The Effect of Organizational Culture on the Relationship of Big Data Analytics Capabilities and Firm Performance: An Explanatory Study
in the Banking Industry Master Thesis
Author: Alexandros Georgallides / Student No. 13302914 Date of Submission: 14/01/2022
MSc in Business Administration: International Business University of Amsterdam
EBEC: 20210924110938 Supervisor: Markus Paukku
Statement of Originality
This document is written by Alexandros Georgallides who declares 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.
Table of Contents
ABSTRACT ... 5
1. INTRODUCTION ... 6
2. LITERATURE ... 9
2.1. BIG DATA ... 9
2.2. RESOURCE-BASED THEORY ... 10
2.3. BIG DATA ANALYTICS CAPABILITIES ... 11
2.4. ORGANIZATIONAL CULTURE ... 13
2.5. THE COMPETING VALUES FRAMEWORK &ORGANIZATIONAL CULTURE ASSESSMENT INSTRUMENT ... 14
2.6. ORGANIZATIONAL CULTURE &BIG DATA ANALYTICS CAPABILITIES ... 17
2.7. FIRM PERFORMANCE ... 17
2.8. TRADITIONAL BANKING VS FINTECH ... 18
3. THEORETICAL FRAMEWORK ... 21
3.1. BIG DATA ANALYTICS CAPABILITIES &FIRM PERFORMANCE ... 21
3.2. BIG DATA ANALYTICS CAPABILITIES &ORGANIZATIONAL CULTURE ... 22
3.3. CONCEPTUAL MODEL ... 26
4. METHODOLOGY ... 26
4.1. RESEARCH STRUCTURE ... 26
4.2. SAMPLE &DATA COLLECTION ... 27
4.3. RESEARCH APPROACH ... 28
4.4. COMMON METHOD BIAS ... 29
4.5. MEASURES ... 29
5. RESULTS ... 34
5.1. DATA ANALYSIS ... 34
5.2. EVALUATION OF THE MEASUREMENT MODEL ... 35
5.3. EVALUATION OF THE STRUCTURAL MODEL:(SEM) ... 40
6. DISCUSSION ... 43
7. CONTRIBUTION ... 45
7.1. THEORETICAL IMPLICATIONS ... 45
7.2. MANAGERIAL IMPLICATIONS ... 46
8. LIMITATIONS AND FUTURE RESEARCH ... 48
9. CONCLUSION ... 51
10. REFERENCES ... 54
11. APPENDICES: ... 66
List of Tables and Figures
TABLE 1: ... 31
TABLE 2 ... 37
TABLE 3 ... 39
TABLE 4 ... 39
TABLE 5 ... 42
TABLE 6 ... 43
FIGURE 1: ... 26
This paper studies how the organizational culture (OC) affects the relationship of big data analytics capabilities (BDAC) and firm performance (FPER). The dimensions of market performance and operational performance were the selected measures of FPER, and they were investigated through the following central question: What effect does organizational culture have on the relationship between big data analytics capabilities and market performance (MP), and what effect does organizational culture have on the relationship between big data analytics capabilities and operational performance (OP). To assess the model, the study adopted a quantitative method of analysis for validation of and the effect of the constructs. Data was collected through surveys that were sent to management and c-executives with a degree of experience in big data. Focusing on the banking industry, the study aims to identify organizational cultural differences between traditional banks and fintech firms. Using partial least square path modeling (PLS-SEM), the study has validated and yielded further support for the relationship of BDAC and FPER. However, results did not indicate any significant effect of organizational culture on this relationship. Despite the complexity involved in analyzing organizational culture, the study yielded further support for the model and provided a new direction for how to assess the role of organizational culture in the adoption of big data analytics capabilities. The study also developed several theoretical and managerial implications that can contribute to the understanding of BDAC and OC, and how these factors influence market and operational performance.
Technology has impacted today’s business world due to its exponential growth. The implementation and usage of today’s technological tools has created many social and organizational opportunities and challenges (Matthias et al., 2017). An important aspect that has drawn the attention of many scholars and practitioners in regard to the growth of technology is the development and introduction of big data. Big data plays a crucial role as a resource. It can contribute to a company’s competitive advantage, but it requires new skills and capabilities to be processed (Davenport and Patil, 2012). According to a forecast provided by Statista, worldwide big data revenue is expected to increase to $103 billion by 2027, while in 2011 it has generated $7.6 billion (Mlitz, 2021). Though businesses may recognize the value big data may offer to access, manage, and analyze bulk volumes of data, many challenges exist due to other factors, such as cultural barriers, organizational adaptations and skill set shortages (Morabito, 2016; Dubey et al., 2019a; Mikalef et al., 2019).
“Information is the oil of the 21st century, and analytics is the combustion engine”.
The existing literature has shown that big data analytics capabilities (BDAC) have a significant impact on a firm’s performance (Rialti et al., 2019; Singh and El-Kassar, 2019;
Gupta and George, 2016). Côrte-Real et al. (2019) have argued that big data analytics (BDA) has been considered the next “frontier” in data science, as it creates potential business opportunities through the interpretation and analysis of data, therefore influencing the strategy.
The argument stated by the authors is supported by other literature that conceptualizes BDA as a technologically enabled organizational capability that can quickly integrate large volumes of complex data and provide valuable insights to organizations, enabling them to create and sustain a competitive advantage (Wamba et al., 2017; Akter et al., 2017). Various advantages can be developed based on the use of big data, as it can be incorporated into the development
of strategy. For example, firms that integrate big data into their strategy may produce real-time forecasts, monitor the business as well as the market, identify crucial hidden points that may impact the decision-making and optimize production and operational costs (UpGrad, 2019).
Though more emphasis in the literature has been placed on technical aspects related to BDAC, less research has been done on the organizational changes required to adopt BDAC and the strategic implications of big data. Thus, more research has been recommended and is needed (Lunde, Sjusdal and Pappas, 2019, pp.164–176; Wang, Kung and Byrd, 2018).
