Critical success factors of Blockchain technology adoption for sustainable and resilient operations in the banking industry during an uncertain business environment

Abstract

Blockchain technology (BCT) adoption in banks can bring resilience and make banking operations sustainable in terms of security in times of increased risks and uncertainty. Along this line, this study aims to identify and model the critical success factors (CSFs) of blockchain technology adoption for sustainable and resilient banking operations. Seventeen CSFs were identified from the literature and semi-structured interviews. After that, 15 CSFs were confirmed through the fuzzy Delphi method, and cause-effect relationships were developed through the Decision-Making Trial and Evaluation Laboratory method. Findings of the study highlighted that ease of local and international legislation and regulation and user data privacy are the most significant CSFs of blockchain technology, and these demand a high level of attention from management and decision-makers to realise sustainability and resilience in the banking sector. The study will guide practitioners and policymakers in understanding the importance of CSFs and thereby help devise a strategy for successful blockchain technology adoption in the banking sector.

Introduction

In recent years, the banking industry has experienced disruptions caused by the COVID-19 pandemic, significantly changing its global operations. Banking and credit institutions have faced upheaval in terms of their relationships with customers, retailers, and businesses. Social distancing norms imposed by the government during the COVID-19 pandemic played a key role in the adoption of novel digital technologies, whose adoption has accelerated significantly compared to pre-crisis periods. Closure of several branches has forced banks to adopt digital technologies which can help bring resilience and sustainability through preventing risk occurrence, reducing the adverse impact of disruption, improving flexibility, and changing the complacent attitude of businesses [40]. Therefore, with the potential wider usage of modern digital technologies, the banking sector is continuously exploring novel ways to perform banking operations that can bring cost efficiency and ensure long-term sustainability. In this context, to transform their business models, banks can consider adopting blockchain technology [44] that relies on an intermediary between two parties. Blockchain technology can transform the banking industry and make day-to-day processes more transparent, secure, and efficient. It works as a decentralised ledger that effectively keeps track of transactions between two parties. It has the potential to eliminate intermediaries and bring transparency and traceability to transactions by streamlining, simplifying, and enhancing conventional business processes [60]. BC implementation improves the trade between two or more parties in terms of privacy, tracking, transparency, and enforceability in smart contracts. In the long run, BC technology makes the SC system more energy-efficient, cost-effective, and high-performance-oriented.

The application of BC technology can improve resource efficiency and effective use. This latest technology will help establish a sustainable and secure supply chain in the banking industry. The Harvard Business Review reported that 60% of financial institutions are willing to use blockchain for international money transfers, 23% for securing clearing and settlement, and 20% for KYC regulations and anti-money laundering services [28]. BC is an emerging technology, and researchers and industry practitioners are trying to implement BC in the field of SC. Considering its unique characteristics, such as privacy, security, smart contract and scalability, the need to make banking operations more efficient becomes essential. This technology is in its infancy phase [46], and limited work has been carried out so far. Extant literature emphasises the potential benefits of a blockchain, while a few studies suggest the critical success factors (CSFs) of blockchain technology in the banking sector [21, 28]. Kaur et al. [31] have tried to analyse barriers to the implementation of blockchain technology for supply chain finance. Despite various empirical studies suggesting a positive relationship between blockchain technology and sustainable and resilient operations [40], studies have yet to identify the CSFs of blockchain adoption in the banking industry, and the literature remains largely theoretical.

The prime focus of this research is to identify CSFs for sustainable and resilient operations in the banking industry in an uncertain business environment and to analyse their inter-relationships using stakeholders’ perspectives. Therefore, the research questions of the study are described as follows:

1. What are the CSFs of blockchain technology adoption for sustainable and resilient operations in the banking industry in an uncertain environment?

2. What are inter-relationships among significant CSFs?

3. What is the ranking of the CSFs that need to be considered by a banking organization?

Therefore, the objective of this study is to identify the CSFs of blockchain technology for sustainable and resilient operations, the inter-relationships among CSFs, and their ranking in banking organisations. To achieve this objective, the study aims to come up with an integrated framework that illustrates the process of identifying the CSFs of blockchain technology applications. The first research question was addressed by reviewing the existing literature, having semi-structured interviews with multiple stakeholders, and analysing those stakeholders’ opinions through the Fuzzy Delphi method (FDM). The second and third research questions were addressed by using the Decision-Making Trial and Evaluation Laboratory (DEMATEL) technique. The study seeks to address the research questions through the following contributions: Drawing from stakeholder theory and CSF theory, the study first identifies and evaluates CSFs for blockchain technology adoptions in sustainable and resilient banking operations through a literature review, semi-structured interviews, and FDM. Second, we develop an inter-related structure of CSFs to explain their inherent interdependence and illustrate the complex relationship between them. The study is the first to provide an in-depth analysis of the interactions between CSFs that facilitate the adoption of blockchain technology in the banking industry. By taking stakeholders’ views in the course of FDM and DEMATEL, the study resolves the high complexity of interconnected CSFs. Third, the study prioritises and ranks CSFs in relation to blockchain technology for sustainable and resilient banking operations. By highlighting the significance of specific CSFs and their effect on other CSFs, it further suggests CSFs that need to be considered for future research. It also suggests practitioners focus on CSFs that should be addressed first to deal with other CSFs with greater ease. Ultimately. The study gives insight into how to facilitate the adoption of blockchain technology in the banking sector for sustainable and resilient operations.

The rest of this paper is organised as follows: Sect. 2 presents a literature review of blockchain adoption for sustainable and resilient operations in the banking industry. Section 3 discusses the research methodology used in this study. Section 4 comprises data analysis and the findings of this study. Section 5 presents the research findings and offers implications for stakeholders, followed by conclusions in Sect. 6.

Literature review

This section presents the theoretical lens used in this study and reviews recent studies on blockchain adoption for sustainable and resilient operations in the banking sector.

Theoretical perspectives

The study used stakeholder theory and CSF theory to address the CSFs of blockchain technology applications for a sustainable and resilient banking industry. A review of these theories is as follows:

Stakeholder theory

Stakeholder theory asserts that stakeholders influence the organisational plans. Therefore, the success or failure of any organisational decision is determined by the perception of multiple stakeholders. Freeman [18] defined stakeholders as “any group or individual who can affect or is affected by the achievement of an organisation’s objective” [18], pp.46). Therefore, stakeholder theory suggests that an organisation manages its relationships with key groups like customers, employees, suppliers, communities, and other groups that can determine the success or failure of an organisation’s objective. The involvement of multiple stakeholders brings complexity; therefore, successful, long-term organisational survival is contingent upon effectively managing stakeholders’ relationships. Stakeholder theory has been widely used in different fields. Shankar et al. [55] applied stakeholder theory to measure the interrelationships among CSFs of a traceability system for food logistics. Kannan [30] developed a sustainable supplier selection problem by integrating a multi-stakeholders view. Using a three-phase methodology, the study assesses Indian suppliers by integrating the sustainability view of various stakeholders such as customers, shareholders, government, and employees.

Stakeholder theory has been employed in this study to obtain insightful information and recommendations on CSFs for the successful adoption of blockchain technology in the banking industry. Organisations should think about the potential CSFs of blockchain technology for a sustainable and resilient banking industry since stakeholder theory implies that organisations must take into consideration the interests of all stakeholders when making decisions [13, 35]. Furthermore, the implementation of blockchain technology may be approached more sustainably and morally using stakeholder theory. Banking organisations may minimise, surmount, and assure the successful adoption of the multistakeholder view in their respective organisations by taking this into consideration. To the best of our knowledge, no empirical study has applied stakeholder theory to assess blockchain adoption in banking sectors. Therefore, the study involves interaction with stakeholders such as blockchain developers, banking professionals, and technocrats who helped identify CSFs for blockchain adoption in the banking sector to ensure sustainable and resilient operations.

