A new theoretical understanding of big data analytics capabilities in organizations: a thematic analysis

Big Data Analytics (BDA) usage in the industry has been increased markedly in recent years. As a data-driven tool to facilitate informed decision-making, the need for BDA capability in organizations is recognized, but few studies have communicated an understanding of BDA capabilities in a way that can enhance our theoretical knowledge of using BDA in the organizational domain. Big Data has been defined in various ways and, the past literature about the classification of BDA and its capabilities is explored in this research. We conducted a literature review using PRISMA methodology and integrated a thematic analysis using NVIVO12. By adopting five steps of the PRISMA framework—70 sample articles, we generate five themes, which are informed through organization development theory, and develop a novel empirical research model, which we submit for validity assessment. Our findings improve effectiveness and enhance the usage of BDA applications in various Organizations.


Study background
Many forms of BDA exist to meet specific decision-support demands of different organizations. Three BDA analytical classes exist: (1) descriptive, dealing with straightforward questions regarding what is or has happened and why-with 'opportunities and problems' using descriptive statistics such as historical insights; (2) predictive, dealing with questions such as what will or is likely to happen, by exploring data patterns with relatively complex statistics, simulation, and machine-learning algorithms (e.g., to identify trends in sales activities, or forecast customer behavior and purchasing patterns); and (3) prescriptive, dealing with questions regarding what should be happening and how to influence it, using complex descriptive and predictive analytics with mathematical optimization, simulation, and machine-learning algorithms (e.g., many large-scale companies have adopted prescriptive analytics to optimize production or solve schedule and inventory management issues) [18].
Regardless of the type of BDA analysis performed, its application significantly impacts tangible and intangible resources within an organization.

Previous studies on BDA
BDA tools or techniques are used to analyze Big Data (such as social media or substantial transactional data) to support strategic decision-making [19] in different domains (e.g., tourism, supply chain, healthcare), and numerous studies have developed and evaluated BDA solutions to improve organizational decision support. We categorize previous studies into two main groups based on non-technical aspects: those which relate to the development of new BDA requirements and functionalities in a specific problem domain and those which focus on more intrinsic aspects such as BDAC development or value-adding because of their impact on particular aspects of the business. Examples of reviews focusing on technical or problem-solving aspects are detailed in Table 1.
The second literature group examines BDA in an organizational context, such as improving firm performance using Big Data analytics in specific business domains [26]. Studies that support BDA lead to different aspects of organizational performance [20,24,25,[27][28][29] (Table 2). Another research on BDA to improve data utilization and decision-support qualities. For example, [30] explained how BDAC might be developed to improve managerial decision-making processes, and [4] conducted a thematic analysis of 15 firms to identify the factors related to the success of BDA capability development in SCM.

Potential applications of BDA
Many retail organizations use analytical approaches to gain commercial advantage and organizational success [31]. Modern organizations increasingly invest in BDA projects to reduce costs, make accurate decision making, and future business planning. For example, Amazon was the first online retailer and maintained its innovative BDA improvement and use [31]. Examples of successful stories of BDA use in business sectors include.
• Retail: business organizations using BDA for dynamic (surge) pricing [32] to adjust product or service prices based on demand and supply. For instance, Amazon uses dynamic pricing to surge prices by product demand. • Hospitality: Marriott hotels-the largest hospitality agent with a rapidly increasing number of hotels and serviced customers-uses BDA to improve sales [33]. • Entertainment: Netflix uses BDA to retain clientele and increase sales and profits [34,35]. • Transportation: Uber uses BDA [36] to capture Big Data from various consumers and identify the best routes to locations. 'Uber eats, ' despite competing with other delivery companies, delivers foods in the shortest possible time. • Foodservice: McDonald's continuously updates information with BDA, following a recent shift in food quality, now sells healthy food to consumers [37], and has adopted a dynamic menu [38]. • Finance: American Express has used BDA for a long time and was one of the first companies to understand the benefits of using BDA to improve business performance [39]. Big Data is collected on the ways consumers make on-and offline purchases, and predictions are made as to how they will shop in the future. • Manufacturing: General Electric manufactures and distributes products such as wind turbines, locomotives, airplane engines, and ship engines [40]. By dealing with a huge amount of data from electricity networks, meteorological information systems, geographical information systems, benefits can be brought to the existing power system, including improving customer service and social welfare in the era of big data. • Online business: music streaming websites are increasingly popular and continue to grow in size and scope because consumers want a customized streaming service [41]. Many streaming services (e.g., Apple Music, Spotify, Google Music) use various BDA applications to suggest new songs to consumers.