Moreover, there are various social, technological, and human factors that have just started to emerge and need to be addressed to understand the capabilities of big data analytics and how it creates a long-term advantage (Matthias et al., 2017). Dubey et al. (2019) have argued that in the big data predictive analytics (BDPA) context how factors such as organizational culture may affect big data capabilities is not well defined, and firms may be reluctant to adopt a data- drive culture, as employees or management may not want to change. Scholars have found that research has been lagging regarding organizational culture. Organizational has a crucial role for organizations that are in the process of or willing to adopt big data capabilities (Lunde, Sjusdal and Pappas, 2019, pp.164–176).
The purpose of this study is to contribute to the current academic literature by focusing on the effect of organizational culture on big data analytics capabilities, as it is a concept that has not received enough attention by scholars on the relationship of big data analytics and firm performance. This study proceeds to identify the cultural profile types that may be more suitable to adopt big data analytics capabilities, as no previous literature was found that showed which cultural types increase big data analytics capabilities (Cameron & Quinn, 2006). In addition, the research provides practical implications for the management positions in the banking industry, as the study defines the differences between the organizational cultures of
traditional banking and fintech firms and analyses the effect these cultures have on the relationship between big data capabilities and firm performance.
The findings of this study were developed by focusing on the banking industry.
Important information from traditional banks and from fintech firms that have been developed based on tech know-how and data-driven culture were gathered and their organizational culture types, which have either been in place for decades or more recently created based on the usage of big data tools and capabilities, were compared.
The existing literature has mainly focused on different types of industries and activities, such as healthcare, retail, manufacturing and supply chain (Santoro et al., 2019; Dubey et al., 2019a; Dubey et al., 2019b; Wang, Kung and Byrd, 2018). The reason this study focuses on the financial industry is due to its evolution across the world. Many start-ups have emerged and introduced innovative technological solutions that fill the gaps in the industry caused by market inefficiencies (Visconti, 2020). Existing literature has defined the historical roots of traditional banks and the new needs of customers around the world. Many gaps have been identified in which opportunities appear for the development and entry of “fintech” firms.
This study contributes to the scientific literature by expanding the empirical findings on how cultural organization affects the relationship between BDA capability and firm performance. The study has been structured in the following manner. First, the existing literature on big data is introduced. Then the theoretical framework model that was developed is outlined, and the methodology the study followed to obtain its findings is discussed. The next section is based on the findings and presents the results. Following is a discussion of the study, and several limitations and future research topics are addressed. Lastly, a conclusion is provided.
Based on the existing literature, this thesis develops a concept of the relationship between big data analytics capabilities and two dimensions of firm performance: market and operational performance. Following, the study examines the effect of organizational culture on this relationship. The current topic is relevant and new, and it is important to incorporate the perspective of organizational culture to understand better how big data analytics capabilities impact firm performance (Grant and Pollock, 2011). In this section, the study analyzes what exists currently in the literature of big data and organizational culture.
2.1. Big Data
Big data has become a very interesting topic among practitioners and scholars, as the volume of data it generates can offer businesses valuable insights, help create significant value and give rise to a competitive advantage over competitors. Big data also allows for the development of data-driven decision-making and the processes to organize, learn and innovate at various levels (Manyika et al., 2011; Wamba et al., 2017). Despite the significant interest on this topic, scholars and practitioners still have different definitions. For example, Dumbill (2013) has defined big data as data that exceeds the capacity of regular database systems organizations keep today. Manyika et al. (2011) have defined big data as “data sets whose size is beyond the ability of typical database software tools to capture, store, manage and analyze.”
Moreover, others have defined big data as data assets that need innovative forms of information processing for enhanced insight and development of decision-making (Matthias et al., 2017).
Big data may not have a universal definition, but most scholars focus their definitions on similar antecedents, and they are discussed in the next section.
Since its inception, big data has been described based on three v’s: variety, velocity and volume. After a while, some scholars introduced to more v’s to the model: value and veracity (Frizzo-Barker et al., 2016). Most scholars and practitioners still refer to only the first three
v’s, as BD hinges on concerns of high volumes of data that have a high variety and a high velocity, meaning data are constantly generated (Matthias et al., 2017). Volume which is the first dimension in the concept of big data, is concerned with an unmatched quantity of available and storable data for firms and organizations. Variety, which is developed as more amounts of data are added, concerns the type of data collected and stored, such as text, numbers, images and videos. The third dimension, velocity, is concerned with the dynamics of the data, mainly the speed of their creation and use (Morabito, 2016).
2.2. Resource-Based Theory
To understand how big data can develop big data analytics capabilities, the resource- based theory (RBT) constructed by Barney (1991) is examined. The antecedents of big data Analytics Capabilities (BDAC) are based on a model developed by Gupta and George, which was influenced and built based on RBT and has been validated by other scholars.
Barney’s (1991) resource-based theory is based on a firm’s resources and includes all types of assets, organizational processes, firm characteristics, information and knowledge owned and controlled by the firm. These assets enable the firm to implement strategies that will enhance its capabilities, efficiency and effectiveness. RBT allows a firm to construct a powerful framework that unites its many resources to develop a competitive advantage. If BDAC falls under the VRIO/VRIN framework, it may develop into a sustained competitive advantage. Firms may integrate those resources and/or capabilities to avoid duplications or imitations. If equity is not available for the acquisition or development of those resources, firms may begin to outsource using exclusive contracts. However, authors have argued that firms with internal resources have a greater determinant of strategic advantage, because they do not depend on the changes of the external environment (Barney, 1986b).
Upadhyay and Kumar (2020) have also suggested that the internal resources of a firm, such as human capital, should develop skills, such as analytical capabilities, to remain
competitive and sustain firm performance. But a firm should define its resources and match the advantages it provides with the external pressures of the environment such as opportunities and threats (Das and Teng, 2000). Resources of a firm may be categorized based on three different categories: physical capital (tangible resources), organizational capital (intangible resources) and human capital (Barney, 1991). Based on these three categories, firms may evaluate their resources and identify what needs to be added or changed to start developing big data analytics capabilities.