Critical success factor (CSF) theory

The study is also grounded on the crucial success factors theory. The critical factors theory refers to firms' specific areas that need to be satisfactory to ensure successful competitive advantage [12]. As per [5], CSFs are “those few things that must go well to ensure success for a manager or an organisation, and thus they represent those managerial or enterprise areas that must be given special and continual attention to bring about high performance”. Critical success areas can be used to devise appropriate strategies and identify critical issues linked with any plan [55]. Adopting blockchain technology is a complex task requiring identification, development, and alignment among several areas. The applications of CSF theory enable one to minimise the complexity associated with decision-making and identify areas that require immediate attention. Critical success theory has been widely used in technology adoption areas. For example, Pozzi et al. [49] used the CSF theory to identify the CSFs of Industry 4.0 technologies. Shankar et al. [54] applied CSFs of sustainable banking operations, and Shankar et al. [55] used them to identify the interrelationship of CSFs of traceability in a food logistics system.

Blockchain adoption in the banking sector for sustainable & resilient operations

Banking and financial institutions are modern society's lifeblood and contribute significantly to economic growth [45]. The application of digital technology in banks for fund transfers, registrations, and maintaining back-end utilities can accelerate the development of blockchain technology adoption in organisations [17]. It acts like a distributed ledger and is open to every network member. Each block is secured using cryptocurrency, which helps improve compliance, minimise cost, and increase security. The technology provides data storage with enhanced security, minimising fraud and cyber-attacks. It is an open, distributed ledger that records transactions between two parties. It is managed by a peer-to-peer network that adheres to a protocol for inter-node communication [50]. Some of the unique features of blockchain, namely decentralised ledgers, tamper-proofing of data, fast execution of transactions, minimised cost, and improved capacity, bring security and resilience to the banking industry's supply chain [28].

Blockchain brings resilience by reducing single points of failure [57]. It also helps bring sustainability to operations [53]. The peer-to-peer model increases redundancy and enables fault tolerance [40]. It also overcomes third-party risks as different nodes serve from various locations. Also, transactions in a blockchain system are challenging to modify; therefore, entire blocks of each peer need to be altered. The blockchain system brings the benefits of improved network capacity. It also facilitates automation, streamlines processes, eliminates manual back-office labour, reduces time, and enhances security. It is helpful for financial transactions. For example, in the case of an interfirm transaction where different ERP systems exist, processes can be streamlined using blockchain. It can improve transparency and minimise the need for reconciliation in monetary transactions [58].

Further, it can offer better payment clearing mechanisms, upgraded credit information, and management systems that may lead to a more efficient banking system. In the traditional banking system, remittance cost accounts for 5 to 20%. Using blockchain, the remittance cost can decrease to 2 to 3% [21]. Blockchain offers a novel institutional technology that affects transaction costs in a decentralised system and refines governance. The use cases of the adoption of blockchain technology in various Indian and foreign banks and the benefits availed by several banks are highlighted in Table 1 and 2, respectively.

Table 1.

Adoption of blockchain technology in indian banks

Bank	Area	Description	Benefit Realized
Axis Bank	Trade finance	Axis Bank issued LoC to ArcelorMittal Nippon Steel on Secure Logistics Document Exchange in a completely digital manner	Paperless processing for LoC Issuance, Transparency and visibility for stakeholders involved in the transaction lifecycle process
ICICI Bank	Trade finance	Documents pertaining to Trade finance process, such as Letter of Credit are shared with the parties over the Blockchain across ICICI bank’s 250 + clients	Paperless processing for LoC Issuance
IndusInd Bank	Cross border payment	IndusInd bank has tied up with ripple for leveraging the latter's platform for cross border remittances and share financial data with other financial institutions	Faster P2P payments and secure transmission of financial information
Kotak Mahindra Bank	Trade finance	Kotak Mahindra bank has reduced the time taken for issuance of Letter of Credit to JP Morgan from 20–30 days to few hours by adopting Blockchain technology	Letter of Credit issuance turnaround time reduction, Paperless processing for LoC Issuance
State Bank of India	Cross border payment	SBI has partnered with JP Morgan to use the latter's Liink blockchain platform for cross border customer payments and sharing financial data with other banks using the platform	Faster P2P payments and secure transmission of financial information
Yes Bank	Operational business processes	Yes Bank leveraged IBM Hyperledger to reduce invoice reconciliation, auditing and recording from 4 days to real-time	PTP business process improvement, Auditing and reporting performance improvement

India's Axis Bank completes first blockchain-powered domestic trade transaction—Aliens: AI Crypto News & Markets Updates. ICICI Bank | ICICI Bank on-boards over 250 corporates on its blockchain platform for trade finance. IndusInd Bank and Ripple tie up for Cross Border Remittances -. Kotak Mahindra Bank re-defines trade finance with blockchain (efma.com). State Bank Of India Ties Up With JP Morgan For Use Of Blockchain (inc42.com). Yes Bank upbeat on using blockchain, to add more processes—The Economic Times (indiatimes.com)

Table 2.

Adoption of blockchain technology by foreign banks

Bank	Affected area	Description	Benefit realized
Bank of Tokyo Mitsubishi UFJ (Japan) (BTMU)	Operational business processes	IoT sensor data from equipment is transmitted across Blockchain to improve service level agreement between two parties IBM and BTMU, for accountability & automation of invoicing & payment process	Automation of invoicing and payment about the order to cash & procure to pay business process
Standard Chartered (UK)	Trade finance	Standard Chartered Bank issued Letter of Credit to Viyalletex (Garments industry) over Blockchain leveraging the CONTOUR blockchain network	Letter of credit issuance turnaround time reduction, Paperless processing for LoC issuance, settlement risk reduction
HSBC (UK)	Trade finance	HSBC is using R3'S Corda platform to store and share digital assets and tokenize paper-based certificates such as LoC safely	Paperless processing for LoC Issuance
Commerzbank (Germany)	Settlement & clearance	Commerzbank and Munich Re traded a commercial paper on R3 Corda Blockchain paper instantly, a process which takes upto 4 days owing to the time taken by traditional clearance systems	Settlement & Clearing turnaround time reduction through a reduction in the no. of intermediaries
Bank Santander (Spain)	Settlement & clearance	Bank Santander issued and redeemed a 20 million US Dollar bond on the Ethereum blockchain network insta.only	Settlement & Clearing turnaround time reduction through the reduction in the no. of intermediaries
Wells Fargo (USA)	Settlement & clearance	Wells Fargo is using the Core-FX distributed ledger to settle Forex trades between itself and HSBC, reducing FX trade settlement time from 2 days to three minutes	Settlement & Clearing turnaround time reduction through the reduction in the no. of intermediaries, Risk exposure to third party reduction

Reduced turnaround time and automation brought sustainability to banking operations and made the entire operations resilient (Table A1).

IBM and Bank of Tokyo-Mitsubishi UFJ Develop Blockchain-Powered Contract Management System | Nasdaq. Standard Chartered Bank Pioneers Bangladesh’s First-Ever Blockchain Letter of Credit – Treasury Management International (treasury-management.com). HSBC Becomes First Financial Institution To Move Corda Enterprise Blockchain Technology On To Google Cloud—Global Cloud Platforms. CommerzBank, KFW, and MEAG Simulate Security Transaction via Blockchain—R3. Santander launches the first end-to-end blockchain bond. HSBC and Wells Fargo use blockchain to settle forex trades | Reuters

CSFs in implementing blockchain

CSFs play an integral part in achieving organisational goals. Despite the broader use of blockchain in secure and sustainable banking operations, there are very few studies on identifying and modelling CSFs necessary to adopt blockchain technology in safe and sustainable banking operations in uncertain environments. Drawing from the literature survey, a total of 17 CSFs were identified that help in the successful adoption of blockchain technology for secure and sustainable banking operations.

