Data Corporation (IDC) (2020).

⁶ Cf. International Data Corporation (IDC) (2020).

⁷ Cf. IWD (2019).

⁸ Cf. United Nations (2019), p. 6.

⁹ Cf. United Nations (2009).

¹⁰ Cf. IFPRI (2015).

¹¹ Dharmaraj & Vijayanand (2018), p. 2122.

¹² Cf. Davenport & Bean (2018), p. 4; cf. DalleMule & Davenport (2017), p. 4.

¹³ Hagiu & Wright (2020).

¹⁴ Cf. DalleMule & Davenport (2017), p. 4.

¹⁵ The agricultural industry is divided into four main sub-sectors: Crops; Livestock; Fisheries/ Aquaculture and Forestry. Cf. Global Facility for Disaster Reduction and Recovery (2012), p. 1.

¹⁶ Cf. DalleMule & Davenport (2017), p. 4.

¹⁷ Top Cloud-Agri Technology Co., Ltd. (TPYN) is a Chinese company in the agricultural industry which has transformed from its traditional business model to a leading digital provider in the agriculture industry.

¹⁸ Cf. DalleMule & Davenport (2017), p. 4–5.

¹⁹ Ibid. p. 4.

²⁰ Single source of truth (see chapter 2.1.1) versus Multiple versions of the truth (see chapter 2.1.2).

²¹ Data Warehouse (see chapter 2.1.1) versus Data Lake (see chapter 2.1.2).

²² Note that the two approaches can be distinguished not only by the aspects mentioned in this paper. However, it is beyond the scope of this paper to provide a holistic analysis of both aspects. Further information can be found in DalleMule & Davenport (2017) : What’s Your Data Strategy?

²³ DalleMule & Davenport (2017), p. 4.

²⁴ Cf. ibid.

²⁵ In Appendix I the main components are highlighted.

²⁶ DalleMule & Davenport (2017), p. 6.

²⁷ In Appendix II the main components are highlighted.

²⁸ DalleMule & Davenport (2017), p. 5.

²⁹ The digital twin is an intelligent digital representation of a real product or process.

³⁰ Cf. Ayaz et al. (2019), p. 129551.

³¹ Cf. Weltzien (2016), p. 66.

³² Synonyms are Agriculture 1.0, Agriculture 2.0, Agriculture 3.0 and Agriculture 4.0.

³³ Maturity level of intelligence and autonomy.