The existing literature has identified various characteristics that define the value and the development of big data analytics capabilities. However, it is important that organizations evaluate all the resources that are needed to build strong capabilities and sustain a competitive advantage (Mikalef et al., 2017a; Gupta and George, 2016). An important aspect noted in the literature was that despite heavily investing in big data, companies may still not be organizationally ready to adapt to structural or operational changes or may not have the knowledge required to understand the intelligence extracted from big data (Gupta and George, 2016). This is an important aspect, as it reflects directly on the human capital resources of a firm, which may lack the knowledge to benefit from the volume, variety and velocity of big data analytics.
2.3. Big Data Analytics Capabilities
In the existing literature, scholars have argued that big data analytics capabilities (BDAC) have a significant influence on a firm’s performance and on the creation of a competitive advantage (Gupta and George, 2016; Wamba et al., 2017; Akter et al., 2017).
Srinivasan and Swink (2017) have defined BDAC as “an organizational facility with tools, techniques, and processes that enable a firm to process, organize, visualize, and analyze data, thereby producing insights that enable data-driven operational planning, decision-making, and execution.” In general, big data analytics capabilities focuses on the ability to manage a large
volume of disparate and unstructured data to allow users to implement data analysis and reaction (Wang, Kung and Byrd, 2018). In other words, BDA refers to a firm’s ability to orchestrate and manage its big data-related resources (Mikalef et al., 2017).
Since the introduction of big data in the business environment, firms have had the opportunity to carry out data-driven decision-making rather than regular decision-making based on intuition, as big data extract results or a decision based on a mass amount and variety of data (Srinivasan and Swink, 2017). Andrew McAfee and Brynjolfsson (2012) have argued that data-driven decision-making may be a better method for making decisions, because it relies on evidence and facts that have been analyzed through a system rather than the intuition of experts or individuals. As Geoffrey Moore, an American consultant and author states, “Without Big data analytics, companies are blind and deaf, wandering out onto the web like deer on a freeway.”
Exploiting big data requires new capabilities to both control information ﬂows that appear internally and externally and transform them into core competencies that can become strategies for products and services that meet market needs (Morabito, 2016). Organizations may outsource these capabilities to retain an advantage if they struggle to create or adopt such capabilities internally due to resistance by management or employees (Raguseo, 2018). To gain the targeted resources related to big data capabilities and technologies in a shorter time frame, organizations may also proceed with acquisition of a firm that has those resources already, thus choosing to grow rather than develop the capabilities organically (Uddin, 2012). Principally, an organization needs to develop and sustain talent, management and tech skills that can process and integrate mass amounts of data to sustain its competitive position and enhance its performance.
Organizations currently are surfing in a large pool of data that is either too voluminous or too unstructured to be analyzed via traditional means (Davenport, Bean and Barth, 2012).
Due to these conditions and the fact that big data is a relatively new concept, organizational culture is a crucial role that needs to be analyzed to define how organizations can adopt capabilities that will help them gain the advantages offered by big data.
2.4. Organizational Culture
Though organizational culture has been studied by many scholars, a clear definition does not exist. Based on Barney (1986), organizational culture is “a complex set of values, beliefs, assumptions, and symbols that define the way in which a firm conducts its business.”
Schein (2016) defines organizational culture as “a pattern of shared basic assumptions that was learned by a group as it solved its problems of external adaptation and internal integration, that has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems.” It is the pattern of basic underlying assumptions that a group or an entire organization has created in learning processes to manage with its external factors and internal integration, and it has worked well enough to be considered valid, and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems and how the organization operates based on the external factors and internal (Schein, 1984; Liu et al. 2010).
Organizational culture can be explained through three levels. The first level relates to artifacts. This describes aspects such as office layout, employees’ dress and appearance at work and publicly available documents. The next level regards the assumptions that exist in the culture of a firm. These are such things as taken-for granted beliefs and norms, and the mental models used by managers and employees to make sense of the environment. The third level concerns the values of a firm. These are elements such as the rules that describe the daily behavior of the employees and management (Schein, 1990; Alavi et al. 2006; Joseph & Kibera, 2019). Studies have found that organizational culture serves as an important intangible resource, as a barrier to imitation and as an element that can have powerful effects on
performance, whereas other scholars suggest that performance can be understood by identifying the organizational culture of a firm (Colyer, 2000; Joseph & Kibera, 2019).
Organizational culture impacts the internal makeup of a firm and its adaptation to the external changing environment. Considering these two facts together, organizational culture represents a valuable factor that impacts performance outcome. Notably, Ogbonna and Harris (2000) have argued that when a firm achieves compliance between organizational culture and strategy, it creates a superior performance. Organizational culture can be stated to be the informal rules related to how things are done that exist in an organization. It is the personality of a company and is seen mainly through the perspective of the employees (O’Reilly, Chatman and Caldwell, 1991).
2.5. The Competing Values Framework & Organizational Culture Assessment Instrument
Despite the complexity of the organizational culture concept and its connection with a firm’s performance, scholars have developed several models to assess and measure specific aspects of organizational culture. In this study, the organizational culture of a firm is assessed and evaluated through the competing values framework (CVF), developed by Kim Cameron and Robert Quinn, in combination with the organizational culture assessment instrument (OCAI) that was developed on the theoretical model of CVF as a measurement instrument (Cameron and Quinn, 2006; Kwan and Walker, 2004). Of the three levels of organizational culture, values are the most easily studied, as basic assumptions are not formally stated or written, and artifacts are hard to translate into normal meanings (Schein, 2016). The chosen model has been identified as one of the most influential and commonly used to understand and assess a company’s organizational culture based on its values (Yu and Wu, 2009).
Cameron and Quinn (2006) developed the CVF model to use two dimensions that illustrate four different cultural types. The first dimension is related to organizational focus.
This ranges from an internal focus, which regards the development of talent and employees in the organization, to an external focus, which is concerned with the development of the organization itself (Yu and Wu, 2009). The second dimension is based on assessing the organizational structure from a perspective of stability and flexibility. Kwan and Walker (2004) have stated that the CVF model has become the dominant model for quantifying organizational culture data. It focuses mainly on staff climate, leadership style, bonding systems and prioritization of goals (Upadhyay and Kumar, 2020). Following the development of these two dimensions, four cultural types were identified: hierarchy, clan, adhocracy and market.
