STRATEGIC INNOVATION IN SUPPLY CHAIN MANAGEMENT: HOW BLOCKCHAIN IS CHANGING THE RULES OF THE GAME
Martina Galli ID 10557121
M.Sc. in Business Administration: Strategy track Thesis Supervisor: Hesam Fasaei
Master Thesis Version: Final Version Date of submission: 22nd of June, 2018
2
Statement of originality
This document is written by Martina Galli who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of contents
1. INTRODUCTION ... 6
2. LITERATURE REVIEW & THEORETICAL FRAMEWORK ... 9
2.1 BLOCKCHAIN... 12
2.2 THE RULES ... 14
2.3 APPLICATIONS ... 15
2.4 SUPPLY CHAIN MANAGEMENT ... 16
2.5 TOWARDS A NEW BUSINESS MODEL ... 19
2.6 VALUE PROPOSITION ... 20 2.7 CUSTOMER ... 22 2.8 INTERNAL PROCESSES/COMPETENCES ... 23 2.9 EXTERNAL POSITIONING... 24 2.10 ECONOMIC MODEL ... 25 2.11 PERSONAL/INVESTOR FACTORS ... 26 3. RESEARCH METHODOLOGY ... 29
3.1 EXPLORATION & CASE SELECTION ... 31
3.2 VALIDATION ... 34
3.3 CASES DESCRIPTION ... 36
4. RESULTS ... 38
5. DISCUSSION ... 50
5.1 HYPOTHESES TESTING ... 50
4
5.3 THEORETICAL AND MANAGERIAL CONTRIBUTIONS ... 57
5.4 LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH ... 59
6. CONCLUSION ... 60
7. REFERENCES ... 65
APPENDIX 1: GARTNER’S HYPE CYCLE ... 70
APPENDIX 2: RESEARCH HYPOTHESES ... 71
APPENDIX 3: OVERVIEW ON COMPANIES AND EXPERTS ... 72
APPENDIX 4: INTERVIEWS STRUCTURE ... 74
APPENDIX 5: NVIVO NODES ... 76
APPENDIX 6: CROSS-CASE ANALYSIS ... 77
5
ABSTRACT
Today, 90% of goods in global trade are carried in the shipping industry, but the supply chain is slowed by the complexity and sheer point-to-point communication across loosely coupled transportation providers, freight forwarders, customs brokers, governments, ports, and ocean carriers (Robinson, 2016). Blockchain has emerged as a solution to create a global, tamperproof system for digitalizing workflow and tracking shipments end-to-end, eliminating frictions including costly point-to-point communication (Robinson, 2016). This paper investigates how blockchain changes the traditional supply chain business model. The supply chain up until this moment has worked as a linear movement of goods, paperwork and communications from the manufacturer to the end consumer. This paper builds on the ecosystem value chain of Weill & Woerner (2015) to argue that blockchain changes the supply chain into an ecosystem where resources, paperwork, and communications are digitalized and shared among all participants. Using Morris, Schindehutte & Allen’s (2005) work, this paper contributes to the literature with a conceptual model in which the impact of blockchain on value proposition, customers, internal processes/competences, external positioning, economic model, and personal/investors potential is analysed. The results show that the interpretation of the supply chain structure is largely subjective; it might be interpreted as a network, a linear process, or as a monitoring mechanism. Furthermore, blockchain resolves many organizational inefficiencies such as agency problems, low cybersecurity, bottlenecks in production, and low consumer trust. However, it was shown that there are instances in which a regular IT system works better than blockchain and that blockchain’s efficacy is largely dependent on the fit with the organization, economic activity, urban infrastructure and political circumstances.
Keywords: Blockchain, IT, Disruptive Technologies, Strategic Innovation, Supply Chain Strategy.
6
1. INTRODUCTION
“No matter what the context, there’s a strong possibility that blockchain will affect your business. The very big question is when” - Harvard Business Review
When we vote, how can we tell our vote has really been counted? When we meet someone online, how can we confirm the identity of that person? When we buy coffee with a “fair trade” label, how can we really tell where that coffee came from? (Institute for the Future, 2016). To ensure these concerns are addressed, we need a system where records can be stored, transactions verified, and security guaranteed (Institute for the Future, 2016). Blockchain technology promises to address these concerns (Iansiti & Lakhani, 2017). With blockchain technology, every agreement, payment or process has a digital identity and a signature that can be tracked, verified, stored and shared with other accounts (Iansiti & Lakhani, 2017). Those professions that function as intermediaries in a transaction, such as lawyers, brokers and bankers, may no longer exist one day. Individuals, companies, and algorithms will freely transact with little friction (Iansiti & Lakhani, 2017). This is the true potential of blockchain technology.
Previous research is superficial and fragmented given that it is limited to discussing cost-benefit analyses and it covers one application per time, including commodities (Apte & Petrovsky, 2016), food safety (Aung & Chang, 2014; Tian, 2016), high tech (Hackius & Petersen, 2017), and retailing (Chakrabarti & Chaudhuri, 2017; Badzar, 2016). While this direction of research is relevant and aligns with global trends, the papers following it are too many and consequently the findings result monotone. Furthermore, blockchain is a new technology that has been introduced in the business context only recently. While the technology has been in circulation for a decade, it is entering the business context only in recent times. Consequently, the research on blockchain for business results scarce. Apart from this, an additional limitation is that blockchain is in constant development and the existing literature is likely to be outdated fast.
7
The literature on blockchain needs greater originality and depth of analysis. Moreover, the scarcity of literature and blockchain’s continuous development require a new, continuous effort by the researchers. An example of a creative approach is Catalini & Gans (2016), which investigated how blockchain changes the structure of the market. Touching a few principles in economics, the authors discover that blockchain reduces the costs of verification and networking. Taking into account the limitations highlighted and the broad view of Catalini & Gans (2016), the following research question arises: How is blockchain changing the traditional business model of the supply chain?
The interest in the supply chain sector was triggered by the joint venture between IBM and Maersk that was launched in the fall of 2017 (Churchill, 2017). The supply chain represents “all the links involved in creating and distributing goods, from raw materials to the finished product that goes into the possession of the consumer” (Dickson, 2016, para. 4). The complexity of a supply chain increases when there are many stages of production and geographic locations involved (Dickson, 2016). This research proposes a conceptual model based on Morris, Schindehutte & Allen’s (2005) study on business models that analyses various dimensions of the supply chain, including value proposition, customer, internal processes/competences, external positioning, economic model, and personal/investor factor. Each dimension contributes to the overall profitability of a firm, therefore it is important to understand the impact of blockchain on each one.