Organization value assessment with BDA
Specific performance measures must be established that rely on the number of organizational contextual factors such as the organization's goal, the external environment of the organization, and the organization itself. When looking at the above contexts regarding the use of BDA to strengthen process innovation skills, it is important to note that the approach required to achieve positive results depends on the different combinations along with the area in which BDA deployed [42].

Organizational development and BDA
To assist organization decision-making for growth, effective processes are required to perform operations such as continuous diagnosis, action planning, and the implementation and evaluation of BDA. Lewin's Organizational Development (OD) theory regards processes as having a goal to transfer knowledge and skills to an organization, with the process being mainly to improve problem-solving capacity and to manage future change.
Beckhard [43] defined OD as the internal dynamics of an organization, which involve a collection of individuals working as a group to improve organizational effectiveness, capability, work performance, and the ability to adjust culture, policies, practices, and procedure requirements. OD is 'a system-wide application and transfer of behavioral science knowledge to the planned development, improvement, and reinforcement of the strategies, structures, and processes that lead to organization effectiveness' [44], and has three concepts: organizational climate, culture, and capability [45]. Organizational climate is 'the mood or unique personality of an organization' [45] which includes shared perceptions of policies, practices, and procedures; climate features also consist of leadership, communication, participative management, and role clarity. Organizational culture involves shared basic assumptions, values, norms, behavioral patterns, and artifacts, defined by [46] as a pattern of shared basic assumptions that a group learned by solving problems of external adaptation and internal integration (p. 38). Organizational capacity (OC) implies the organization's function, such as the production of services or products or maintenance of organizational operations, and has four components: resource acquisition, organization structure, production subsystem, and accomplishment [47]. Organizational culture and climate affect an organization's capacity to operate adequately ( Fig. 1).

Research methodology
Our systematic literature review presents a research process for analyzing and examining research and gathering and evaluating it [48] In accordance with a PRISMA framework [49]. We use keywords to search for articles related to the BDA application, following a five-stage process.

Stage1: design development
We establish a research question to instruct the selection and search strategy and analysis and synthesis process, defining the aim, scope, and specific research goals following guidelines, procedures, and policies of the Cochrane Handbook for Systematic Reviews of Intervention [50]. The design review process is directed by the research question: what are the consistent definitions of BDA, unique attributes, objections, and business revolution, including improving the decision-making process and organization performance with BDA? The below table is created using the outcome of the search performed using Keywords-Organizational BDAC, Big Data, BDA (Table 3).

Stage 2: inclusion and elimination criteria
To maintain the nuances of a systematic review, we apply various inclusion and exclusion criteria to our search for research articles in four databases: Science Direct, Web of Science, IEEE (Institute of Electrical and Electronics Engineers), and Springer Link. Inclusion criteria include topics on 'Big Data in Organization' published between 2015 to 2021, in English. We use essential keywords to identify the most relevant articles, using truncation, wildcarding, and appropriate Boolean operators (Table 4).

Stage 3: literature sources and search approach
Research articles are excluded based on keywords and abstracts, after which 8062 are retained ( Table 5). The articles only selected keywords such as Big Data, BDA, BDAC, and the Abstract only focused on the Organizational domain.

Stage 4: assess the quality of full papers
At this stage, for each of the 161 research articles that remained after stage 3 presented in Table 6, which was assessed independently by authors in terms of several quality criteria such as credibility, to assess whether the articles were well presented, relevance which was assessed based on whether the articles were used in the organizational domain.

Stage 5: literature extraction and synthesis process
At this stage, only journal articles and conference papers are selected. Articles for which full texts were not open access were excluded, reducing our references to 70 papers 1 ( Table 7).