Hierarchical organizations are more process oriented, as they seek control and share similarities with stereotypical bureaucratic corporations (Tharp, 2009). Such organizations are defined by their degree of stability and control, and by their internal focus and integration.
These organizations maintain a clear organizational structure, formal rules and procedures, and well-defined responsibilities (Yu and Wu, 2009). This cultural type has been traced to the concept that was developed in early works made by Weber in 1947 on modern organizational management. This study assumes that firms with a hierarchical culture may be reluctant to make in changes in their business models and strategies to adopt big data capabilities due to a need for stability.
Clan culture seeks collaboration by emphasizing flexibility and discretion. Scholars have argued that this type of culture builds an atmosphere of collectivity and emphasizes enhancement and employee development. Companies with this type of culture may operate based on semi-autonomous teams to maintain flexibility (Tharp, 2009). As this culture type enhances the flexibility and development of employees, it may be more inclined towards the adoption of big data analytics capabilities.
Adhocracy is more related to a dynamic, entrepreneurial culture, but it too emphasizes flexibility and discretion (Hofstede et al., 1990). Organizations under this type of culture are more focused on differentiation. Moreover, due to various social, economic and technological changes in the business context, this culture type values flexibility over older corporate attitudes that may be less efficient (Tharp, 2009). Moreover, this culture understands that to find new opportunities in the environment and sustain a competitive advantage, organizations need to adapt by developing new resources. Both adhocracy and clan culture can be seen as the organizational culture types that represent firms that are willing to adopt BDAC to sustain a competitive advantage and be influenced by the technological growth and change in the business context.
While some cultural types focus internally, a market cultural type develops an external oriented approach to elaborate on differentiation and respond to competitive challenges that arise either from the local market or from overseas. This cultural type is focused on developing relationships with suppliers and stakeholders to achieve success (Yu and Wu, 2009).
Furthermore, market culture has been found in organizations operating in uncertain and unstable environments, thus where flexibility is needed (Maher, 2000).
Due to the differences that exist in the units of an organization, in regard to functions, location and team members, an organization may follow a combination of the described cultural types. For example, a department of an organization may need to be more task- oriented, while other departments may need to keep strong inter-relationships to enhance flexibility in uncertain times and avoid conflicts. In general, there is not a specific culture that is superior to another, and a company may operate within all types of culture, to some extent, to meet the external pressures of the environment (Maher, 2000).
2.6. Organizational Culture & Big Data Analytics Capabilities
As mentioned earlier, organizational culture plays a crucial role in the adoption and deployment of big data capabilities. Bean (2021) conducted a survey of 85 Fortune 1000 companies and found out that despite the high level of committed investments organizations make towards big data and artificial intelligence initiatives, firms still struggle to extract value from their big data and transform themselves into data-driven cultural organizations. Bean argues that the organizational environment, business processes, skill sets and traditional organizational cultures are reluctant to make the necessary changes that need to be done.
Alavi, Kayworth and Leidner (2006) have found that organizational culture has the power to influence technology selection and appropriation in an organization. Despite the emergence of scholars studying the concept of big data through many management and organizational theories, via systematic literature review, other scholars have found that organizational culture is still a new aspect in the field of big data (Lunde, Sjusdal and Pappas, 2019, pp.164–176). In addition, practitioners have argued that cultural factors that exist in organizations are the reason why adopting a data-driven culture remains difficult (Waller, 2020). Management and employees find it hard to switch from their current business or organizational model to a data-driven culture. Many suggestions have been given by practitioners to solve this challenge; however, firms still struggle to adapt.
2.7. Firm Performance
Firm performance as a construct has been studied by many scholars and based on different measurements. Despite the existing literature and the contribution of scholars, there is not a clear definition nor specific dimensions that construct firm performance (Gupta and George, 2016; Rai, Patnayakuni and Seth, 2006; Selvam et al., 2016; Taouab and Issor, 2019).
Selvam et al. (2016) have stated that firm performance is one of the categories that falls under the concept of organizational effectiveness, and they argue that it is defined by two
aspects: operational and financial outcomes. However, other scholars to have determined firm performance through profitability performance, which focuses on ROA, NET income revenues or economic value added (EVA). Still, other authors have used environmental performance as a measure of the firm performance, which focuses on the number of projects to improve the environment and level of energy intensity. Thus, as there is no specific definition or dimensionality, this study’s measure of firm performance is based on Gupta and George’s framework that defines firm performance “as the extent to which a ﬁrm generates superior performance with respect to its competitors.”
Gupta and George (2016) developed their dimensionality using existing literature of international strategy by measuring market performance (MP) and operational performance (OP). Among operational performance, many scholars have included aspects such as financial results generated by sales, profitability and return on investment, while market performance has been related to information about a firm’s success in new entries and the introduction of new products or services in existing and new markets (Gupta and George, 2016; Rai, Patnayakuni and Seth, 2006). These dimensions have not only been used by scholars, but they have also been validated in empirical studies (Ravichandran & Lertwongsatien, 2005).
2.8. Traditional Banking vs Fintech
Due to the rapid growth of many sectors in the digital era, financial services have been transforming through the power of technology. Today, new start-ups, referred to as FinTech, offer services comparable to those of traditional banks. Dorfleitner et al. (2017) have stated that the term “FinTech” describes companies that offer financial services supported and developed through innovative technologies, such as artificial intelligence, machine learning and big data. Moreover, the financial stability board defines fintech as “technologically enabled innovation in financial services that could result in new business models, applications, processes or products with an associated material effect on financial markets and institutions
and the provision of financial services” (Financial Stability Board, 2019). FinTech firms build digital products and services that mirror the products and services traditional banks offer via traditional methods. In addition, such start-ups may provide peer-to-peer to lending, alternative payment options, digital bank accounts and digital wallets. To simplify the definition for the purposes of this study, fintech firms are understood as a firm that provides any financial service conducted only online, through technology and technological services.