This research is qualitative and make use of semi-structured interviews and secondary data such as white papers, academic journals, and scientific articles for the data collection. The data is analysed through comparative analysis, consisting of searching for similarities and differences across the data employed (Harding, 2013). Qualitative research and semi-structured interviews were preferred because exploratory research makes more sense in instances where the existing literature is scarce (Tracy, 2010), while semi-structured interviews allow greater freedom of expression for respondents and empower the interviewer to guide the discussion (Leech, 2002). The interviewer can also decide to extend the discussion on a specific topic through prompt questions (Leech, 2002).
8
The research process was divided into three stages: exploration, case selection and validation. Exploration served to gather an overview on the research articles and academic studies on blockchain, while the purpose of case selection was to identify the companies that held enough expertise to answer the research question. Validation consisted of comparing different sources of information on blockchain and identifying common patterns and similarities. The reliability of data was ensured by recording the interviews on tape and transcribing them on paper. The interview transcripts will be made available online to ensure that future researchers can retrieve them.
The conceptual model was linked to the theories taught in strategy and digital business (i.e. disruptive technologies, circular or ecosystem business models, the hype cycle of Gartner, interest alignment and agency theory) in order to propose ten hypotheses. The research design is based on a single, holistic case study: blockchain. In order to analyse blockchain and the supply chain, three companies with extensive experience and different participation roles were chosen: IBM, Maersk, and Seal Network. From each company, a group of professionals was interviewed, which included experts on blockchain, experts on logistics, and users of blockchain products. IBM, Maersk and Seal Network operate in international trade, software development and consumer services respectively. Comparing different operational backgrounds and different perspectives on blockchain was useful to strengthen the research credibility and make the findings generalizable.
This thesis contributes to the academic and managerial literature in a few ways. First, it adds structure and provides a broad overview on blockchain and the supply chain. Previous research was superficial and fragmented provided that it is limited to specific industries and to analysing certain aspects of blockchain and the supply chain. It failed to consider the supply chain as a whole. The framework proposed in this paper considers companies with different operational backgrounds and it extends the analysis of this topic to six dimensions, including aspects of an economic nature such as competitive positioning, economic model and investors attractiveness. Second, it extends the research on blockchain and shares insights from a private company and two open corporations. Testing blockchain is useful for the development and expansion of the technology. As firms get to
9
know more about it, they make an effort to understand it and implement it. Consequently, those firms that implement it successfully will maximize their efficiency of business operations and increase their profitability. Third, it brings attention to the phenomenon of digitalization. As Weill & Woerner (2013) claim, digitalization is a phenomenon that cannot be arrested and companies that fail to adapt and digitalize their business models will soon find themselves out of the market. This research could encourage companies to invest in information technology (IT) and motivate them to innovate their operational systems.
In addition, managers could find this research useful in assessing their match with blockchain technology. For the companies that are able to benefit from it, blockchain becomes a strategic tool to increase the efficiency of business operations, increase the security of sensitive data, reduce the frictions between the management and employees, increase consumer trust in the firm’s products, fight counterfeit activities and recover losses in productivity.
The paper is structured as follows. The next section examines digitalization, disruptive technologies and emphasizes the importance of IT strategic renewal. It then passes to discuss what is blockchain, how it works and to which industries it can be applied. Other than that, it focuses on the supply chain sector and discusses multiple applications to the supply chain. Once the research focus is established, it introduces the conceptual framework used to analyse blockchain and the supply chain. The conceptual framework proposed is based on six business model dimensions and for each dimension it puts forward one or more hypotheses to test. The third section presents the research design and its stages: exploration, case selection, and validation. It explains the methods for data collection and analysis and describes the companies that were interviewed. It also explains how the research intends to attain data reliability and validity. The fourth section visualizes the words frequency and reports the relevant quotations from the interview transcripts. The fifth section discusses the results, the extent to which these answered the research questions, the contributions to the academic and managerial literature and the limitations and suggestions for future research. Then, the concluding remarks are presented.
10
2. LITERATURE REVIEW & THEORETICAL FRAMEWORK
In recent years digitalization introduced innovations such as blockchain, cryptocurrencies, big data and Internet of Things (IoT). Digitalization originated from three trends that begun in the 60s: the fast-paced development of information technology; the increasing IT literacy across generations; and the passage from “word of mouth” to “word of mouse” (Weill & Woerner, 2013). Its progress transferred business practices from a physical world, tangible and oriented towards consumer transactions, to a digital world, where content, packaging and infrastructure are intangible and structurally different (Weill & Woerner, 2013). Content is virtually distributed and no longer requires physical space to be sold. Property rights changed. In some cases, such as journal articles, the buyer purchases the right to access a resource, but not a physical copy of the article itself. This right is temporary and requires a new monthly or annual payment to continue to enjoy the product. Generations born in this digital age begin using digital devices in their childhood. Learning at a younger age is more effective than learning at a later stage of life, conferring these individuals an affinity with technology unimagined before (Weill & Woerner, 2013). The natural consequence of close acquaintance with technology is the incremental usage of digital channels of communication (Weill & Woerner, 2013). Consumers are now able to exchange opinions and reviews on the products sold online. Moreover, they can confront a broad range of products in a matter of seconds, thereby speeding the whole process of purchasing. The combination of increased communication and broader choice result in a higher consumer bargaining power (Quix & van der Kind, 2016). While in the physical world consumers spend time and energy to move from one store to another, in a digital world they can visit multiple stores by clicking on an ad, external link, or by typing a store’s name in one of the web search engines. The combination of these three trends increased the difficulty of keeping a business into place (Weill & Woerner, 2013); consumers are confronted with better means of comparison across stores and have much lower switching costs between brands (Quix & van der Kind, 2016).
11
The threat of digitalization is further increased by the rise of disruptive technologies; technologies that “create new business opportunities and destroy business models that have been successful for a long time” (Weill & Woerner, 2015, p. 31). If a firm does not have the readiness and the competences to dominate a rising disruptive technology, these will soon experience losses and will be driven out of business. The distinguishing features of disruptive technologies are two: these enter the market with a set of attributes that consumers do not value; and continue to improve until invading the established markets (Clayton Christensen cited in Gallaugher, 2017). An example of a disruptive technology is Polaroid and the introduction of digital photography. Polaroid was the first company to introduce instant photography and it was firmly convinced that consumers valued printed photos (Tripsas & Gavetti, 2000). Digital photography entered the market with low quality and limited capacity and consumers were not interested in the new product. However, as the quality and capacity of digital photos increased, consumer demand shifted and Polaroid did not manage to offer an equally valuable digital product (Tripsas & Gavetti, 2000). Other examples of disruptive technologies include steam engine, mobile phones, and e-books. Critical to cope with disruptive technologies is the ability to realize their potential on time and develop the competences required to compete in the new market.