Cf.1. Management leadership buy-in

Management/leadership buy-in is defined as the ability of the management or the leadership to support blockchain use case adoption and the implementation of this technology [63]. It refers to the organizational capability pertaining to getting the right people to sponsor use cases of blockchain technology and a program to manage the implementation of blockchain technology. As per Holotiuk and Moormann [24], the top management considers blockchain potential positively. It also mentions that awareness needs to be created among the people by the management to highlight its advantages. Organizational capability pertaining to getting the right people to sponsor use cases of blockchain technology and a program to manage the implementation of blockchain technology is essential.

Cf.2. Transaction cost efficiency

Transaction cost-efficiency refers to the cost of sending money either in a P2P fashion or a cross-border country route. For a blockchain network to be adopted, it should be able to lower transaction costs. Owing to its decentralized structure, blockchain has the ability to reduce transaction costs by removing the intermediary. However, it has been noted by [9] that all blockchain implementations do not reduce costs. An article in CNBC) mentioned that transaction fees for mining bitcoins had increased exponentially. One person stated that for a transfer of 25 USD worth of bitcoin, the transaction fees were 16 USD.1

Cf.3. Transaction storage/energy efficiency

Transaction storage signifies the storage costs involved in transactions owing to mining & verification operations across the platform. Blockchains consume energy, costing up to 15 million US dollars per day, to compute and verify transaction nodes [9]. According to [4] from a processing perspective, each transaction per second requires 6.75 GB, and the recommendation is to have a shared storage network.

Cf.4. Scalability

Scalability refers to the ability to scale to the tune of large transaction volumes, ensure platform throughput, and achieve low latency levels transactionally. As the size of the blockchain network grows, there is a need for the platform to be scalable. Scalability is also required to reduce latency in transaction processing because each transaction currently takes at least 10 min to be confirmed. Prasad et al. [50] mention that scalability is one of the top factors required for the adoption of blockchain technology by organizations. The scalability trilemma of the blockchain resonates across various literature, which states that only two of the three characteristics, namely decentralisation, security and scalability, can be achieved in blockchain technology [10].

Cf.5. Security and integrity

A blockchain network is susceptible to various attacks such as Double Spending attacks (a miner who controls computational power can replicate a transaction which leads to a digital token being spent two times), 51% Attacks (several pools of miners that can reverse or update a valid transaction or confirm an invalid one), and private key security attacks or a private key security [58]. Ensuring an impenetrable platform results in the lack of data breaches, leading to the maintenance of the integrity of the network across which transactions are happening.

Cf.6. User data privacy

Blockchains operating over public ledgers can lead to possible intrusions into the privacy of users. The spending habits of the users can be tracked on a public blockchain ledger. Ćirić et al. [9] also mention that blockchain suffers from a lack of anonymity. Personal data can be compromised and therefore cause material/immaterial damages to the person involved. EU GDPR was initiated to protect against misuse of personnel data, and companies are fined to the tune of millions in case of a breach.

Cf.7. User engagement & desirability

User engagement and desirability refer to the ability to switch from traditional modes of payment to a blockchain system. For blockchain technology to be successful, users need to be engaged with issues pertaining to awareness of this technology. Many users confuse blockchain technology with Bitcoins, indicating a lack of awareness. As per Ćirić et al. [9], users should be comfortable storing their personal data on a public ledger.

Cf.8. Ease of local & international legislation and regulation

With increasing local & international regulation and some countries banning cryptocurrency, adoption of blockchain technologies is hindered. The local and international laws pertaining to blockchain are currently in a state of flux. Some states have banned applications of blockchain, such as cryptocurrency; China is a case in point. Chinese banks are also adherent to national law, with the People’s Bank of China outlawing cryptocurrency transactions [51]. On the other hand, other countries like India are against cryptocurrencies. The RBI is a staunch opponent of the instrument but supports exploring various use cases pertaining to blockchain technology.2

Cf.9. Personnel training

Personnel training refers to the ability to upskill the blockchain ecosystem's personnel. As per [9], investment in upskilling blockchain technology resources is required to increase the adoption rate. Currently, industry-wide skills related to blockchain are not pervasive.

Cf.10. Availability of funds for implementation

Capital investment by management is essential for the implementation and post-implementation management of services for a blockchain platform. Sufficient capital is one of the CSFs for the implementation and adoption of blockchain technology [63]. The cost of implementing blockchain can range from 5000 US Dollars to 200,000 US Dollars.3 Depending on use cases & smart contract functionality, as well as the need for external entities to support blockchain technology implementation, the costs can go beyond this.

Cf.11. Professional consultation & advisory capability

Professional consultation & advisory capability signifies getting help from external sources for blockchain use case definition, proof of concept, and implementation. Zhou et al. [63] state that companies can approach professional consulting organizations for assistance in blockchain implementation. This factor is inferred as the ability of a professional services firm/technology development organization to evaluate use cases pertaining to blockchain technology implementation in a sector and implement or manage such projects. Table 3 outlines cases of professional services firms and the companies that have required assistance from the former.

Table 3.

Professional services/technology firms engaged blockchain related engagements

Bank	Firm	Nature of work
Bank of Tokyo Mitsubishi UFJ	IBM	Smart contract proof of concept & prototyping
ICICI Bank	Infosys	Blockchain deployment assistance
Kotak Mahindra Bank	Deloitte	Blockchain PoC development
Yes Bank	IBM	Blockchain technology implementation
BNP Paribas	Ernst & Young	Blockchain pilot program

BNP Paribas and EY explore private blockchain to optimize the bank’s global internal treasury operations—BNP Paribas United Kingdom

Cf.12. Blockchain talent availability

Blockchain talent availability means an organisation’s ability to tap into a pool of people capable of understanding the nuances of blockchain, cryptocurrency, and digital tokens. Organizations looking to implement and manage blockchain technology need personnel skilled in the area of technology as well as legal and financial experts, as per Ćirić et al. [9]. India currently has an insufficiency in the total software workforce associated with blockchain technology, highlighting a shortage in the talent pool [29].

Cf.13. Integration with other cloud services/e-commerce platforms

A key success factor for blockchain technology would be integrating it with cloud systems, in-house ERP systems, and e-commerce platforms. It can improve various business processes such as PTP, OTC, etc. Organizations adopt cloud computing to increase flexibility and speed and reduce operational costs [11]. HSBC Bank moved its CORDA R3 blockchain enterprise platform to the cloud to reduce operating costs [25]. As per Nguyen et al. [46], a cloud can provide scalable support to the blockchain as the number of transactions increases because cloud computing can offer on-demand computing resources for data processing services.

Cf.14. Incentives for miners

Blockchain is based on decentralization, and a consensus has to be reached whenever a new update is made on this network. Miners are validators that provide computational resources to verify such updates and achieve consensus [7]. Miners are rewarded with cryptocurrency tokens for validation purposes. For large-scale adoption, it is inferred that an incentivization strategy is required so that miners who can provide computational resources and are incentivized to continue doing the same. Organisations need to develop strategies to incentivize miners to verify a transaction by rewarding them.

Cf.15. Smart contract robustness and business case deployment

Smart contract robustness and business case development refer to using programmable code in the form of smart contracts to achieve mutual agreement between two or more parties triggered by an event. As per IBM, smart contracts are defined by a set of pre-defined contracts deployed on the blockchain network when certain conditions are met with the use case of releasing funds applicable to the banking industry [26]. Smart contracts are beneficial because they can achieve significant operational benefits in automation and eliminate intermediaries.

Cf.16. Full ecosystem interoperability

Blockchain technology adoption encounters various risks of errors and failures because blockchain technology lacks interoperability and universal standards. The lack of a universal standard creates problems in communication because different organisations use different blockchain platforms in communication, information sharing, and data protection. Interoperability refers to the possibility of freely sharing value across all blockchain networks without the requirement of intermediaries. It also helps interact with users from other blockchain networks and reduces spending resources and efforts on the translation or wasting time on downtime. Information can be sent or received from members when required.