Fintechs are not hyper-regulated deposit institutions like traditional banks. Such firms are more focused on tech investments, while traditional banks may be seen as more interested in capital and labor (Visconti, 2020). Another difference between traditional banks and fintech firms is that traditional banks have a physical presence with an established HQ, regional HQ and branches across the world. In contrast, FinTech firms have an online-only presence, thanks to advancements of tech, and are able provide faster and more flexible customer-centric service (Orlando, 2017; Nicoletti, 2017). Despite their online-only presence, FinTech firms may have the opportunity to establish physical operations in territories. However, laws and regulations would need to be considered in the scenario of running a physical operation like a traditional bank (Financial Stability Board, 2019). In addition, fintech firms have the advantage of lower operating costs, consequently it may not be advantageous to absorb costs created due to a physical presence.
Notably, fintech firms have managed to gain a significant market share that was once owned by traditional bank. FinTech firms provide services that traditional financial institutions are less efficient in performing and can explore a wider pool of users due to global internet connections accessed through many devices, such as smartphones (Navaretti et al., 2018).
While traditional banks have been building their services and long-term relationships with stakeholders for decades, FinTech firms process their information using big data analytics that data is collected through digital channels. According to a survey conducted by McKinsey &
Company, FinTech firms have managed to build a significant degree of customer trust and have thus created a threat to the relationship between customers and traditional banks (Krivkovich et al., 2020). Moreover, the CEO of Morgan Chase, Jamie Dimon, in an interview has stated that fintech firms have created “enormous competitive threats” to traditional banks due to their advantage of processing big data and building financial products digitally (as cited in Locke, 2021).
Despite the threat, many traditional banks have invested in the development of digital and analytical tools that would allow them to remain competitive. To sustain competitiveness and create value for their current and potential new customers, many traditional banks are transforming their business models to be more similar to those of fintech businesses (Navaretti et al., 2018). However, many big tech firms and traditional retail firms have started providing financial services as fintech activities, for example Amazon, Apple, General Electric and Tesco. With their respective knowledge and resources in tech and customer experience, these firms are able to provide faster and more efficient services through “robo-adviser” services /(Panetta, 2018; Navaretti et al., 2018).
The financial services industry serves as a strong example for the purposes of the study.
It provides two different types of organizations that contain differences in organizational culture due to the different antecedents that created them. Following Gupta and George’s (2016) opinion on the disruption on various industries because of new business models, such as Airbnb or Uber in the taxi industry, the study emphasizes on the entrance of FinTech firms in the financial services industry.
Based on the existing literature, there has been a lack of attention by scholars on the role of organizational culture in the relationship between big data analytics capabilities and firm performance. Firms that were founded in the digital era may have a more data-driven culture than traditional organizations that have long historical roots. Moreover, due to the
development and differences of FinTech firms in the financial industry, this study focuses on understanding the effects of organizational culture and compares the cultures of the two types of financial institutions that exist today, FinTech and Traditional Banks. The following are the research questions this study uses to address the gap in the literature:
RQ1: What effect does organizational culture have on the relationship between big data analytics capabilities and market performance?
RQ2: What effect does organizational culture have on the relationship between big data analytics capabilities and operational performance?
3. Theoretical Framework
In this section, the theoretical framework and hypotheses related to the topic are developed. Previous literature has addressed that organizational culture plays a critical role in the adoption of big data analytics capabilities and has often cited organizational culture as one of the main reasons why big data initiatives fail (Lunde et al., 2019, pp. 164–176).
Organizational culture plays an important role as it is a concept incorporated in a firm’s strategy, structure and processes. However, it is a concept that has not yet got enough attention.
3.1. Big Data Analytics Capabilities & Firm Performance
When reviewing the existing literature, scholars are shown to strongly agree that there is a positive relationship between big data analytics capabilities (BDAC) and firm performance (FPER). The benefits big data may offer to an organization, allow the firm to optimize its operations by processing unstructured data and thus increasing its performance both market- wise and operational-wise (Garmaki et al., 2016; Gupta & George, 2016). However, to be able to develop and exploit such capabilities, firms need to be aware of the antecedents required, such as tangible resources, intangible resources and human skills. Gunasekaran et al. (2017) have found that BDA has a positive influence on several dimensions of a firm’s performance.
Other scholars have examined the relationship of BDAC and FPER in specific industries and found different direct and indirect positive impacts (Wang et al., 2018). BDAC can enhance an organization’s performance and its business model, because it is able to optimize activities of a firm. BDAC may influence both market and operational performance. Therefore, based on the existing literature and results, the following hypotheses were tested in this study:
H1a (+): Big data analytics capabilities (BDAC) have a positive effect on market performance (MP)
H1b(+): Big data analytics capabilities (BDAC) have a positive effect on operational performance (OP)
3.2. Big Data Analytics Capabilities & Organizational Culture 3.2.1. Hierarchy Culture
Hierarchy culture is the oldest of the four cultural types developed by Max Weber. It is a culture based on the bureaucratic system and is characterized by order, meritocracy, hierarchy, impersonality and accountability (Cameroon & Quinn, 2006). This culture describes a more formal and structured workplace where overt rules and policies are applied to keep the organization together (OCAI Online, 2019). Although hierarchy culture may offer different benefits to an organization, the formal procedures of the culture generally lack innovation and entrepreneurship (Davis & Cates, 2018). Based on these characteristics and the bureaucracy that exists in traditional global banks, the study assumes that hierarchy cultural is more likely to be found in traditional banks than in fintech firms. In addition, due to a lack of innovation, the study assumes a negative relationship on the adoption of big data analytics capabilities and firm performance. The following are the proposed hypotheses:
H2a (-): Hierarchy culture negatively moderates the relationship between big data analytics capabilities (BDAC) and market performance (MP)
H2b (-): Hierarchy culture negatively moderates the relationship between big data analytics capabilities (BDAC) and operational performance (OP)
3.2.2. Market Culture
The market culture type has been argued by authors as the foundation of organizational effectiveness (Cameroon & Quinn, 2006). Its focus is on competition and productivity. This type of culture refers to an organization that functions based on the external environment rather with internal environments. Thus, a market culture firm is focused on transactions and relationships more than vertical integrating activities. The goal of this organizational culture type is to beat the competition and penetrate the market by creating a competitive advantage and market superiority (Barton & Court, 2011). Although a market culture firm may outsource several activities that could be related to big data, resources, such as human skills, are still needed internally in the organization to understand and interpret the BDA provided by a third party. Due to the evolving market conditions and the development of big data in many industries, such cultural type seeks to create a competitive advantage, market superiority, and penetration of the market. The proposed hypotheses are the following:
H3a (+): Market culture positively moderates the relationship of big data analytics capabilities (BDAC) and market performance (MP)
H3b (+): Market culture positively moderates the relationship of big data analytics capabilities (BDAC) and operational performance (OP)
3.2.3. Clan Culture
Clan culture as a cultural type focuses internally, keeps flexible processes and creates a friendly workplace environment (OCAI Online, 2019). Its main characteristics consist of increasing employee empowerment by participating and being involved from different levels of the hierarchy in decision-making, thus providing more opportunities for employee recognition (Davis & Cates, 2018). Organizations that are based on clan culture emphasize
human development and the role of the leader as a mentor and coach. Moreover, this culture type allows employees to constantly make job rotations to understand all activities of an organization (Cameroon & Quinn, 2006). An example of the clan cultural type was found in Japanese organizations after World War II. A characteristic of why such a cultural type was adopted in Japan was the rapidly changing, turbulent environment that made it difficult for managers to plan far in advance and to reduce uncertainty in decision-making. Thus, coordination between employees was important to make sure everybody shared the same values, beliefs and goals. Keeping a flexible environment in the organization, allows such culture to enhance human development, sustain success and reduce uncertainty. The following are the proposed hypotheses for this culture type:
H4a: Clan culture positively moderates the relationship of big data analytics capabilities (BDAC) and market performance (MP)
H4b: Clan culture positively moderates the relationship of big data analytics capabilities (BDAC) and operational performance (OP)
3.2.4. Adhocracy Culture
Adhocracy is the fourth cultural type and is developed based on an external focus and a degree of flexibility. The main objective of the adhocracy culture is to maintain and enhance a central concept of creativity, innovation and entrepreneurship (OCAI Online, 2019). Leaders of organizations with this culture type are seen as innovators and willing to take high risks. The key long-term goal of adhocracy is to develop new superior resources that will outperform the competition (Cameroon & Quinn, 2006). This cultural type has been found frequently in tech start-ups and technology-driven industries, such as Airbnb and Uber (OCAI Online, 2019).
Scholars have argued that innovation and creativity enhance a firm’s ability to manage challenging situations and successfully adapt (Naranjo‐Valencia et al., 2011; Davis & Cates, 2018).
Given the drive towards innovation and entrepreneurship connected to this culture type, this study assumes that firms with an adhocracy culture are in favor or may have already adopted the necessary capabilities for big data analytics. Striving to be innovative, these organizations may have also already developed the human skill and necessary resources to integrate big data. Many fintech firms may use this organizational culture type. However, as traditional banks compete in the same industry as fintechs, some may have chosen to adopt adhocracy to sustain their market competitiveness. Although traditional banks may want to follow adhocracy, the bureaucratic system they maintain may create a reluctancy to change into a more flexible and entrepreneurial system. The study’s proposed hypotheses for this culture type are the following:
H5a: Adhocracy culture positively moderates the relationship of big data analytics capabilities (BDAC) and market performance (MP)
H5b: Adhocracy culture positively moderates the relationship of big data analytics capabilities (BDAC) and operational performance (MP)
3.3. Conceptual Model
To provide a visual representation of the variables and hypotheses that will be tested the following conceptual model was developed:
This section provides an overview of the study’s research design. The main objective of this study was to assess and understand how organizational culture can affect a firm’s big data analytics capabilities and performance. The data for this research were collected through surveys sent to management employees working in either traditional banks or fintech firms in the US and EU. Following the explanation, the research structure of the study is explained, including what type of data was collected and how it was collected. Next, information regarding the data’s reliability and validity, and the analytical procedure the study followed is provided.
4.1. Research Structure
Firstly, research philosophy is explained to understand how knowledge is developed and what its nature is, as important assumptions needed to be considered to underpin the
research strategy and its structure (Saunders et al., 2016). The study builds on the philosophy of positivism, because it allows the research to conduct quantifiable observations and proceed to statistical analyses. In addition, the positivism philosophy allows the research to be concerned with facts associated with production or performance (Saunders et al., 2016). The study’s research was approached through an explanatory structure to understand and explain how organizational culture influences the relationship of BDAC and FPER. By collecting quantifiable observations and correlations regarding the variables developed in the framework of explanatory research, the study may be able to identify which organizational culture type creates better results for firm performance (FPER).
An exploratory research structure was not considered for this work, because it seeks new insights to assess a specific phenomenon or investigate a problem via qualitative methods, such as through a review of the literature or focus group interviews with experts. As the research was designed to observe quantitative data, the study was based on a discripto- explanatory approach (Saunders et al., 2016). Moreover, the explanatory structure of the study was based on a deductive approach. Therefore, support was provided to test the hypotheses developed through the theoretical framework so that the original theories introduced in the study could be tested (Trochim et al., 2016). An inductive approach to the research would have required an exploratory structure to gather observations and the theories to be developed based on the observations. Thus, an inductive approach was not feasible for the purposes of the study.
4.2. Sample & Data Collection
Due to the limited availability of time, the research was developed as a cross-sectional study that would describe the data of a chosen population at specific point in time. If more time were available for the study to be conducted, then various developments and changes in the characteristics of the chosen variables based on the targeted population could have been detected. The study’s data were based on a quantitative survey of a sample population. The
goal of the survey was to gather the data, analyze it and provide answers to the study’s hypotheses by obtaining responses from a large representative population. Performing a qualitative analysis would have given the study more of an exploratory structure and decreased its external validity due to a smaller sample.
4.3. Research Approach
The sample that was used in the study was chosen based on specific criteria that would support the validity of the responses. As mentioned, the study focused on firms from the EU and the US. Thus, managers employed in such firms with a degree of experience in big data analytics or overseeing big data projects were chosen as the appropriate sample for the study’s analysis. The study focused on managers, because they have worked in their firms for several years and have relevant experience related to big data. If surveys had been sent to analysts or associates, the results may have contained a higher bias due to the workers’ fewer years of experience in such positions. The method chosen to select and approach the sample was based on convenience sampling. By using this sampling method, the study’s bias was reduced, and focus could be placed on a sample that would provide an external validity for firms in the financial industry established either in the US or in Europe.