Digitalization and the emergence of disruptive technologies suggest an urgency for firms to understand and embrace IT. In the face of these threats an aspect that firms should not forget is to maintain an active process of strategic renewal. Strategic renewal “includes the process, content and outcome of refreshment or replacement of attributes and organizations that have the potential to substantially affect its long-term prospects” (Agarwal & Helfat, 2009, p. 282). As Agrawal & Helfat (2009) argue, strategic renewal may involve major transformations in the firm’s core business (radical) or continuous adjustments through routines, organizational structure, and incentives to conduct ongoing renewal activities (incremental). Although radical and incremental strategic renewal imply different levels of risk, both processes are relevant and should be chosen according to the situation. IBM is a company that engaged numerous times in both activities. The company
12
first transformed itself from an electromechanical accounting equipment company into an electronic computing company during the years 1940 to 1965 (Agarwal & Helfat, 2009). In recent years, it switched from a hardware-based computing company with a substantial personal computer business to a business computing services company (Agarwal & Helfat, 2009). In addition, IBM never stopped innovating itself by means of “. . . small, frequent investments and . . . learn from them” (Herrald et al. 2007, p. 41; cited in Agarwal & Helfat, 2009).
Blockchain arose quietly and is rapidly improving and finding new applications across different industries. In future blockchain might be preferred over standard IT. Therefore, it deserves a spot light in the academic literature for the purposes of enriching our academic and managerial understandings.
2.1 Blockchain
Blockchain was introduced by Satoshi Nakamoto in the white paper on Bitcoin in 2008 (Kakavand & Kost De Sevres, 2016). It has been defined as a “distributed, shared, encrypted database that serves as an irreversible and incorruptible repository of information” (Wright, 2015; cited in Kakavand & Kost De Sevres, 2016, p. 4). Instead of simply exchanging digital money between two parties, the technology goes a step further: it records all the transactions executed without the supervision of a third party. As Natamoto (2008) explains, “what is needed is an electronic payment system based on cryptographic proof instead of trust, allowing any two . . . parties to transact . . . with each other without the need for a trusted third party” (p. 1). Satoshi Nakamoto recognized the importance of trust in financial transactions and sought a manner to replace it with computing power. Financial transactions are validated by means of solving a complex mathematical puzzle; those transactions that are “computationally impractical” will not be recognized by the system (p. 1).
13
Blockchain is strongly associated with cryptocurrencies and in particular with Bitcoin (Crosby, Pattanayak, Verma & Kalyanaraman, 2016), however it has a much broader scope and it can be employed in numerous industries. Even if it originated in the public domain as a system to handle financial transactions, it was adapted to the business context and renamed “permission blockchain”, or “blockchain for business” (Crosby, Pattanayak, Verma & Kalyanaraman, 2016, p. 13). The difference between blockchain and permission blockchain or blockchain for business is that the first is publicly accessible and users identities are semi-private, while the latter has limited accessibility and users identities are known (Crosby, Pattanayak, Verma & Kalyanaraman, 2016). Semi-private identification means that the names of users are not explicitly stated, but users identity can be traced back by means of an eight-digit code assigned to each user (Crosby, Pattanayak, Verma & Kalyanaraman, 2016). Permission blockchain is blockchain adapted to manage other types of transactions, including the exchange of documents, certifications, signatures, and product testing results (Crosby, Pattanayak, Verma & Kalyanaraman, 2016). Besides finance, blockchain expanded to energy, healthcare, pharmaceuticals, and supply chain management.
Blockchain for business brings many advantages to industries. For example in supply chain and logistics this system allows the shortening of time and costs of delivering a product to consumers (Robinson, 2016). Even if blockchain does not physically bring the product from the supplier to the consumer, it facilitates the exchange of documents and approvals between one stage to another of the supply chain (Robinson, 2016). In large supply chains the complexity of communication and exchange of documents and approvals easily results in slower times of delivery and increased financial costs (Abeyratne & Monfared, 2016). If these same operations can be virtually executed in a safe and trustworthy manner, there is great potential for reducing costs and increasing productivity. Furthermore, blockchain improves the visibility of the activities occurring in the supply chain (Robinson, 2016), since the peer-to-peer system shares all the information and paperwork with the users of the platform. Greater visibility and transparency allow fraud prevention (Robinson, 2016). The transactions recorded in the ledger cannot be altered or deleted, ensuring that
14
no user in the system can ever cheat the platform. Blockchain is secure “as long as honest nodes collectively control more power than any cooperating group of attacker nodes” (Nakamoto, 2008, p. 1).
A major limitation is the immaturity of the technology. The excitement over blockchain is growing faster than its concrete development, resulting in widespread expectations that do not meet reality yet (Iansiti & Lakhani, 2017). In the case of Bitcoin the popularity of the cryptocurrency and the large volume of users slowed down the process of resolving mathematical puzzles – hence verifying transactions. At the moment, it takes around 15 minutes for miners to verify a single transaction (Norman & Antonopoulos, 2018). Furthermore, the concept of digital money has raised concerns on security. Digital wallets and everything they contain are protected by a private key – or password – that is in principle vulnerable to hacker attacks (Norman & Antonopoulos, 2018). Digital currencies have also been embraced by narcotics traffickers, and consequently have earned a negative reputation worldwide. Narcotics traffickers took advantage of the novelty and unregulated status of cryptocurrencies to conduct their business undisturbed. Cryptocurrencies have now been regulated and are accepted as a method of payment in many countries including China, Japan, the U.S. and Russia (Norman & Antonopoulos, 2018). Many other countries including Algeria, Bolivia, Ecuador, Nepal, and Cambodia banned cryptocurrencies (Norman & Antonopoulos, 2018).
2.2 The rules
Intangible goods including digital currencies, documents, certificates, authorizations, and test results are transferred in the blockchain network by means of digital signatures (Nakamoto, 2008). In order for a good to be transferred, all servers have to agree that the good exists and is in possession of the account holder that wishes to transfer it (Nakamoto, 2008). For example, person A wishes to transfer 25 Bitcoins to person B. We can only be sure that person A has this amount of money by checking the last mentioned transactions and by looking for a hint that this amount was already spent. This hint is given by the timestamp, a sequence of characters that reveals the day and
15
time a certain sum was spent (Nakamoto, 2008). Therefore, through the timestamp server, it is possible to verify how much virtual cash is present in the wallet of person A.
Figure 1: Timestamp Server (Nakamoto, 2008, p. 2).
Each timestamp “includes the previous timestamp in its hash, forming a chain, with each additional timestamp reinforcing the previous one” (Nakamoto, 2008, p. 2). The timestamp hash chain is our blockchain itself, with each block proving that a transaction exists and validating it. With each block newly added, the volume of the blockchain increases. The computing power needed to manage transactions increases with the length and volume of the chain of blocks (Nakamoto, 2008). When Satoshi Nakamoto introduced Bitcoin, a regular computer could process transactions efficiently; however, now the extensive length of the chain requires extremely powerful computers to process the transactions.