Meta-analysis of selected papers
Of the 70 papers satisfying our selection criteria, publication year and type (journal or conference paper) reveal an increasing trend in big data analytics over the last 6 years (Table 6). Additionally, journals produced more BDA papers than Conference proceedings (Fig. 2), which may be affected during 2020-2021 because of COVID, and fewer conference proceedings or publications were canceled.

Results
Thematic analysis is used to find the results which can identify, analyze and report patterns (themes) within data, and produce an insightful analysis to answer particular research questions [51]. The combination of NVIVO and Thematic analysis improves results. Judger [52] maintained that using computer-assisted data analysis coupled with manual checks improves findings' trustworthiness, credibility, and validity (p. 6).   [53] "A large volume of digital data found in government organizations which require different speed, i.e., velocity based on the requirements of the government application domains with a wide variety of data types and sources, and these government big data must be able to guarantee veracity to extract desired value for the target government organization" (p. 41) [54] "Big data in information technology is a set of approaches, tools, and methods for processing structured and unstructured data of huge volumes and considerable diversity for obtaining human-perceptible results" (p. 1) [55] "Volume: This is the most significant aspect that characterizes Big data-the huge amount/volume. Velocity: Velocity means two things here. The first one is the flow of data-a constant stream of data. The other one is the possibility of making use of real-time data. Variety: It indicates varying characteristics of the data (unstructured data, or data in different structures) as well as of sources that present these data" (p. 853) [56] "Big data are data sets so complex that cannot be managed or analyzed using traditional data analysis software. These data sets share 7 common characteristics, the 7Vs: V1-Volume, V2-Velocity, V3-Variety, V4-Veracity, V5-Value, V6-Variability & V7-Visualization" (p. 1094) [57] "Veracity, which refers to data accuracy that relates to quality. After which it became possible to develop more sophisticated data analysis software to fulfill the needs of handling the information explosion according to the way it is accessed, searched, processed and managed" (p. 364) Sabharwal and Miah Journal of Big Data (2021) 8:159 Defining big data Of 70 articles, 33 provide a clear replicable definition of Big Data, from which the five representative definitions are presented in Table 8.

Defining BDA
Of 70 sample articles, 21 clearly define BDA. The four representative definitions are presented in Table 9. Some definitions accentuate the tools and processes used to derive new insights from big data.

Defining Big Data analytics capability
Only 16% of articles focus on Big Data characteristics; one identifies challenges and issues with adopting and implementing the acquisition of Big Data in organizations [42]. The above study resulted that BDAC using the large volumes of data generated through different devices and people to increase efficiency and generate more profits. BDA capability and its potential value could be more than a business expects, which has been presented that the professional services, manufacturing, and retail have structural barriers and overcome these barriers with the use of Big Data [60]. We define BDAC as the combined ability to store, process, and analyze large amounts of data to provide meaningful information to users. Four dimensions of BDAC exist data integration, analytical, predictive, and data interpretation (Table 10). Table 9 BDA definitions Source Definition [58] "Big data analytics defined in six components-data generation, data acquisition, data storage, advanced data analytics, data visualization, and decision-making for value-creation), its typical tools, techniques and technologies, and its main domains of application" (p. 755) [56] "Big data has been used to describe datasets so complex that they cannot be managed or analyzed using traditional data analysis software" (p. 1095) [11] "Big data analytics enables large-scale data sets integration, supporting people management decisions, and cost-effectiveness evaluation of healthcare organizations" (p. 1) [59] "Big data analytics is defined as a process to analyze the large data volumes to capture value for the businesses and employees" (p. 229) Table 10 Big Data analytics capability definitions Source Definition [61] "The ability of an organization to collect and analyze data to generate insights, by effectively developing its data, technology, and talent through organization-wide processes, roles, and structures" (p. 2) [57] "The ability to acquire, store, process and analyze a large amount of data" (p. 13) [62] "An organization's ability to mobilize and deploy data analytics-related resources combined with marketing resources and capabilities" (p. 3) [16] "Big data capability is defined as firms need a combination of certain tangible, human and intangible resources to build BDA capability" (p. 1050) Sabharwal and Miah Journal of Big Data (2021) 8:159