The survey itself was developed using SurveyMonkey and was sent to the chosen employees via LinkedIn, whereas sample’s email addresses were collected through Rocketreach. After sending a first email requesting the employee to answer the survey, up to three reminders were sent to those who still had not replied. These reminders were sent 10 days apart from each other.
In the following section, an explanation of the adopted theoretical models is provided.
Importantly, these models represent the construct of the study’s measurements.
4.4. Common Method Bias
Participants in behavioral research often respond to questions in a socially preferable way that can cause measurement errors and generate biased results. This effect on research is referred to as common method bias (Podsakoff et al., 2003). To reduce and avoid such bias, the respondents were allowed to answer the survey anonymously, so they would not feel as though they were being judged. Regarding reliability, the study focuses on constructs that have been studied, tested and validated by previous research papers. To avoid any incomplete surveys, the respondents were required to answer all questions. Lastly, questions related to the independent and dependent variable were randomly positioned in the survey to prevent respondents from maintaining a consistency in their responses to questions that focus on a similar topic, the so-called consistency motif (Podsakoff et al., 2003).
The survey contained questions related to each variable in the framework and control variables constructed for this study. A complete list of the items used in the survey is found in Appendix A.
4.5.1. Independent Variable
To assess big data analytics capabilities, the study adopted the measurement scale model developed by Gupta and George that is based on the resource-based theory. This measurement consists of 32 individual items that are divided into three dimensions: tangible resources, such as data technology and basic resources; human skills, such as managerial and technical skills; and intangible resources, such as a data-driven culture and intensity of organizational learning (Gupta and George, 2016). The survey questions asked the respondents to report the extent to which they agreed or disagreed that their organization possessed or had developed the necessary resources such as tangible, intangible and human skills.
BDAC is a three-dimensional reflective-formative construct that was developed and validated in studies by Gupta and George (2016). Previous scholars have discouraged the use of formative constructs in business disciplines because they conceptualize formative constructs and the presence of interpretational confounding, such as the difference that exists between empirical and nominal meanings of a construct. However, other scholars have recommended formative construct’s feasibility by stating “when grounded theoretically and analyzed properly, formatively specified constructs can play a valuable role in IS research.” (Petter et al., 2012).
The independent variable BDA capability was developed as a three-dimensional formative construct and based on resources related to big data, such as tangible, intangible, and human skills (Gupta and George, 2016). All constructs were developed and measured on a seven-point Likert scale. Gupta and George (2016) have developed and validated the chosen constructs. Some of the constructs are formative, while others are reflective. Formative constructs can be formed via a linear combination of causal indicators. These indicators define their construct and do not reflect it (Bollen & Bauldry, 2011). For example, first-order constructs such as data, technology and basic resources are formative whereas data-driven culture, managerial skills, technical skills and organizational learning are reflective constructs.
Second-order constructs that are formative include tangible resources, human skills and intangible resources. The third-order construct that was formative was BDA capabilities.
Characteristics of the first-order BDA capability constructs that have been used and validated by scholars can be found in Table 1. To sustain a degree of credibility, the study did not incorporate any other formative or reflective constructs in the main model Gupta and George (2016) developed. In Appendix B, the framework of the BDAC construct can be found.
The model was developed by Gupta and George (2016) and has been supported by Cronbach’s alpha, as scores were above 0.8 for all constructs.
The model developed by Gupta and George has been empirically validated and supported. The questions provided in the survey that are related to this construct were answered on a 7-point Likert scale (1 = strongly disagree; 7 = strongly agree).
Characteristics of First-Order BDA Capability Constructs (Gupta & George, 2016)
4.5.2. Dependent Variable
Firm performance was based on the measurement of operational and market performance. A clear definition of what constructs a firm’s performance has not yet been specified, although several scholars who have studied this topic have identified different dimensions based on several characteristics to determine a firm’s performance. These dimensions include profitability performance, market value performance, growth performance, employee satisfaction and environmental performance (Selvam et al., 2016). Nevertheless, specific dimensions have not yet been used to specify what exactly firm performance is.
Therefore, this study relied on Gupta and George’s model, because the scholars have already examined and validated the dimensions of operational performance and market performance in relation to big data analytics capabilities (Ravichandran & Lertwongsatien, 2005).
Both operational performance and market performance are reflective, and each consists of four items that allowed the study to quantify them based on a 7-point Likert scale (from strongly disagree to strongly agree). Market performance and operational performance were assessed separately, as each construct corresponds to a different aspect of a firm’s performance.
Based on these two dimensions, the study could present a relative comparison between fintech firms and traditional banks.
The organizational culture assessment instrument (OCAI) was developed by Cameron and Quinn. It has been empirically supported and validated by authors via various methods, such as confirmatory factor analyses (Heritage et al., 2014). Over 10,000 companies have used this tool to assess their organizational culture.
The OCAI model consists of a tool in which the respondent distributes 100 points between four important “competing values,” known as the ipsative scale. These values correspond to the four organizational culture types. In this tool, adhocracy culture is the creative and entrepreneurial culture, clan culture is the collaborative culture, hierarchy culture is more control oriented, and market culture is based on competitive attributes (OCAI Online, 2019). In fact, the respondent performs two assessments in this model. First, they distribute the 100 points based on their perceptions of the current culture of their organization. Next, the respondent distributes 100 points according to how they would ideally like the culture to be in the future. Through this assessment structure, organizations can understand the aspects of their culture that need to change and in which direction (David et al., 2018)
The OCAI tool consists of six dimensions. These include dominant characteristics, organizational leadership, management of employees, organizational glue, strategic emphases and criteria of success. Each dimension has four alternatives that represent one of the organizational culture types (clan, adhocracy, market and hierarchy). Furthermore, this instrument has been found to contain strong psychometric properties and could be elaborated as a scientific tool to assess organizational culture (David et al., 2018). Many alternative models to evaluate organizational culture exist. However, due to the complexity in the concept of organizational culture, the study adopted the OCAI instrument, as it can provide empirical and
quantitative data regarding the type of organizational culture that most favorably impacts the relationship of BDAC and FPER.