To spread the practice of reviewing timestamps on a large network of computers (referred to as “nodes”), a proof-of-work system is implemented to ensure that the nodes are validating the transactions properly (Nakamoto, 2008). The functioning of the proof-of-work system is best explained by looking at the previous example. When person A orders the transfer of 25 Bitcoins to person B, the timestamp server generates a hash (or a code specifying the transaction details) and assigns it a random and unique number containing a puzzle to solve. The nodes then compete to find an answer to the puzzle and the first one arriving to the correct answer receives a compensation (Nakamoto, 2008).
16
Figure 2: Proof-of-Work (Nakamoto, 2008, p. 3).
The answer has to be compatible with the hash that generated the random number in order for the nodes to accept the transaction order (Nakamoto, 2008). If the solution is compatible with this number, the transaction is approved and therefore allowed to take place. In the opposite situation the transaction is rejected and therefore deemed invalid. This process of solving a puzzle locks the blocks of the chain together. It is technically possible to reverse a transaction by identifying the answer that was previously given by the nodes and removing it (Nakamoto, 2008). However, the blocks in the chain are connected in such a way that encoded information on a block becomes available throughout the whole chain. To eliminate a single block, one must work across the entire chain of blocks to find this encoded information and remove it. The longer the chain of blocks, the safer the entire system becomes. In a long chain it becomes impractical to go through each block to remove a single transaction or group of transactions (Nakamoto, 2008). Therefore, blockchain results immutable and consequently trustworthy.
2.3 Applications
Blockchain became an interesting innovation for many sectors including finance, healthcare, pharmaceuticals, energy and supply chain management. A commonality between all these sectors is that the management of large scale operations is based on paperwork. The problem with relying on paper is that it causes many added expenses and delays, and makes it easier for crime to divulgate (Forbes, 2017). The financial sector is leading the way in terms of high tech innovation and relies on blockchain for the clearing and settling of financial assets, commercial financing, operational risks, smart contracts and other operations (Laurence, 2017). Blockchain proved to be useful also in healthcare, where it is used to optimize the management of medical records, the authorization of payments, the settlement of insurance claims, and other operations (Laurence, 2017). Most importantly, blockchain presents a secure solution to store sensitive, private information in healthcare such as Electronic Medical Records (EMRs). With its “shared, immutable and
17
transparent history of all the transactions” (Dubovitskaya, Xu, Ryu, Schumacher & Wang, 2017, p. 1), blockchain allows the development of a secure and trustable EMRs data management and sharing system. Another critical sector is the pharmaceutical one, where sometimes shortening the time between drug production and delivery to the consumer may save a human life (Kurki, 2016). This sector is subject to strict governmental regulation and each drug must obtain a patent before being commercialized. Obtaining a patent usually requires a couple of years and blockchain may effectively speed up the entire process (Kurki, 2016). In addition, academic research suggests to employ blockchain to set up a global market to exchange energy (Mannaro, Pinna & Marchesi, 2017). The idea originated from the "inadequacy of the regional energy distribution system in relation to the new demands arising from the advent of distributed production" (Mannaro, Pinna & Marchesi, 2017, p. 2). Users would purchase tokens on the blockchain-powered platform to buy and sell energy according to their needs.
Even though every sector is fascinating, this research focuses on supply chain management. This choice was inspired by the joint venture between IBM and Maersk, finalized to uncover the possibilities of gaining efficiency in supply chain operations.
2.4 Supply chain management
The literature on blockchain and the supply chain examines its potential benefits for manufacturers (Apte & Petrovsky, 2016), food suppliers (Aung & Chang, 2014; Tian, 2016), scientists (Hackius & Petersen, 2017), retailers (Chakrabarti & Chaudhuri, 2017), and logistics operators (Badzar, 2016). Blockchain is a simple way of passing information from one user to another, whereby the user receives complete and non-falsifiable records of transactions (Apte & Petrovsky, 2016). Through tagging and registering a product across the value chain, it is possible to verify how and where a product was assembled and made (Apte & Petrovsky, 2016). Therefore, blockchain denies the market for illegal and counterfeit products. This is a major advance for the reliability of excipient
18
supply chains and consumers’ trust in these, but it does not eliminate the need for quality audits (Apte & Petrovsky, 2016). Besides luxury products and the counterfeit economy, the transparency and reliability of supply chains helps food providers enormously. When it comes to food, a perishable commodity, consumers “call for high quality food with integrity, safety guarantees and transparency” (Aung & Chang, 2014, p.173). By attaching a radio-frequency identification (RFID) device to a chocolate cake and by scanning the device, consumers will receive information concerning the production, processing, distribution and retail of that chocolate cake (Tian, 2016). RFID and blockchain technologies guarantee the standards of quality and safety of food “from the farm to the fork” (p. 6). Moreover, blockchain is considered a solution to connect and manage IoT devices reliably (Hackius & Petersen, 2017). IoT will concern several devices in the future, including vehicles, shipments, etc., therefore supply chain might be one of the most promising applications of blockchain. Walmart, for example, was recently granted a patent to improve last mile logistics by connecting delivery drones to blockchain (Hackett, 2017; cited in Hackius & Petersen, 2017). In the retail sector blockchain can be used to assist retailers in finding any lost or damaged product in the shipments (Chakrabarti & Chaudhuri, 2017). Retailers will have access to all the details on the delivery of a product, such as the date and time, the personal details of who’s handling the shipment, temperature, condition of the package/product, and more (Chakrabarti & Chaudhuri, 2017). The technology can also be used to build customer profiles based on each consumer’s transactions and use them to improve marketing and promotions. In addition, these might be shared between stores in order to collect deeper analytics on customer records (Chakrabarti & Chaudhuri, 2017). Furthermore, Badzar (2016) argues that smart contracts can be used to automatically validate and execute the sustainability clauses in transport contracts. The benefit of this application is twofold: on the one hand, CO2 emissions in the environment will be lower; on the other hand, firms will save the costs of contracting, documentation and enforcement of contracts (Badzar, 2016).
19
Finally, it was demonstrated that the principal purpose for using blockchain in the supply chain is to improve the flow of information (Hackius & von See, n.d.). 88% of managers considers the blockchain ledger and smart contracts the most important features, while time stamping is considered as a means to integrate business to business (B2B) and IoT (M2M) transactions (Korpela, Hallikas & Dahlberg, 2017).
2.5 Towards a new business model
A business model is a “concise representation of how an interrelated set of decision variables in the areas of venture strategy, architecture, and economics are addressed to create sustainable competitive advantage in defined markets” (Morris, Schindehutte & Allen, 2005; cited in Zott, Amit & Massa, 2011, p. 727). According to Morris, Schindehutte & Allen (2005), a business model is composed of few elements such as value proposition, customer, internal processes/competencies, external positioning, economic model, and personal/investor factors (p. 727). Up until this moment, the business model used in the supply chain resembled figure 3.