Cf.17. Technological maturity and standardisation

Technical investment and maturity are essential for blockchain technology adoption. Several technical issues in the developer community range from blockchain size to latency to security and throughput. Organisations also do not have a consensus on using blockchain technology; few organisations favour using a new blockchain, whereas others would like to improve the Bitcoin blockchain. The technical community is not consistent in their choice of using blockchain technology. Also, blockchain technology is immature. Therefore, technology investment and maturity in all organisations are critical for blockchain technology adoption. Table 4 presents the CSFs for blockchain technology adoption in banking.

Table 4.

CSFs for blockchain implementation

CSF	CSF Description	References
Cf.1	Management/leadership buy in	Ćirić et al. [9], Zhou et al. [63], Prasad et al. [50], Holotiuk and Moormann [24], Madhusudhanan and Reka [38]
Cf.2	Transaction cost efficiency	Ćirić et al. [9], Prasad et al. [50], Madhusudhanan and Pon Reka, n.d.), Ali et al. [3]
Cf.3	Transaction storage/energy efficiency	Ćirić et al. [9], Prasad et al. [50], Madhusudhanan and Reka [38], Sahoo et al. [53]
Cf.4	Scalability	Ćirić et al. [9], Ali et al. [3]
Cf.5	Security & integrity	Ćirić et al. [9], Prasad et al.[50, 60], Madhusudhanan and Reka [38], Ali et al. [3]
Cf.6	User data privacy	Ćirić et al. [9], Prasad et al.[50], Madhusudhanan and Reka [38], Ali et al. [3]
Cf.7	User engagement & desirability	Ćirić et al. [9, 60], Madhusudhanan and Reka [38]
Cf.8	Ease of local & international legislation & regulation	Ćirić et al. [9], Zhou et al. [63], Prasad et al. [50], Madhusudhanan and Reka[38]
Cf.9	Personnel training	Ćirić et al. [9], Zhou et al.[63]
Cf.10	Availability of funds for implementation	Zhou et al. [63]
Cf.11	Professional consultation & advisory capability	Zhou et al. [63]
Cf.12	Blockchain talent availability	Prasad et al.[50], Ali et al. [3]
Cf.13	Integration with other cloud services/ E-commerce platforms	Prasad et al. [50], Madhusudhanan and Reka [38], Holotiuk and Moormann [24], Mulhall [45], Nguyen et al.[46]
Cf.14	Incentives for miners	Prasad et al.[50], Madhusudhanan and Reka [38], Chiu and Koeppl [7], Chiu and Koeppl [8]
Cf.15	Smart contract robustness & business case deployability	Prasad et al.[50], Madhusudhanan and Reka [38]
Cf.16	Interoperability and standardization	14, 15, Ali et al. [3]
Cf.17	Technology investment and maturity	14, 15

Research methodology

The study was carried out in three phases. A research framework was developed to analyse the complex cause-and-effect relationships among CSFs affecting blockchain adoption for sustainable and resilient banking operations (Fig. 1). Each of these phases is described in this section.

Fig. 1.

Research framework

Integrated approach

Phase I- identification of CSFs

An initial set of CSFs for blockchain adoption in the banking sector has been identified from the literature and then these CSFs have been verified through the semi-structured interviews with 10 experts. Section 2.3 has already presented and discussed these factors.

Phase II- finalisation of CSFs using the FDM

Initially proposed by Ishikawa et al. [27], the FDM combined fuzzy set theory with the conventional Delphi method. The conventional Delphi technique is a lengthy, time-consuming, and consensus-making process. FDM overcomes the limitations of conventional Delphi techniques as experts’ opinions are collected in fuzzy numbers, and experts are not needed to review their judgment [42]. Recently, FDM has been widely used in the literature (Table 5). In this study, FDM has been applied to assess the CSFs identified through the literature review and experts’ interviews. FDM involves three significant steps: data collection, conversion of decision-makers opinions into Triangular fuzzy numbers (TFNs) using Table A2, and calculation of the fuzzified importance weight followed by defuzzification of fuzzy scores. Once fuzzy scores are calculated, a threshold value for selecting CSFs is determined, and the chosen CSFs are used for the next phase of the study.

Table 5.

Application of FDM

Author(s)	Application area	Methodology
Rejeb et al.[52]	Industry 4.0 adoption in the circular economy	Multiple industry focus, sample size 12, TFNs
Agarwal and Singh [2]	Sustainable sub-indicator of textile wastewater treatment	The textile industry, sample size 5, TFNs,
Padilla-Rivera et al., [48]	Social indicator of circular economy	Multiple industry focus, sample size 45, TFNs
Abdul-Hamid et al.[1]	Industry 4.0 implementation in the circular economy	Palm& oil industry, sample size 14, TFNs

Table A2.

Decision makers involved in FDM method

Particular	Number	Particular	Number
Domain		Locations
Retailer & commercial banking	8	Delhi, India	4
Blockchain developer	4	Mumbai, India	6
Lawyer & legal consultant	3	Pune, India	3
Blockchain researcher	4	Bangalore, India	5
		Hyderabad, India	2
Total years of experience in blockchain domain
3 – 5 years	6
More than 5 Years	14

Steps involved in FDM

Step 1: Identification of CSFs for blockchain adoption in the banking sector.

In the first phase of the study, a literature review was carried out to identify CSFs for blockchain technology adoption for secure and sustainable banking operations. Then, semi-structured interviews were conducted to confirm these CSFs with industry practitioners (5 banking professionals, 2 blockchain developers, 3 blockchain researcher & legal consultant) and identify other essential CSFs not documented in the literature.

Step 2: Gathering experts’ opinion

After finalising the CSFs, in the second phase, practitioners were approached to participate in the study. The study goal was to identify and finalise CSFs for blockchain technology adoption in the banking industry based on multiple stakeholders’ views. Therefore, a wide range of respondents from different domains, namely retailers and commercial banks, blockchain developers, lawyers and legal consultants, and blockchain researchers, were chosen for this study to get a multistakeholder perspective. The profile of experts involved in the FDM process is given in Appendix Table A3. Decision-makers assessed the significance of CSFs using linguistics variables listed in Table A2. TFNs were utilised for assessing the CSFs. In addition, a geometric mean model was used to aggregate the group decision of experts.

Table A3.

Linguistic rating and TFNs

Linguistic scale	TFNs
Very unimportant (VU)	$\left(113\right)$
Unimportant (U)	$\left(135\right)$
Moderate (M)	$\left(357\right)$
Important (I)	$\left(579\right)$
Very Important (VI)	$\left(799\right)$

Source: Zhang [61]

Step 3: Calculation of CSFs for sustainable and resilience banking operations.

The CSFs for sustainable and resilient banking operations were finalised by the aggregating group weight (${\tilde{n}}_k$) of each CSF and then aggregate weights. The threshold value of $\tilde{n}$ was compared.

$\tilde{n}$ signifies the average aggregate group importance weight of all CSFs (${\tilde{n}}_k$).

The defuzzification of ${\tilde{n}}_k$ and $\tilde{n}$ is done for calculating the crisp value using the formula mentioned below.

$$\text{If}{\tilde{n}}_k=\left(lmu\right)\text{then Crisp value}\left({\tilde{n}}_k\right)=\frac{l+4m+u}{6}$$ 1

$$\text{If}\tilde{n}=\left(abc\right)\text{then Crisp value}\left(\tilde{n}\right)=\frac{a+4b+c}{6}$$ 2

The criteria to assess a CSF for a sustainable and resilient banking operation is calculated as follows:

$$k^{\mathit{th}}\text{CSF is used for the study}\text{if}{\tilde{n}}_k\ge \tilde{n}$$ 3

$$k^{\mathit{th}}\text{CSF is not used for the study},\text{if}{\tilde{n}}_k<\tilde{n}$$ 4

Phase III interrelationship among CSFs for sustainable and resilient banking operations using DEMATEL

In the third phase, the interrelationships among CSFs for sustainable and resilient banking operations were calculated using DEMATEL techniques. DEMATEL, developed by Battelle Memorial Institute of Geneva, is used to compute the interrelationships among CSFs to form and analyse structural models of cause-effect relationships [16]. DEMATEL has been widely used in various studies in different functional areas (Table 6).