4.5.4. Control Variables
For the purposes of the study, information regarding firm size, age of the firm and years of experience in BDAC were requested from the respondents. Prior studies in the BDA topic have used this information as control variables (Garmaki et al., 2016; Gunasekaran et al., 2017).
Similarly, these aspects were used as control variables in this study to reduce and avoid any biases in the sample.
The variable of firm size has been used in previous studies as a control variable (Garmaki et al., 2016; Gunasekaran et al., 2017). By considering firm size, the study could reduce any effects related to firm size. Moreover, the study could distinguish between traditional banks and fintech firms in the sample. A reliable and accurate comparison was important for the study’s objective.
Furthermore, firms that are larger in size may have antecedents that are different from smaller firms. Thus, a larger firm’s procedure to adopt BDAC may be more complex. Large organizations also tend to have higher revenues, more resources and more opportunities than smaller firms, as such organizations may have been established for a longer period in the industry.
Age of the Firm
As discussed in the literature, fintech firms are a new type of organization that has emerged with the development of digitalization and the benefits of the internet. They can be active worldwide without a physical presence. To keep the study’s sample reliable and accurate, respondents were asked to report the age of their firm to determine how old traditional banks are in comparison to fintech firms. In addition, this variable allowed the study to define
whether the age of the firm had an impact on the adoption of BDAC. For example, firms that have been in the industry for more than 50 years may maintain a traditional culture that does not favor innovation and the adoption of tech tools such as big data analytics. Previous studies have shown that the age of a firm impacts its performance both in the short-term and in the long-term (Galbreath & Galvin, 2008).
Years of Experience in BDA
By knowing the years of experience in BDA, the study may be able to understand its impact on why some firms generate or produce higher performance rates in comparison to firms that have the same age or same firm size. In addition, via years of experience in BDAC, the study may be able to understand if fintech firms recruit employees that are not only experienced and knowledgeable in the financial services sector but also highly skilled in handling big data.
The BDAC a firm has may influence its ability to coordinate organizational processes and its efficiency. Thus, years of experience may impact a firm’s ability to succeed.
For the purposes of this study, years of experience have been grouped into age range groups based on the existing literature. To analyze their impact, the age range groups have been transformed into dummy variables.
In this section, the data gathered via the questionnaire is assessed and analyzed to test the hypotheses of the study. To do so, the data is evaluated based on three important elements.
Next, the validity and reliability of the measurement model is assessed. Lastly, before testing the hypotheses, the structural model is evaluated regarding its validity.
5.1. Data Analysis
Out of the 123 responses received, 21 were excluded from the study because they were incomplete. This left 102 potential responses. To assess the engagement of the participants, a
minimum threshold of 0.2 was established regarding the variance in all the data per response.
As a result, 18 more responses were excluded from the study, leading to a final sample size of 84 responses. Characteristics of the sample can be seen in Table 2.
The next step was to analyze the skewness and kurtosis of the data. Upon examining all the indicators concerned with the independent, moderator and dependent variables, several variables were noted to have possible excess kurtosis or skewness. Trochim et al. (2016) have argued that kurtosis and skewness can acceptable if the values fall between -2 and 2. Based on this reasoning, D2, D3, T4, T5, T6, BR1, TS1, TS5 and OL2 have an excess kurtosis, indicating the possibility of a lack of variance. Moreover, variables CSE1 and CCS3, which are part of the moderator, also had an excess kurtosis above 2. Problems regarding these excesses were tested and assessed during the factor analysis.
5.2. Evaluation of the Measurement Model
Before testing the hypotheses, evaluation of the measurement model regarding its construct validity and reliability needed to be done. The purpose of testing a model’s construct validity is to illustrate the extent to which the variables that were to be measured have actually been measured (Fink, 2010, pp. 152–160). Reliability concerns the consistency of the measurements used in the survey, as it shows whether consistent responses yielded the same results.
This section is divided into two parts. The first part regards the reflective construct. In the second part, the assessment of the formative construct is discussed.
5.2.1. Reflective Construct Validity
To establish construct validity, the study emphasized both convergent and discriminant validity. Convergent validity allows for the assessment of the reflective indicators and whether they significantly lead on theoretical constructs. Discriminant validity tests the measurements that are not supposed to have any relation to prove that they are indeed unrelated.
Running the PLS-algorithm and a confirmatory factor analysis (Appendix D), the study was able to extract important information regarding its convergent validity. Cepeda-Carrion et al.
(2019) have indicated that outer loadings should have at least 0.7. The only reflective indicators that reported 0.7 and above were DD3, MS2, MS5, OL2 and TS4. However, all of the reflective indicators were shown to be significant (p<0.05). Because none of the outer loadings fell below 0.40, the assessment determined whether the rest of the indicators needed to be deleted by examining their impact on internal consistency reliability, for example, if removing an indicator increases the measure(s) above the threshold (Hair et al., 2017). After removing all the potential indicators, none were found to increase the measures. Therefore, the study retained all its reflective indicators.
Focusing on the average variance extracted (AVE), which indicates how much variation exists in the measures of the construct, the study’s findings showed that all AVE for the reflective constructs were below 0.5. Therefore, there was not much variation in the measures, and it may not confirm its discriminant validity.
To better assess the study’s discriminant validity, both Fornell and Larcker’s (1981) criterion and the heterotrait-monotrait ratio (HTMT) introduced by Henseler et al. (2015) were used. According to the findings, the square root of the AVE of each reflective construct was greater than the correlations with any other construct. Therefore, Fornell and Larcker’s (1981) criterion for discriminant validity was satisfied (Table 3). In contrast, the HTMT ratio that uses correlations between reflective constructs did not seem to satisfy the criterion for discriminant validity. Sufficient discriminant validity is illustrated when correlations across constructs are below 0.85 (Table 4).
Overall, considering the findings regarding the AVE and the findings of the HTMT ratio, one could argue that there is not sufficient validity, as one of the models did not support the measure for discriminant validity.