Figure 3: Traditional supply chain.
The process of value creation is clear: the supplier provides the raw materials, the manufacturer transforms these into commodities, the distributor brings the products to a group of stores and the stores sell these products to consumers (Abeyratne & Monfared, 2016). The price that consumers pay then goes to the firm, which makes a profit once it pays off its labour and administration costs (Abeyratne & Monfared, 2016). A layer of paperwork travels together with the products, whose SUPPLIER MANUFACTURER DISTRIBUTOR RETAILER CONSUMER
Physical products +
Paperwork
20
speed might be faster or slower (Dickson, 2016). Communication is sporadic. Each block communicates with the blocks nearby in the supply chain. This simple design becomes more complicated once the supply chain expands over numerous stages and geographic locations (Dickson, 2016).
Assessing the impact of blockchain on the supply chain business model is complicated by two factors. First, there is no single identity for the supply chain; it is an assemblage of physical goods, busienss processes, paperwork, firm competences, competitive strategy and communication (Dickson, 2016). Second, the complexity increases with the length and geographical expansion. A supply chains might involve dozens of production stages and it might extend across many countries (Dickson, 2016). The solution is to analyse the supply chain along the dimensions proposed in the definition of business model (Morris, Schindehutte & Allen, 2005): value proposition, customer, internal processes/competencies, external positioning, economic model, and personal/investor factors. The conceptual framework proposed simplifies the analysis of the supply chain; it gives structure to the research; it allows a comprehensive analysis; and it leads to targeted, concise suggestions for managerial practices. The framework can be seen below:
Figure 4: Conceptual Framework.
2.6 Value proposition
BLOCKCHAIN SUPPLY CHAIN MODEL
Value proposition Personal/Investor factors External positioning Economic model Internal processes/ Competences Customer influence
21
Value proposition is the value that the firm is offering to consumers (Morris, Schindehutte & Allen, 2005). It includes “the nature of the product/service mix, the firm’s role in production or service delivery, and how the offering is made available to customers” (p. 729). More organizations cooperate in a supply chain: each acts as a customer when buying materials from the suppliers and as a supplier when delivering products to consumers (Waters, 2003). In this journey, materials might move through “raw materials suppliers, manufacturers, finishing operations, logistics centres, warehouses, third party operators, transport companies, wholesalers, retailers, and a whole range of other operations” (p. 8). Occasionally, the products go beyond the end consumer for its materials to be recycled and reused. Materials are accompanied by paperwork along a linear pattern with each stage of the supply chain confined to its operations and administrative tasks (Waters, 2003). Access to information is limited for each stage. For example, retailers cooperate with wholesalers and consumers, but it is uncommon for them to do so with the other supply chain participants, such as manufacturers and logistics centres.
Digitalization changes the formula for value creation and delivery. Digitalization breaks down industry barriers and creates new opportunities, while destroying long successful business models (Weill & Woerner, 2015). This process is also described as digital disruption (Weill & Woerner, 2015; Clayton Christensen cited in Gallaugher, 2017). With digitalization documentation becomes intangible and physical movements are replaced by virtual transfers (Weill & Woerner, 2013). As Weill & Woerner (2015) argue, organizations are increasingly opting for digital ecosystems and the processes that once seemed linear, nowadays resemble a network of organizations. Moore (1993), pioneer of this concept, explains business ecosystems as “a network of interdependent organizations that coordinate each other in order to create success” (cited in Sherwani & Tee, 2018, p. 15). Based on the arguments of Weill & Woerner (2015) and Moore (1993), it can be inferred that with blockchain the linear movement of goods in the supply chain becomes a circular process, or a digital ecosystem.
22
2.7 Customer
Every company has to define who they wish to sell their products to, or the customer target (Morris, Schindehutte & Allen, 2005). Trust in business transactions is fundamental in order to maintain a firm’s economic activity profitable (Jones, 1995). However, firms are confronted with two challenges: digitalization and low consumer trust. Digitalization made it more difficult for firms to maintain their businesses (Weill & Woerner, 2013) since consumers gained bargaining power (Quix & van der Kind, 2016). Consumers gained better access to information and their switching costs between brands decreased.
Additionally, digitalization placed firms business operations and marketing strategies under the spot light (Quix & van der Kind, 2016). It is not uncommon for firms to exaggerate the merits of their products in their marketing strategy. In the digital era this type of “cheating” is no longer possible because consumers will find out (Quix & van der Kind, 2016). As a consequence, the word of companies and institutions is no longer considered reliable. Only 25% of consumers believes that the information provided by companies and institutions is trustworthy (Mintel European Consumer Trend Report, 2018).
Several authors (Aung & Chang, 2014; Apte & Petrovsky, 2016; Robinson, 2016; Tian, 2016) suggest that firms can take advantage of digitalization and that low consumer trust is reversible. Blockchain records transactions in a shared ledger between the platform users and renders these records immutable (Robinson, 2016). Apte & Petrovsky (2016) claim that manufacturers can increase trust by tagging and registering their products in the blockchain platform and making these records available to consumers. This procedure was recommended also for food (Aung & Chang, 2014; Tian, 2016), where asymmetric information causes great discontent among consumers.
Hypothesis 1: The creation and delivery of value is changing from a linear process to a circular process or “ecosystem”.
23
Therefore, the second hypothesis will be:
2.8 Internal processes/Competences
Internal processes are the activities that a firm performs regularly (Morris, Schindehutte & Allen, 2005). These activities involve large flows of paperwork and information. In a physical world, it takes time to move paperwork and information around; however, in a digital world, the times of delivery are significantly shortened (Weil & Woerner, 2013). Digitalization alone is not enough. Organizations must invest in strategic renewal in order to maintain their operations efficient (Agarwal & Helfat, 2009).
As compared to a traditional system, blockchain offers few additional features. Vulnerability is reduced. Shared visibility and ownership of assets allow any user to retrieve documentation and information when needed (Crosby, Pattanayak, Verma & Kalyanaraman, 2016). In addition, smart contracts enhance operational efficiency by triggering automatic transactions when the required contractual obligations are performed (Crosby, Pattanayak, Verma & Kalyanaraman, 2016).
Therefore the expectation is that:
A core competence is “an internal capability or skill set that the firm performs relatively better than the other firms” (Hamel, 2001; cited in Morris, Schindehutte & Allen, 2005, p. 730). These competences, especially when combined with other valuable capabilities, can grant firms a competitive advantage (Prahalad & Hamel, 2000). It is hard to predict how blockchain will impact these capabilities, since there is no single capability or skills set falling in this category. A core
Hypothesis 2: Blockchain is expected to increase consumer trust in the market through shared visibility and fixed recordkeeping.