Table 6.

Application of DEMATEL technique

Author(s)	Application area	Methodology
Liang et al. [36]	Determinants of economic operations of electric vehicle charging stations	Electric vehicle charging stations, Sample size
Braga et al. [6]	Determinant of smart city	Smart city, sample size 2 expert groups
Mishra [41]	Factors influencing omnichannel retailing adoption	The apparel industry, sample size 12
Garg [20]	E-waste mitigation strategies	Electrical and electronic industry, sample size not applicable
Zhao et al. [62]	Critical factors for the sustainable energy sector	Energy sector, sample size 15

A large-scale study was difficult to carry out because blockchain adoption is in its infancy and novel in its application. Therefore, five presential group sessions were conducted using the purposive sampling technique to get the multistakeholder view. Each group comprised a homogenous set of participants, with five members in three groups and four and three members in the other two groups. Within each group, a discussion took place on the success factors of BCT adoption in banking, and the responses were collected through the consensus method. Each group was homogenous and comprised respondents with experience in a similar domain–payment and credit card, retail banking, commercial banking, blockchain developer, and lawyer & legal consultant. A blockchain researcher & cryptocurrency start-up co-founder were approached. The profiles of the decision-makers in each group are given in Table A4.

Table A4.

Decision maker profiles

Decision maker	No. of experts	Domain	Description
Group 1	5	Payments & credit cards	Experience with one of the world’s Top 3 payment service providers focused on big data analytics, credit analytics and user acquisition
Group 2	3	Retail banking	Experienced in retail banking software development with a bank technology firm
Group 3	5	Commercial banking	Experience in commercial banking software development with a Top UK based and a global private bank
Group 4	5	Blockchain developer	Experience in developing proof of concepts through blockchain for banking clients
Group 5	4	Lawyer & legal consultant, blockchain researcher & cryptocurrency start-up co-founder	Experience in advising clients pertaining to digital tokens, cryptocurrency, Nfts As well as research on blockchain technology from a legal perspective

Step 1: Determine average matrix

To collect responses, p experts (stakeholders) from diverse organisations were collected and rated on a scale of 0 to 4 (0 = no influence, 1 = low influence, 2 = medium influence, 3 = high influence, 4 = very high influence) to indicate the degree of direct influence each factor/element $i$ exerts on each factor/element $j,$ which is represented by $x_{\mathit{ij}}$.

The score given to each expert gave an $n\times n$ non-negative answer matrix $X^k$ with $1\le k\le p.$ Hence, $X^1,X^2,\dots .X^p$ are resultant matrices derived from p experts, and each element of $X^k$ is an integer represented as $x_{\mathit{ij}}^k$. The diagonal elements of each resulting matrix $X^k$ are equal to 0. The average matrix A is given as follows:

$$a_{\mathit{ij}}=\frac{1}{p}\sum _{k=1}^p{x}_{\mathit{ij}}^k$$ 5

Step 2: Compute normalised initial direct matrix

The normalised direct matrix D is calculated by normalising the average matrix A as follows:

$$D=sA$$ 6

where s is a constant and represented as follows.

$$s=\frac{1}{ma{x}_{i\le i\le n}{\sum }_{j=1}^1{a}_{\mathit{ij}}}$$ 7

Step 3: Calculate the total relationship matrix T

The total relationship matrix is calculated as follows:

$$\text{T}=\text{D}{\left(\text{I}-\text{D}\right)}^{-1},\text{where I is an}n\times n\text{identify matrix}.$$ 8

The sum of rows and columns of matrix T is represented by vectors D and R, respectively.

Step 4: Compute degree of influence

Let $D_i$ be the sum of the $ith$ row in matrix T. Thus, $D_i$ signifies the total given by both the direct and indirect effects that CSF $i$ has on other CSFs. $R_j$ is the sum of the $jth$ column in matrix T which represents the total received from both the direct and indirect effects that all other CSFs have on CSF $j;$when $i=j$, the sum $\left(D_i+R_j\right)$ signifies the total effects given and received by CSF $i$; thus it signifies the degree of importance of CSF $i$. The difference $\left(D_i-R_j\right)$ signifies the net effect that factor $i$ contributes to the system. If $\left(D_i-R_j\right)$ is + ve, then the influence factor $i$ is a net cause, while if it is –ve, $i$ is a net receiver.

Step 5: Draw impact relation map

An impact relation map is drawn by mapping all values of $\left({\text{D}}_{\text{i}}+{\text{R}}_{\text{j}}\right)$ and $\left({\text{D}}_{\text{i}}-{\text{R}}_{\text{j}}\right)$ on coordinate sets to visualise the complex interrelationships and to understand the significance of CSFs in decision-making.

Rationale of using integrated FDM and DEMATEL technique

The study used the FDM to finalise the CSFs for sustainable and resilient banking operations to overcome the vagueness and uncertainty associated with decision-making. Then the DEMATEL technique was used to determine the cause-and-effect relationships among CSFs [31, 59]. The DEMATEL technique has a significant advantage over other multicriteria decision-making methods such as Elimination and Choice Expressing Reality (ELECTRE), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR), and grey relational analysis (GRA). DEMATEL assists in ranking the options, identifying significant assessment criteria, and calculating their weights. These MCDM methods are used to prioritise the CSFs but fail to determine the cause and effect factors [56]. For example, AHP does not involve the computation of dependencies and interrelationships among criteria. Furthermore, the analytical network process (ANP) deals with the dependencies. However, it considers identical weights for each segment to compute a super-weighted matrix, which is impractical [19].

Moreover, it facilitates decision-makers in understanding and visualising the mutual relationships among criteria. It assists in analysing both the direct and indirect mutual relationships between criteria and helps in developing a cause-effect relationship in decision-making, thereby helping decision-makers in comprehensive decision-making. Integrated FDM and DEMATEL techniques have been applied in diversified fields of management. Table 7 presents applications of integrated FDM and DEMATEL approaches.

Table 7.

Applications of integrated FDM and DEMATEL techniques

Studies	Focus areas
Mohandes et al. [43]	Casual factors contributing to construction accidents in a developing country
Hashemi et al. [23]	Analysis of fears of failure in the international entrepreneurship ecosystem
Khan et al. [32]	Analysis of elements of Halal supply chain management and their significant risk dimensions
Singh and Sarkar [56]	Sustainable product development in the Indian automobile industry
Gardas et al. [19]	Challenges to sustainability in the textile and apparel
Kumar et al. [33]	Significance of social media in polio prevention
Mangla et al. [39]	Benchmarking critical factors related to logistics management

Results

The results of this study have been presented in two phases. In the first phase, CSFs of blockchain technology adoption for a sustainable and resilient banking industry were finalised using semi-structured interviews and the FDM. To get a multi-stakeholder view, responses from different stakeholders were gathered and analysed. RQ1 was addressed in the first phase of the study. It provides CSFs for blockchain technology adoption for sustainable and resilient banking operations. In the second phase, RQ2 and RQ3 were addressed by identifying the interrelationships among CSFs and finalising the ranking of these CSFs using the DEMATEL technique.

Finalization of CSFs for blockchain technology adoption using interviews & FDM

The semi-structured interviews were used to explore and confirm CSFs that facilitate blockchain technology adoption for sustainable and resilient banking operations from multi stakeholder perspective. Experts confirmed all the CSFs identified from a thorough literature review. Experts indicated that all identified CSFs are comprehensive and include all the essential factors for blockchain technology adoption. The coverage of two CSFs ‘personnel training’ and ‘blockchain talent availability’ were overlapping; however, they were retained for FDM analysis.