Research theme Summary
BDA resource concinnity 12% of articles focus on implementing BDAC, but limited information is provided on orchestrating and handling such resources to build BDAC within an organization. The potential to strategically leverage these resources must be increased. Future research must identify how to build proficiency BD proficiencies 7% of articles focus on organization proficiency rather than Big Data proficiency; these two must be differentiated. An organization can implement and develop substantial BDA proficiency but may utilize it only in specific functions such as finance and marketing. Future research should focus on building both BDAC and organization proficiency Transforming BD into actionable insights To comprehend the value of BDAC, data must be transformed into actionable insights to improve organization decisions. 15% of research articles focused on transforming BDAC into action. Organizations may depend upon many factors which can be manipulated experimentally in future studies using BDAC and OD frameworks Management attitude towards BD and its impact 18% of studies revealed that an organization's management normally decided to implement Big Data but did not fully support the capability of the decision-making process. Management use prediction technique over BDA result which has been studied and found in only 12% articles presented management. Investment in BDA is affected by a lack of understanding of its capabilities by high-level management Organization value assessment with BDA Specific performance measures must be established that rely on the number of contextual factors and the area in which BDA is deployed. 7% of articles considered organizations were ready for BDA implementation, but organization structure must be supportive to enable the realization of BDA benefits

BDA themes
It is feasible to identify outstanding issues of research that are of excessive relevance, which has termed in five themes using NVIVO12 (Fig. 3). Table 11 illustrates four units that combine NVIVO with thematic analysis for analysis: Big data, BDA, BDAC, and BDA themes. We manually classify five BDA themes to ensure accuracy with appropriate perception in detail and provide suggestions on how future researchers might approach these problems using a research model. Manyika et al. [63] considered that BDA could assist an organization to improve its decision making, minimize risks, provide other valuable insights that would otherwise remain hidden, aid the creation of innovative business models, and improve performance.
The five themes presented in Table 11 identify limitations of existing literature, which are examined in our research model (Fig. 4) using four hypotheses. This theoretical model identifies organizational and individual levels as being influenced by organization climate, culture, and capacity. This model can assist in understanding how BDA can be used to improve organizational and individual performance.

The Research model development process
We analyze literature using a new research method, driven by the connection between BDAC and resource-based views, which included three resources: tangible (financial and physical), human skills (employees' knowledge and skills), and intangible (organizational culture and organizational learning) used in IS capacity literature [65][66][67][68]. Seven factors enable firms to create BDAC [16] (Fig. 5).
To develop a robust model, tangible, intangible, and human resource types should be implemented in an organization and contribute to the emergence of the decision-making process. This research model recognizes BDAC to enhance OD, Fig. 4 The framework of organizational development theory [64] strengthening organizational strategies and the relationship between BD resources and OD. Figure 6 depicts a theoretical framework illustrating how BDA resources influence innovation sustainability and OD, where Innovation sustainability helps identify market opportunities, predict customer needs, and analyze customer purchase decisions [69].
Miller [70] considered data a strategic business asset and recommended that businesses and academics collaborate to improve knowledge regarding BD skills and capability across an organization; [70] concluded that every profession, whether business or technology, will be impacted by big data and analytics. Gobble [71] proposed that an organization should develop new technologies to provide necessary supplements to enhance growth. Big Data represents a revolution in science and technology, and a data-rich smart city is the expected future that can be developed using Big Data [72]. Galbraith [73] reported how an organization attempting to develop BDAC might experience obstacles and opportunities. We found no literature that combined Big Data analytics capability and Organizational Development or discussed interaction between them.
Because little empirical evidence exists regarding the connection between OD and BDA or their characteristics and features, our model (Fig. 7) fills an important void, directly connecting BDAC and OD, and illustrates how it affects OD in the organizational concepts of capacity, culture, and climate, and their future resources. Because BDAC can assist OD through the implementation of new technologies [15,26,57], we hypothesize:

H1: A positive relationship exists between Organizational Development and BDAC.
OC relies heavily on OD, with OC representing a resource requiring development in an organization. Because OD can improve OC [44,45], we hypothesize that:

H2: A positive relationship exists between Organizational Development and Organizational
Capability.
With the implementation or adoption of BDAC, OC is impacted [46]. Big data enables an organization to improve inefficient practices, whether in marketing, retail, or media. We hypothesize that:

H3: A positive relationship exists between BDAC and Organizational Culture.
Because BDAC adoption can affect OC, the policies, practices, and measures associated with an organization's employee experience [74], and improve both the business climate and an individual's performance, we hypothesize that:

H4: A positive relationship exists between BDAC and Organizational Climate.
Our research is based on a need to develop a framework model in relation to OD theory because modern organizations cannot ignore BDA or its future learning and association with theoretical understanding. Therefore, we aim to demonstrate current trends in capabilities and a framework to improve understanding of BDAC for future research.