Hypothesis 3a: Blockchain is expected to increase the efficiency of internal processes within the supply chain.
24
competence may assume disparate forms, including technical skills, managerial skills, or a specific organizational culture, for example (Prahalad & Hamel, 2000).
The literature identifies few areas that blockchain is able to improve, including greater visibility and transparency, greater operational efficiency, fraud prevention (Robinson, 2016), secure storage of documentation and assets (Abeyratne & Monfared, 2016), and immutability of business transactions (Nakamoto, 2008). It is logical that when a firm’s core competences are related to one or more of these areas, blockchain will strengthen them. The extent to which blockchain can enhance a firm’s core competences depends on the number of areas that it influences.
Thus, the following hypothesis is:
2.9 External positioning
The positioning school of thought has long sustained that firms gain a competitive advantage when they can position themselves better than their competitors in the industry. Porter (1979) leads this school of thought with his five forces model. According to this model, a firm’s positioning is determined by five forces: the threat of new entry, consumer bargaining power, the threat of substitute products, supplier bargaining power, and competitive rivalry (Porter, 1979). A firm can improve its positioning with three generic strategies, namely cost-leadership, differentiation and focus (Porter & Advantage, 1985). Firms should either lower the price of products, increase the quality of products, or focus on a niche market. Low prices attract new consumers and drive them away from competitors. Quality encourages consumers to pay a premium price for the products. A niche market is small and specialized in products and services that are far from competitors attention.
The positioning school is often criticized for underestimating the role of time. Christensen & Rosenbloom (1995) sustain that time contributes to determine a firm’s positioning within the
25
industry. When a new product is introduced in the market and consumer demand is high, the first firm that gathers the attention of consumers is likely to make substantial returns. Until the product introduced is new and unique, there are no substitutes for that product. Therefore, the first mover in the market is free from any competition. Explained in Christensen & Rosenbloom’s (1995) words, the first-mover advantage derives from “detecting an emerging business opportunity and attracting the customers of that market before other firms” (p. 234). The utility of blockchain has been proved in many fields, including finance, healthcare, energy, pharmaceuticals, and supply chain management (Laurence, 2017). Since it is a new concept in the business context (Iansiti & Lakhani, 2017), it is likely that the first firm which invested in it is now occupying a better position than its competitors in the industry.
The fourth hypothesis will be:
2.10 Economic model
At the basis of a firm’s business model there is its economic model (Linder & Cantrell, 2000; cited in Morris, Schindehutte & Allen, 2005). The concern is with generating profits. An economic model is “a statement of how a firm will make money and sustain its profit stream over time” (Stewart & Zhao, 2000; cited in Morris, Schindehutte & Allen, 2005, p. 727). In determining profits, the sources of revenue, pricing strategy, cost structure, margins expected, and volume of sales matter. A firm can influence its economic model positively by either increasing revenue or reducing costs (Morris, Schindehutte & Allen, 2005).
As Robinson (2016) explains, blockchain increases operational efficiency by reducing the time and costs of moving documents and information around the supply chain. Costs are reduced because firms no longer need customs to inspect and release their products across borders (Robinson, 2016). Automated, trustworthy transactions substitute the necessity to spend money on
Hypothesis 4: The first firm which commercialized blockchain for business is likely to occupy a better position in the market than the other firms.
26
inquiries and approval releases. Time is reduced because the exchange of documents and information occurs instantaneously. In addition, visibility and transparency reduce the chances of incurring into time-consuming, costly errors (Robinson, 2016).
The gains in time and costs are achieved through what Malone, Laubacher and Dellarocas (2010) call “collective intelligence genome”, or peer-to-peer system. In such a system, individuals work together to achieve a common objective either voluntarily or in exchange of compensation. Blockchain uses peer-to-peer participation, meaning that each node is both a publisher and a subscriber (Laurence, 2017). Each participant can send and receive transactions and the data is automatically shared through the entire network.
Hence, the next hypothesis will be:
2.11 Personal/Investor factors
It is important to understand the expected lifetime of a business as well (Morris, Schindehutte & Allen, 2005). Blockchain technology is immature (Iansiti & Lakhani, 2017). A way to grasp what are the expectations concerning its development and peak of productivity is by looking at Gartner’s hype cycle (Appendix 1). Gartner’s (1995) model illustrates the celebrity of a new disruptive technology, from the moment of its introduction to the phase of standardization (cited in Gallaugher, 2017). After the introduction of a new technology, expectations become unrealistically high. Once people realize this, the expectation (perceived value) changes from hype to disillusion (Gallaugher, 2017).
Iansiti & Lakhani (2017) suggest that although blockchain and smart contracts promise to revolutionize the business context, “we are decades away from the widespread adoption of blockchain and smart contracts” (para. 31). The authors emphasize that the promises made by blockchain require institutional buy-in, a “tremendous degree of coordination”, and widespread
Hypothesis 5: Peer-to-peer participation improves firms economic model by reducing administration costs and paperwork and shortening the product’s time of delivery.
27
understanding of the technology (Iansiti & Lakhani, 2017, para. 31). On the other hand, the excitement over blockchain is undeniable. The large body of literature reviewed in this research indicates that the peak of expectations is close. Among all the research articles reviewed, scepticism on blockchain was rare.
Therefore, the expectation is the following:
Revenue streams are influenced by risk and by the firm’s growth rate (Morris, Schindehutte & Allen, 2005). Risk is largely recognized as a negative influence for a business and it comes under multiple scenarios, including conflicting interests (Chakrabarti & Chaudhuri, 2017), counterfeit activities (Apte & Petrovsky, 2016), internal mistakes and security breaches (Dubovitskaya, Xu, Ryu, Schumacher & Wang, 2017).
Conflicting interests will be explained through agency theory. A principal-agent relationship is formed when a party (principal) delegates a task to another party (agent) (Eisenhardt, 1989). The agency problem arises when the agent does not perform as agreed, or if the principal cannot verify the agent’s performance (Eisenhardt, 1989). Blockchain functions as a monitoring mechanism (Chakrabarti & Chaudhuri, 2017): it gives the principal greater control in the relationship through consensus and provenance. Consensus means that blockchain’s network recognizes a business transaction only if its members have approved it (Laurence, 2017). Provenance allows the participants in the platform to know where an asset is coming from and how its ownership status has changed overtime (Laurence, 2017).
The seventh hypothesis will be:
Hypothesis 7: Blockchain proposes an alternative monitoring mechanism to resolve the agency problem.
Hypothesis 6: The blockchain hype is approaching the peak of expectations. It will reach a stage of “disillusionment” in the coming 3-5 years.