After identifying CSFs from the literature and interviews with practitioners, CSFs were finalised for detailed analysis using FDM. The multi-stakeholders view was important because it offered a holistic view of adopting blockchain technology. Only a few experts were involved in the study, but Ocampo et al. [47] found no association between the quality of decisions and the number of experts in their research work. The CSFs for adopting blockchain technology were evaluated using the scale in Table A3, and geometric means were used to aggregate the experts’ judgements [37]. The threshold value was computed by averaging the importance weight of all CSFs for the finalisation of CSFs. Two CSFs, namely user engagement & desirability (6.46) and personnel training (6.28), were dropped because the crisp value was found to be less than a threshold value (7.78). Therefore, 15 of 17 CSFs were finalised for the next phase of the study because their value was above the threshold value (Table A5).

Table A5.

FDM Results

Notation	Factors	Value
Cf1	Management/leadership buy in	7.86
Cf2	Transaction cost efficiency	7.78
Cf3	Transaction storage/energy Efficiency	7.86
Cf4	Scalability	7.81
Cf5	Security & integrity	7.84
Cf6	User data privacy	8.03
Cf7	User engagement & desirability	6.46*
Cf8	Ease of local & international legislation & regulation	7.89
Cf9	Personnel training	6.28*
Cf10	Availability of funds for implementation	7.95
Cf11	Professional consultation & advisory capability	8.03
Cf12	Blockchain talent availability	7.84
Cf13	Integration with other cloud services/E-commerce platforms	7.95
Cf14	Incentives for miners	8.21
Cf15	Smart contract robustness & business case deployability	7.78
Cf16	Interoperability and standardization	8.30
Cf17	Technology investment and maturity	8.48
	Threshold value	7.78

*Value less than threshold value was dropped for further analysis

Inter-relationships among CSFs for blockchain technology adoption using DEMATEL

The interrelationships among CSFs for blockchain technology adoption were established based on their influence on one another. Steps 5 to 8, delineated in Subsect. 3.1, of DEMATEL were used to develop a direct influence matrix, a normalised direct influence matrix, a total relation matrix, and the degree of influence. The resulting matrices are presented in Tables 8, 9, 10 and 11, respectively. The values of D + R and D-R are presented in the impact relation map in Fig. 2.

Table 8.

Average direct relationship matrix

CSF	C1	C2	C3	C4	C5	C6	C8	C10	C11	C12	C13	C14	C15	C16	C17
C1	0	2	3	2	1	1	1	2	3	3	3	4	1	3	4
C2	3	0	3	3	2	2	1.4	3	3.8	3.6	4	3.8	3.6	2	4
C3	2	2	0	2	2	2	1.4	3	3.8	3.6	4	3.8	3.6	2	3
C4	4	3	3	0	2	2	1.2	4	3.4	3.6	4	3.8	3	3	3
C5	4	3	3	3	0	3	1.6	4	3.8	3.6	4	3.8	3.6	4	4
C6	3.2	3.2	3.2	3.2	3	0	1.6	4	4	3.8	4	4	3.6	4	4
C8	3.6	3.2	3.2	3.2	3.2	3.4	0	3.2	4	3.8	4	4	4	3.6	4
C10	4	3	3	2	2	2	0.8	0	4	4	4	1	3	3	3
C11	3	1.4	1.4	1.4	1	1	0.6	0	0	3.2	3	3	2.4	1	3.2
C12	2.2	1.4	1.4	2	1.2	1.2	0.6	0	2	0	1	2	1	1.2	2
C13	2.4	1.2	1.2	2	2	1	0.8	0	3	3	0	3	4	1	3
C14	1.8	1.6	0.8	2	1	1	0.8	0	3	3	3	0	2	2	2
C15	3	1.2	3	3	1	1	0.8	0	3	3	3	3	0	1.4	3
C16	3	2	2	1.6	2	1	0.4	4	2	4	4	4	3.4	0	2
C17	2	1	2	1	1	1	0	0	1	2	2	1	1	1	0

Table 9.

Normalised direct relationship matrix

CSF	C1	C2	C3	C4	C5	C6	C8	C10	C11	C12	C13	C14	C15	C16	C17
C1	0	0.04	0.06	0.04	0.02	0.02	0.02	0.04	0.06	0.06	0.06	0.079	0.02	0.06	0.079
C2	0.06	0	0.06	0.06	0.04	0.04	0.03	0.06	0.075	0.071	0.079	0.075	0.071	0.04	0.079
C3	0.04	0.04	0	0.04	0.04	0.04	0.03	0.06	0.075	0.071	0.079	0.075	0.071	0.04	0.06
C4	0.079	0.06	0.06	0	0.04	0.04	0.02	0.079	0.067	0.071	0.079	0.075	0.06	0.06	0.06
C5	0.079	0.06	0.06	0.06	0	0.06	0.03	0.079	0.075	0.071	0.079	0.075	0.071	0.079	0.079
C6	0.063	0.063	0.063	0.06	0.06	0	0.03	0.079	0.079	0.075	0.079	0.079	0.071	0.079	0.079
C8	0.071	0.063	0.063	0.06	0.063	0.067	0	0.063	0.079	0.075	0.079	0.079	0.079	0.071	0.079
C10	0.079	0.06	0.06	0.04	0.04	0.04	0.02	0	0.079	0.079	0.079	0.02	0.06	0.06	0.06
C11	0.06	0.028	0.028	0.03	0.02	0.02	0.01	0	0	0.063	0.06	0.06	0.048	0.02	0.063
C12	0.044	0.028	0.028	0.04	0.024	0.024	0.01	0	0.04	0	0.02	0.04	0.02	0.024	0.04
C13	0.048	0.024	0.024	0.04	0.04	0.02	0.02	0	0.06	0.06	0	0.06	0.079	0.02	0.06
C14	0.036	0.032	0.016	0.04	0.02	0.02	0.02	0	0.06	0.06	0.06	0	0.04	0.04	0.04
C15	0.06	0.024	0.06	0.06	0.02	0.02	0.02	0	0.06	0.06	0.06	0.06	0	0.028	0.06
C16	0.06	0.04	0.04	0.03	0.04	0.02	0.01	0.079	0.04	0.079	0.079	0.079	0.067	0	0.04
C17	0.04	0.02	0.04	0.02	0.02	0.02	0	0	0.02	0.04	0.04	0.02	0.02	0.02	0

Table 10.

Total influence matrix

CSF	C1	C2	C3	C4	C5	C6	C8	C10	C11	C12	C13	C14	C15	C16	C17
C1	0.09	0.1	0.127	0.107	0.073	0.067	0.05	0.089	0.152	0.162	0.159	0.173	0.105	0.123	0.172
C2	0.175	0.081	0.15	0.147	0.107	0.101	0.07	0.123	0.196	0.204	0.207	0.198	0.178	0.126	0.203
C3	0.146	0.112	0.085	0.121	0.1	0.095	0.06	0.115	0.185	0.191	0.195	0.186	0.168	0.118	0.172
C4	0.197	0.141	0.154	0.094	0.109	0.103	0.06	0.145	0.193	0.209	0.212	0.203	0.171	0.148	0.189
C5	0.213	0.152	0.167	0.162	0.081	0.129	0.08	0.155	0.217	0.227	0.23	0.22	0.197	0.179	0.224
C6	0.2	0.157	0.171	0.167	0.138	0.074	0.08	0.156	0.221	0.232	0.232	0.224	0.198	0.179	0.225
C8	0.213	0.161	0.176	0.172	0.145	0.141	0.05	0.145	0.228	0.239	0.238	0.232	0.211	0.177	0.232
C10	0.185	0.132	0.145	0.123	0.102	0.096	0.05	0.063	0.191	0.202	0.198	0.14	0.159	0.138	0.177
C11	0.124	0.073	0.081	0.08	0.059	0.055	0.03	0.037	0.072	0.139	0.133	0.132	0.108	0.07	0.135
C12	0.098	0.066	0.072	0.081	0.056	0.053	0.03	0.034	0.097	0.065	0.083	0.1	0.072	0.066	0.099
C13	0.121	0.075	0.084	0.097	0.081	0.059	0.04	0.042	0.136	0.144	0.084	0.139	0.144	0.076	0.139
C14	0.102	0.076	0.069	0.09	0.058	0.054	0.04	0.038	0.126	0.134	0.131	0.074	0.101	0.087	0.111
C15	0.137	0.079	0.12	0.118	0.066	0.062	0.04	0.046	0.142	0.15	0.147	0.146	0.075	0.087	0.145
C16	0.156	0.106	0.117	0.107	0.095	0.072	0.04	0.129	0.144	0.189	0.185	0.18	0.156	0.074	0.145
C17	0.084	0.051	0.075	0.055	0.046	0.044	0.02	0.028	0.068	0.091	0.09	0.071	0.063	0.054	0.05

Table 11.