Discussion
Despite the hype that encompasses Big Data, the organizational development and structure through which it results in competitive gains have remained generally underexplored in empirical studies. It is feasible to distinguish the five prominent, highly relevant themes discussed in an earlier section by orchestrating a systematic literature review and recording what is known to date. By conducting those five thematic areas  Fig. 7, provide relation how they are impacting each other's performance and give some ideas on how researchers could approach these problems. The number of published papers on Big Data is increasing. Between 2015 and May 2021, the highest proportion of journal articles for any given year (21%) occurred until May 2021 with the inclusion or exclusion criteria such as the article selection only opted using four databases: Science Direct, Web of Science, IEEE (Institute of Electrical and Electronics Engineers), and Springer Link and included only those articles which titled as 'Big Data in Organization' published, in the English language. We use essential keywords to identify the most relevant articles, using truncation, wildcarding, and appropriate Boolean operators. While BDAC can improve business-related outcomes, including more effective marketing, new revenue opportunities, customer personalization, and improved operational efficiency, existing literature has focused on only one or two aspects of BDAC. Our research model (Fig. 7) represents the relationship between BDAC and OD to better understand their impacts on OC. We explain that the proposed model education will enhance knowledge of BDAC and that it may better meet organizational requirements, ensuring improved products and services to optimize consumer outcomes.
Considerable research has been conducted in many different contexts such as the health sector, education about Big Data, but according to past literature, BDAC in an organization is still an open issue, how to utilize BDAC within the organization for development purposes. The full potential of BDA and what it can offer must be leveraged to gain a commercial advantage. Therefore, we focus on summarizing by creating the themes using past relevant literature and propose a research model based on literature [61] for business.
While we explored Springer Link, IEEE, Science Direct, and Web of Science (which index high-impact journal and conference papers), the possibility exists that some relevant journals were missed. Our research is constrained by our selection criteria, including year, language (English), and peer-reviewed journal articles (we omitted reports, grey journals, and web articles).

Conclusion
A steadily expanding number of organizations has been endeavored to utilize Big Data and organizational analytics to analyze available data and assist with decision-making. For these organizations, influence the full potential that Big Data and organizational analytics can present to acquire competitive advantage. In any case, since Big Data and organizational analytics are generally considered as new innovative in business worldview, there is a little exploration on how to handle them and leverage them adequately. While past literature has shown the advantages of utilizing Big Data in various settings, there is an absence of theoretically determined research on the most proficient method to use these solutions to acquire competitive advantage. This research recognizes the need to explore BDA through a comprehensive approach. Therefore, we focus on summarizing with the proposed development related to BDA themes on which we still have a restricted observational arrangement.
To this end, this research proposes a new research model that relates earlier studies regarding BDAC in organizational culture. The research model provides a reference to the more extensive implementation of Big Data technologies in an organizational context. While the hypothesis present in the research model is on a significant level and can be deciphered as addition to theoretical lens, they are depicted in such a way that they can be adapted for organizational development. This research poses an original point of view on Big Data literature since, by far majority focuses on tools, infrastructure, technical aspects, and network analytics. The proposed framework contributes to Big Data and its capability in organizational development by covering the gap which has not addressed in past literature. This research model also can be viewed as a value-adding knowledge for managers and executives to learn how to drive channels of creating benefit in their organization through the use of Big Data, BDA, and BDAC.
We identify five themes to leverage BDA in an organization and gain a competitive advantage. We present a research model and four hypotheses to bridge gaps in research between BDA and OD. The purpose of this model and these hypotheses is to guide research to improve our understanding of how BDA implementation can affect an organization. The model goes for the next phase of our study, in which we will test the model for its validity.