28
Provenance can defeat counterfeit activities as well. RFID technology and NFC chips can be used to verify the transactions occurring throughout the supply chain (Apte & Petrovsky, 2016). The origins and processes involved in production are made visible for the firm’s employees and management team. This monitoring mechanism obstacles the introduction of counterfeit goods within the supply chain (Apte & Petrovsky, 2016).
Thus, the eighth hypothesis will be:
This research moreover proposes that blockchain aids data maintenance and security and that it resolves bottleneck problems. Blockchain makes handling information within an organization easier (Hackius & von See, n.d.). Peer-to-peer participation improves communication and the flow of information within the supply chain. Since the data and transaction records on the blockchain are shared with all the participants in the network, the breakdown of one machine or the mishandling of data will not cause an irreversible loss of data (Hackius & von See, n.d.).
For the same reason a cyberattack finalized to access the data and records of one computer will not succeed in the subtraction of intangible assets. Blockchain architecture proposes incentives for the nodes to behave according to the rules (Nakamoto, 2008). As Nakamoto (2008) explains, blockchain is secure as long as a majority of honest nodes is in control. Altering the transactions in the ledger is possible, but it requires an unbearable amount of work and a single, dishonest node cannot succeed alone.
Therefore, the expectations are the following:
Hypothesis 9: Blockchain eliminates the risk of losing critical data or the risk of the latter being subtracted from the firm.
Hypothesis 8: Blockchain reduces counterfeit activity through the implementation of NFC chips and RFID technology.
29
A last situation in which blockchain might come handy is bottlenecks. A bottleneck refers to a situation in which a computer or person is accountable for a large proportion of tasks in their department. Then, when this machine breaks down, or when the person is missing, the firm suffers severe losses in productivity (Chakravorty & Atwater, 2006). Through shared visibility and ownership of assets blockchain allows other machines and people to take over the large amount of work.
The last hypothesis will be:
An overview of all the research hypotheses is available in Appendix 2.
3. RESEARCH METHODOLOGY
The research process adopted in this paper consists of three stages: exploration, case selection, and data validation. Exploration consists of an in-depth analysis of the literature and a preliminary list of candidates that could suit the research. Case selection involves narrowing down the list of participants to a few ideal candidates for the research. Lastly, data validation serves to assess the validity and reliability of the research and will rely on the method of triangulation.
The research stages are shown below:
30
Figure 5: Research Design.
Qualitative approaches are preferable in instances where the information available is limited (Tracy, 2010). The limitations in the existing literature on blockchain and its applications suggest that an explorative approach is preferable. Furthermore, qualitative approaches seek an answer to why and how certain phenomena develop. The research question “how does blockchain change the traditional supply chain business model?” suggests the necessity to carefully examine a real-life phenomenon: blockchain. Therefore, this research is meant to be qualitative.
Case study research focuses on understanding the dynamics of a particular phenomenon (Eisenhardt, 1989b). This type of research can be conducted on a single case study or on multiple case studies, and may include different levels of analysis (e.g. firm, industry, or both levels). The focus of the research is on a holistic case study, the digitalization of the supply chain.
1. Exploration
Conceptual model and long list of candidates
2. Case Selection
General assessment and selection of few candidates
3. Data Validation
Comparison of responses from different angles: experts on blockchain, experts on logistics,
consumers of blockchain products
Literature Analysis
Semi-structured interviews Secondary data
31
Yin (1994) claims that case studies are “rich, empirical descriptions of particular instances of a phenomenon that are typically based on a variety of data sources” (p. 34). Moreover, according to Yin (1994), cases can be historical accounts, but they are more likely to be contemporary explanations of recent events. They have the advantage of relying on multiple sources of evidence and benefitting from the data collection methods used in previous research.
Case studies typically combine data collection methods such as archives, interviews, surveys, and observations (Eisenhardt, 1989b). The evidence may be in words (qualitative), in numbers (quantitative), or it may be both (mixed methods research) (Eisenhardt, 1989b). The preferred methods for data collection will be semi-structured interviews and secondary data (i.e. books, research articles, white papers). The reason for these choices is that this research wishes to investigate certain aspects of blockchain and not a general notion of blockchain; additionally, semi-structured interviews allow the interviewer to intervene with further questions where needed (Leech, 2002).
The central notion is to use case studies as a basis to develop theory inductively (Eisenhardt & Graebner, 2007). Using case studies serves to test the theory in different contexts, or experiments, and the inferences made provide a benchmark for further analysis. While laboratory experiments isolate the phenomena from their contexts, case study research emphasizes the contexts in which the phenomena occur, ensuring that the information collected is rich and genuine (Eisenhardt & Graebner, 2007). Case study research is sometimes deemed “subjective” because it foresees the research and selection of a pre-determined sample. Nevertheless, if done properly, theory building from case studies can be very objective, because the strict adherence to the information provided by the cases maintains the researcher “honest” (p. 25).
The single-case approach was chosen because it addresses the following points. First, because theoretical sampling for single-case research is straightforward. The cases chosen are unusually revelatory, extreme exemplars, or opportunities for unusual research access (Yin, 1994). Second, as compared to multiple cases, a single case study provides qualitative evidence that is
32
richer and more persuasive (Eisenhardt & Graebner, 2007). The case is discussed in detail from beginning to the end, fitting each quotation and peculiarity of the narrative to the emerging theory. Presenting a complete and unbroken narrative for multiple cases is unfeasible; if the researcher relates theory to the narrative of each case, “then the theory is lost and the text balloons” (p. 29). Finally, single-case research enables the creation of complicated theory because the researchers can fit their theory to the many details of a single case (Eisenhardt & Graebner, 2007). Developing theory that is too simplistic might disappoint the readers that would prefer more insights from the research.
The following sections present the research stages of exploration, case selection and data validation.
3.1 Exploration & case selection
During the research and analysis of the literature, many online journals, press releases and newspaper articles were considered. A diversity of sources is recommended to strengthen and enhance the validity of the research (Eisenhardt, 1989b). Newspaper articles and press releases were a focal point given that blockchain is an infant technology and is currently applied on a small scale. Therefore, newspaper articles were a great way to “spot” the firms that have implemented/are implementing it. The selection of candidates was based on their involvement with blockchain and the supply chain. Firms that have been working with blockchain were strongly preferred over others.
The selection process consisted of gathering greater information on the industry, size and organizational culture and assessing the candidates’ compatibility with the research question. The procedure resulted in three companies, operating in software development, international trade, and consumer services. The companies have different sizes and ownership structures: two large public companies with over 10,000 employees and a start-up with 10 to 50 employees. Three to four experts per company were interviewed either in person or by phone. An overview of the companies
33
and experts can be found in Appendix 3. The interviews took place in Amsterdam, Rotterdam, Groningen and Cape Town (South Africa).