DEMATEL ranking and cause effect categorization

CSF	D	R	D + R	D—R	Cause/effect	Rank
C1	1.748	2.24	3.988	−0.492	Effect	2
C2	2.26	1.562	3.822	0.698	Cause	7
C3	2.05	1.793	3.843	0.257	Cause	6
C4	2.328	1.722	4.05	0.607	Cause	1
C5	2.628	1.315	3.943	1.314	Cause	4
C6	2.648	1.205	3.853	1.444	Cause	5
C8	2.759	0.722	3.482	2.037	Cause	13
C10	2.101	1.346	3.447	0.756	Effect	14
C11	1.334	2.368	3.701	−1.034	Effect	9
C12	1.072	2.579	3.651	−1.507	Effect	11
C13	1.46	2.525	3.985	−1.065	Effect	3
C14	1.286	2.418	3.704	−1.132	Effect	8
C15	1.562	2.105	3.667	−0.542	Effect	10
C16	1.893	1.7	3.594	0.193	Cause	12
C17	0.885	2.417	3.302	−1.532	Effect	15

Fig. 2.

Impact relation map

Sensitivity analysis

Sensitivity testing was performed as per the guidelines given by Xia et al. (2015), where the decision-makers weights were altered. The sensitivity analysis methodology for cases 1, 2, 3, 4 & 5 is similar to that followed by Kumar et al. [34], where the weights of the decision-makers were reduced proportionately, as well as kept unequal, with one decision-maker always having more weight than the others in the domain of the case (Table A6). The sensitivity analysis results are given in Table 12 and Fig. 3.

Table A6.

Sensitivity analysis weights as per cases

Case	Group1	Group 2	Group 3	Group 4	Group 5
1	0.3	0.25	0.2	0.15	0.1
2	0.25	0.2	0.15	0.1	0.3
3	0.2	0.15	0.1	0.3	0.25
4	0.15	0.1	0.3	0.25	0.2
5	0.1	0.3	0.25	0.2	0.15

Table 12.

Results of sensitivity analysis

CSF	Di + Rj	Di−Rj	Di + Rj	Di−Rj	Di + Rj	Di−Rj	Di + Rj	Di−Rj	Di + Rj	Di−Rj
C1	4.073	−0.509	3.858	−0.48	4.102	−0.525	3.885	−0.461	4.031	−0.489
C2	3.887	0.705	3.69	0.683	3.922	0.715	3.74	0.677	3.883	0.71
C3	3.907	0.276	3.697	0.262	3.949	0.251	3.764	0.238	3.907	0.258
C4	4.118	0.639	3.902	0.613	4.156	0.617	3.967	0.552	4.115	0.614
C5	4.009	1.331	3.802	1.256	4.036	1.347	3.846	1.285	4.031	1.352
C6	3.905	1.457	3.689	1.354	3.969	1.503	3.774	1.427	3.938	1.482
C8	3.56	2.016	3.256	2.125	3.487	2.122	3.445	2.003	3.669	1.92
C10	3.49	0.762	3.327	0.712	3.55	0.784	3.387	0.755	3.489	0.767
C11	3.75	−1.08	3.54	−1.057	3.857	−0.993	3.624	−0.987	3.747	−1.055
C12	3.713	−1.482	3.525	−1.484	3.76	−1.605	3.524	−1.486	3.744	−1.481
C13	4.049	−1.08	3.848	−1.056	4.112	−1.083	3.891	−1.041	4.037	−1.064
C14	3.788	−1.144	3.628	−1.105	3.762	−1.174	3.59	−1.112	3.76	−1.127
C15	3.735	−0.517	3.534	−0.5	3.767	−0.563	3.587	−0.566	3.721	−0.567
C16	3.662	0.176	3.474	0.157	3.674	0.188	3.515	0.216	3.652	0.229
C17	3.351	−1.552	3.2	−1.48	3.406	−1.583	3.225	−1.5	3.336	−1.549

Fig. 3.

Results of sensitivity analysis

Discussion

The FDM and DEMATEL techniques were utilised in the study to identify and evaluate the CSFs of blockchain technology adoption in the banking sector. Stakeholder viewpoints are important because many parties are engaged in the effective adoption of blockchain technology and since that adoption heavily depends on cooperation, support, customer involvement, risk management, and achieving regulatory compliances. Participants in this study included stakeholders like banks (commercial and retail banking), blockchain developers, lawyers, blockchain consultants, and blockchain researchers. From a stakeholder viewpoint, the findings of the study highlight the significance of each CSF and group these CSFs into "cause" and "effect" categories.

Cause group

Eight CSFs were found to be in a cause (facilitators) group, showing they have more influence on other CSFs than they receive from other CSFs. The maximum $″D-R″$ score is for ease of local & international legislation & regulations (C8), followed by user data privacy (C6), security & integrity (C5), availability of funds for implementation (C10), transaction cost efficiency (C2), scalability (C4), transaction storage (C3), and interoperability and standardization (C16). Working on these CSFs facilitates improving CSFs falling into the effect (dependent) group. Therefore, banks should emphasise developing CSFs associated with the cause (facilitator) groups on a priority basis to adopt blockchain technology for sustainable and resilient banking operations.

Effect group

Seven CSFs of blockchain technology adoption belonged to the effect (dependent) group. The technology investment and maturity group (C17) has the most dependency on other CSFs belonging to the cause group. The remaining CSFs of the effect group, in decreasing order, are blockchain talent availability (C12), incentive for miners (C14), integration with other cloud services (C13), professional consultation & advisory capability (C11), smart contract robustness & business case deployability (C15), and management/leadership buy-in (C1). The CSFs belonging to the effect group strongly depend on the cause group. These CSFs can be improved by working on the cause group. The findings corroborated the experts’ opinion that CSFs of effect group for sustainable and resilient banking operations are essential. However, organisations can improve these CSFs while working on the CSFs belonging to the cause group. For instance, professional consultation & advisory capability is essential for successful blockchain technology adoption; however, it is largely contingent upon the ease of local & international legislation & regulation and the availability of funds. Similarly, smart contract robustness & business case deployability are largely affected by user data privacy and security and integrity services offered by blockchain technology.

Ranking

The values of $\left(D+R\right)$ represent “prominence”, signifying the “total cause and effect”. A higher value of $\left(D+R\right)$ show the greater overall prominence (visibility/importance/influence) of CSFs concerning the overall relationship with other CSFs. On the other hand, the values of $\left(D-R\right)$ represents the “net effect”. The ranking based on the value of prominence indicates prioritisation based on total cause and effect, and classification based on net effect shows the influential order (net effect) of CSFs.