As Leech (2002) claims, “there are many types of interviews with many styles of questions, each appropriate in different circumstances” (p. 665). Unstructured interviews resemble a casual conversation, with dynamic topics, steering in unexpected directions. This type of interviews is most suitable when the interviewer knows very little about a certain topic (Leech, 2002). When the interviewer is only interested in counting how many respondents fall into a certain category, structured interviews with closed-ended questions are most appropriate (Leech, 2002). However, sometimes the interviewer might be interested in discovering only specific details. A style that is used in such circumstances is semi-structured interviews with open-ended questions (Leech, 2002). This approach is the most appropriate for this research because the focus is not on blockchain technology in general, but on the digitalization of the supply chain.
The expertise of Leech (2002) offers great help with all the variations of semi-structured interviews. She recommends giving attention to the emotional status of the interviewees, understanding their responses, and organizing the order of questions strategically. McCracken (1988) suggests that in order to break the ice the interviewer may “appear slightly dim” or “play dumb” (cited in Leech, 2002, p. 38). Adopting this strategy is recommended, but without losing one’s own level of expertise in the subject.
Moreover, it is useful to repeat whatever the respondent says in order to remember and avoid interpreting later on. When the response is unclear, it is better to ask for an application of the topic that is being discussed, instead of asking for a repetition of the concept (Spradley, 1979; cited in Leech, 2002). It is better to ask first the easy questions to put the respondent at ease and then proceed to the difficult or inconvenient questions (Leech, 2002).
To gather more information, certain types of questions are more effective, such as grand tour questions and prompts. Grand tour questions involve asking the description of a typical day in a specific environment. They are used to give a sense of how is the daily life of the interviewee
34
(Leech, 2002). Prompts are planned or unplanned questions that were triggered by the curiosity of an argument. They are used to keep the interviewee talking and rescue dry responses (Leech, 2002).
The interviews structure used in this research can be found in Appendix 5. The questions were designed following the suggestions of Eisenhardt (1989b) and Leech (2002). The preferred method of interviewing was face-to-face because it produced the richest and most interesting content. In fact, more factors than just words came into play. The interviewer was able to observe facial expressions and perceive the respondents’ level of comfort.
Besides being useful for data collection, asking questions directly in front of someone empowers the role of the interviewer in the process (Leech, 2002). If the respondent is nervous and skips many details that could be useful, the interviewer could try to put the other person at ease. Phone calls are more limited in this respect. The interviewer cannot grasp the details that come from body language and runs the risk of missing out relevant information. Nevertheless, the interviewer can still guide the conversation. An advantage of phone calls is that they can take place in any location, without the struggle of reaching the location of the interview. This method was preferred for the interviews that took place outside Amsterdam.
3.2 Validation
The validation of research is done through construct validity, external validity, internal validity, and reliability (Gibbert & Ruigrok, 2010). Construct validity is the “extent to which the research investigates what it claims to investigate” (p. 712). The positivist literature suggests two strategies to ensure construct validity: the triangulation method and establishing a clear chain of evidence (Gibbert & Ruigrok, 2010). Triangulation is the “combination of two or more data sources, investigators, methodologic approaches, theoretical perspectives . . . or analytical methods . . . within the same study” (Thurmond, 2001, p. 253). Whereas multiple types of triangulation exist, this research makes use of data triangulation. As Denzin (1970) claims, “the three types of data sources recognized in the literature are time, space, and person” (cited in Thurmond, 2001, p. 253).
35
This research consults three different groups of respondents, each participating to the digital supply chain. The three groups include experts on blockchain, experts on the supply chain, and users of blockchain products. Instead, whether a clear chain of evidence is provided depends on how well the process of data collection is described. The planned and actual procedures, as well as the times and locations of the interviews should be included. Times, duration and the specific roles of the interview candidates are provided in Appendix 3.
External validity or “generalizability” is grounded in the intuitive belief that researchers must prove their theories valid not only for the settings in which they are studied, but also in other settings (Gibbert & Ruigrok, 2010). External validity does not allow statistical generalization, or generalization from observing a phenomenon, but it accepts analytical generalization, or generalization from empirical observations. Therefore, to ensure external validity, a researcher must consider a sufficient number of data sources. It is also important to explain the rationale for choosing determined data sources and give ample details on the context (Gibbert & Ruigrok, 2010). A maximum of five candidates per company was established. Aiming at five interviews per company, a total of twelve interviews was achieved. IBM was the first-mover in the market for permission blockchain and, together with Maersk, created a new product to improve the efficiency of the supply chain. In terms of expertise with this subject, they are the companies that are most informed in the market. Seal Network is a start-up in blockchain products for consumers. Although many start-ups have had the same idea, Seal is one of the few ones with a functioning product in the market.
Also called “logical validity”, internal validity refers to the causal relationship between an independent and one or more dependent variables (Gibbert & Ruigrok, 2010). Establishing internal validity depends on the extent to which researchers manage to build strong arguments in order to support their conclusions. Whether researchers will be persuasive depends on basing their arguments on a “critical investigation of all their data and . . . not . . . on a few well-chosen examples” (Silverman, 2005; cited in Gibbert & Ruigrok, 2010, p. 10). The literature on qualitative
36
research (Denzin & Lincoln, 1984; Eisenhardt, 1989; Yin, 1994) suggests constructing internal validity through three methods. First, drawing diagrams and explanations in words of the relationships observed might help. Second, it is common to contrast the relationships observed with those from previous research and different contexts. A third option finally could be adopting multiple perspectives (triangulation) on the inferences made.
Silverman (2005) describes reliability as ‘‘the degree of consistency with which instances are assigned to the same category by different observers or different occasions’’ (cited in Gibbert & Ruigrok, 2010, p. 210). With qualitative research the problem is that often the reader depends on what the writer is telling about the research. To avoid this problem, he suggests including the data with a minimum of description, instead of including a summary of the data. In the case of interviews this is done by recording all the face-to-face interviews on tape, carefully transcribing the words, using fixed-choice answers, double checking the coding of open ended questions, and providing long extracts of data in the research paper. The intention is to present an accurate and detailed documentation and clarification of the research procedures. The interviews will be recorded on tape and transcribed on a digital document. Then, the interview transcripts, preliminary conclusions, and the narratives gathered during the study will be stored on an online database, so that future researchers will have access to the same type of information.
The next section provides further details and background information on the companies selected.
3.3 Cases description
IBM & Maersk. In 2014, a team of IT specialists from Maersk analysed a shipment of avocados and roses from Kenya to the Netherlands (Churchill, 2017). The purpose of this experiment was to report the maze of physical processes and paperwork that affect every shipment and contribute to the total costs of international trade. The research uncovered that excessive interactions and