Scalability (C4) has the maximum $D+R$ score, followed by $C1>C13>C5>C6>C3>C2>C14>C11>C15>C12>C16>C8>C10>C17$. Scalability is the most significant CSF, and it needs maximum attention for successful blockchain adoption for sustainable and secure banking operations. Based on the value of $D-R$, CSFs can be organised in decreasing order as $C8>C6>C5>C10>C2>C4>C3>C16>C1>C15>C11>C13>C14>C12>C17$. The prioritisation of CSFs based on the $″D-R″$ value suggests that CSF management should be given utmost importance to CSFs belonging to the cause group because they can influence the effect group CFSs to change for the better. The most influencing CSF is “ease of local & international legislation & regulation”, which has the maximum effect on adopting blockchain technology in the banking sector. The study finds support from previous studies that universal regulations are the key criterion for the successful implementation of blockchain challenges [3]. Successful blockchain technology adoption encourages novel financial services; however, laws can regulate transactions. Furthermore, regarding data privacy, the inherent unique features of blockchain technology enhance vulnerabilities. Therefore, management and regulators must develop new privacy and security policies to encourage effective blockchain models [14, 50].

Validation of findings

The validity of the findings was ensured by sharing the results with a group of experts having more than 5 years of experience in existing organisations and more than ten years of total experience. The knowledge and experience of experts ensured that they were appropriate respondents and that their suggestions and feedback would be unbiased and close to reality.

The experts agreed that CSFs associated with the “cause group” should be given the utmost importance to realise the benefits of blockchain adoption for sustainable and resilient banking operations. They also suggested that moving from conventional banking operations to blockchain-enabled banking requires scalability, leadership buy-in, and integration with other cloud services. Local and international legislation & regulations, user data privacy, security and integrity, availability of funds, transaction cost efficiency, and interoperability and standardization are essential to enable an overall sustainable and secure blockchain-enabled system. In the words of a blockchain researcher & cryptocurrency start-up co-founder, ‘Scalability is the most essential thing for implementation of Blockchain Technology which is currently being looked on by Ethereum 2.0/Blockchain Lightning Network’. The opinions shared by the experts were consistent with the findings of the study. Therefore, validation using the insights of experts and published literature ensures the robustness of the findings.

Conclusion

The applications of blockchain technology in banking operations are identified as a significant innovation in the digitalization of banking operations and have the potential to make banking operations sustainable and resilient. While a significant amount of grey literature on blockchain technology is available, the state-of-the-art of CSFs for blockchain technology applications in banking operations has received limited attention in the form of rigorous and empirical studies [3]. While CSFs that are relevant in blockchain applications for sustainable and resilient banking operations have been studied fragmentarily in recent years [21, 22], empirical investigation related to evaluation and prioritisation has been largely overlooked. In addition, the interrelationships of these CSFs remain unclear, calling for further research. To respond to these deficiencies in the academic literature, the study took the perspective of stakeholder theory and critical factor analysis to answer the research questions. In this study, an integrated approach has been used based on the subjective data collected from stakeholders to assist banking organisations in the successful adoption of blockchain technology in their operations.

To address the RQs, an integrated research approach consisting of a literature review, semi-structured interviews, FDM, and DEMATEL was applied. To get the multi-stakeholder perspective, participants from different domains like retail and commercial banking, blockchain developers, blockchain researchers, lawyers, and consultants were considered in this study. While the literature review and semi-structured interviews revealed a set of 17 relevant CSFs, they also answered the first research question. The results of the FDM helped in understanding the CSFs associated with successful blockchain technology adoption in sustainable and resilient banking operations. DEMATEL provided important insights for researchers and practitioners by visualising the interrelationships among CSFs in a causal diagram and prioritising and ranking them, thereby addressing the second and third research questions. The DEMATEL method places eight CSFs in the cause group and seven in the effect group, reflecting driving and dependence relationships.

Theoretical implications

The study contributes to a growing body of knowledge within the blockchain technology adoption in the banking sector and paves the way for extensive research. First, it highlights the causation analysis of CSFs of blockchain technology adoption in the banking sector by putting forward a comprehensive list of CSFs contributing to the adoption of blockchain technology. Second, the CSFs and their interrelationships can facilitate the devising of strategies for adopting blockchain technology for sustainable and resilient banking operations. Third, the interrelationships among the causal enabling factors uncovered in this study and the influence of causal enabling factors on each other offer a good foundation for future studies towards developing conceptual frameworks on various aspects of blockchain technology adoption for sustainable and resilient banking operations. Empirical studies are scarce on blockchain technology adoption in the banking industry [3], especially on CSFs. In this context, this study helps answer what and how questions in theory building by transforming the poorly articulated mental model of CSFs into a well-defined model using the FDM-DEMATEL approach. The study offers answers related to “how” and “why” and provides a strong foundation for a statistical model for future research on blockchain technology adoption in the banking sector. Further, several studies have used stakeholder theory in the field of technology adoption and applications. But the grounding of stakeholder theory in the evaluation and prioritisation of CSFs for blockchain technology applications is scarce. Further, the use of CSFs theory is also novel in the field of blockchain technology. Therefore, to fill the gaps, this study has used stakeholder and CSF theories to examine the CSFs of blockchain technology adoptions. Above all, researchers aiming to validate their findings can benefit from the procedure discussed in this study for external validation.

Managerial implications

The study is helpful for practitioners planning blockchain technology adoption for sustainable and resilient banking operations. First, the potential causes of CSFs of blockchain technology adoption for sustainable and resilient banking operations are put forward. Also, the significance and ranking of these CSFs for successful blockchain technology are provided. Thus, the study gives insight on how to encourage the successful adoption of blockchain technology in banks, the essential CSFs for its adoption, and a priority basis for which CSFs to focus on more than others. The developed model is intended to assist decision-makers and policymakers in understanding each CSFs’ significance and devising the appropriate strategies or policies to overcome the same. Second, the findings indicate that giving importance to CSFs associated with the cause group would be essential due to their influence on the CSFs belonging to the effect group. The most important CSF that positively affected blockchain technology adoption in the banking sector is the ease of local & international legislation & regulation (C8). Therefore, to adopt blockchain technology, practitioners should stress the requirements of the regulations to achieve a higher level of adoption. Besides, user data privacy (C6) is essential for adopting blockchain technology.

Moreover, it is essential to recognise the importance of security and integrity (C5). This implies that practitioners should work on ensuring data privacy, security, and integrity of the blockchain-enabled system. The novel insights offered by this study would guide practitioners towards making sagacious decisions for improving influential causes, which would enhance the other passive causes.

Limitations and future research

Despite several implications of the study, limitations of the study need to be noted. The interrelationships derived from this study are not statistically validated. Applications of multivariate techniques on large-scale data and a focus on different industrial settings can help develop the findings' replicability. The CSFs identified in this study are high-level CSFs, and these factors can be further classified into sub-factors; therefore, detailed analysis of these sub-factors can be done in future research. Also, the study was conducted in India, and the specific conditions of other developing countries need to be incorporated into the model. Thus, the findings may apply to other countries in a similar manner. However, this study provides the opportunity to empirically verify these findings in different regions. Future research can also be carried out by expanding the scope of the study to geographically located countries to minimise region-specific biases. The study has not considered the dynamics among the identified causes. Therefore, future research should consider the dynamism associated with the CSFs of blockchain technology adoption for sustainable and resilient banking operations. Lastly, because the evaluation of FDM and DEMATEL subjective judgements was done by experts, the study can suffer from self-bias. Therefore, future research can use a fuzzy DEMATEL approach to minimise the issue of self-bias.

Appendix 1

See Table A1, A2, A3, A4, A5 and A6

Table A1.

Turnaround time reduction after blockchain technology adoption

Use case	Pre-implementation	Post implementation
LoC issuance	Days/months	Hours
FX trade settlement	Days	Minutes
Operational finance reporting	Days	Seconds

Declaration

Conflict of interest

There is no conflict of interest to declare for this manuscript.