Information Technology Managers’ Strategies for Implementing Data Governance

Information Technology Managers’ Strategies

for Implementing Data Governance

Section 1: Foundation of the Study

Background of the Problem

Data plays a crucial role in the information economy by providing the knowledge

needed for business decisions and scientific and technical processes. In the financial

sector, data-driven transformation is evident, highlighting the central and critical role of

data (Fryczak, 2020). Information technology (IT) managers in businesses and

organizations rely on data quality, including availability, reliability, accuracy, validity,

and security (Cichy & Rass, 2019). Meeting these requirements necessitates the

establishment of a data governance process, which ensures data availability, usability,

integrity, and security based on standards, policies, and control (Abraham et al., 2019).

The 21st century is characterized by the era of big data in which businesses aim

to harness and utilize vast amounts of data to drive decisions and establish strong

customer relationships (Fan et al., 2019). Artificial intelligence (AI) is playing an

increasingly significant role in organizing and leveraging big data, particularly in the

retail sector (Duan et al., 2019). In the realm of AI and big data, effective data

governance enhances the consistency and trustworthiness of data for big data algorithmic

systems (Janssen et al., 2020). Regulated industries such as health care require data

governance (DG) to safeguard patient records from misuse, especially with the

exponential growth of digital health care data worldwide. Tiffin et al. (2019) proposed a

DG framework to address this challenge, particularly in developing countries where data

security and privacy are in their infancy.

The discussed use cases clearly demonstrated that business leaders and IT

managers have a vested interest in the quality of business data. DG is essential not only

for making data-driven decisions but also for protecting data against misuse. The current

study aimed to gain an in-depth understanding of the strategies employed by IT

managers to implement DG. The findings may reveal the driving factors, challenges, and

opportunities associated with DG implementation.

Problem Statement

In the era of Industry 4.0, data plays a crucial role in enterprise software

architecture, functioning as a valuable business asset and a key component of data-as-

aservice. However, DG has often been informal, fragmented, and inadequate when solely

driven by the IT department (Al-Ruithe et al., 2019). Insufficient DG affects data quality

at the organizational level, which is a significant concern in the information economy

(Abraham et al., 2019). Studies indicated that nearly half of new data sets are error prone

due to the lack of stringent DG, emphasizing the need for organizations to prioritize data

quality improvement (Nagle et al., 2020).

The general IT problem was that many companies fail to fully integrate DG with

IT governance. This oversight can result in inconsistencies in data availability,

compromises in data integrity and security, and various costs such as data loss, legal

actions, and compliance penalties. The specific IT problem was the lack of strategies

among some IT managers to effectively implement DG.

Purpose Statement

This qualitative pragmatic study aimed to investigate the strategies employed by

IT managers in effectively implementing DG. By exploring the factors that facilitate or

impede the implementation of DG within organizations, the study sought to enhance

knowledge in this area. The targeted population consisted of IT managers in the United

States, specifically those affiliated with companies that have achieved successful

implementation of a comprehensive DG process. The potential positive outcomes of this

research include advancements in data quality and security for businesses and

individuals, as well as the promotion of accessible DG solutions through affordable

applications, leading to its democratization.

Nature of the Study

The study aimed to explore the strategies used by IT managers to implement DG.

Research methods include quantitative, qualitative, and mixed methods. Qualitative

research is inductive and often involves interviews to gain a deeper understanding of

cultural aspects or individuals’ perceptions (Small, 2021). On the other hand,

quantitative research is deductive and relies on numerical data to test hypotheses (Lima

& NewellMcLymont, 2021). Mixed-methods research (MMR) combines qualitative and

quantitative approaches, which was not suitable for the current study due to time and

resource constraints. MMR arose in the 1980s from the need to mitigate bias and

weaknesses inherent in quantitative and qualitative methods by using triangulation of

data sources to provide a more balanced view of the phenomenon (Timans et al., 2018).

MMR has gained momentum in the social scientific circles. MMR could be appropriate

for providing a quantitative follow-up to validate driving factors for the implementation

of DG by IT managers.

In the current study, the qualitative method was selected to explore the “what,”

“why,” and “how” of IT managers’ DG strategies. Qualitative methodology allowed for

a comprehensive exploration of the phenomenon and the generation of new knowledge.

Although MMR could have provided quantitative validation, it was not feasible for this

study’s scope.

The pragmatic qualitative design was deemed most suitable for this study because

it helped me gain an in-depth understanding of IT managers’ DG strategies. The case

study design, under the pragmatic approach, enables detailed examination through data

collection methods such as observations, interviews, and secondary data analysis. The

case study design allows for studying a bounded case or multiple cases without

interference (Takahashi & Araujo, 2020). Ethnographic and phenomenological designs

were not chosen because they did not align with the focus on IT managers’ strategies or

data engineering topics. The strength of the ethnographic design is in understanding the

shared patterns of a cultural group (Gherardi, 2019). This was not the focus of the current

study. The phenomenological design was not appropriate because it focuses on the life

experiences of individuals (Greening, 2019). Therefore, this design was not suitable for a

data engineering topic such as DG.

To answer the research questions, I conducted a qualitative study involving

interviewing IT managers. This approach provided insights into the strategies used for

implementing DG. A pragmatic design was justified because it facilitated individual

interviews with IT managers from various institutions without requiring partner

organizations to be approved by the institutional review board (IRB).

Recruiting IT managers for individual interviews and collecting qualitative data

were essential for the planned research design. Interview questions and protocols were

developed to align with the problem statement and purpose of the study. The data

collected included IT managers’ responses regarding their strategies for successful DG

implementation.

Interviewing IT managers at institutions with DG programs may have required

approval from both Walden University’s IRB and the company’s IRB. However, Walden

University offers a pragmatic approach in which doctoral students can interview selected

participants with IRB approval without necessitating partner organization approval. The

main challenge was finding organizations that had successfully implemented DG

programs and IT managers willing to share their insights.

Research Question

The central research question was the following: What strategies do IT managers

use for implementing DG programs?

Interview Questions

Interview questions are among the typical data collection mechanisms for a

qualitative pragmatic study. The questions must be designed to collect the data needed to

answer the main research questions and must be in strict alignment with the conceptual

framework of the study. For the current study, the research question addressed the

strategies that IT managers use for implementing successful DG programs. By exploring

strategies that IT managers use to drive successful DG programs, this study aimed to

provide insights others can use to implement similar programs. The conceptual

frameworks guiding this study were, Abraham et al.’s (2019) DG conceptual framework

and Khatri and Brown’s (2010) unified framework for DG. Table 1 shows an alignment

of the two frameworks by grouping elements that belong together; this helped me

simplify the formulation of interview questions to cover the two frameworks at the same

time.

Table 1

Alignment of Abraham et al.. (2019) Conceptual Framework for Data Governance and

Khatri and Brown (2010) Data Governance Framework

Abraham et al.

(2019) dimension

Description Khatri and Brown

(2010) dimension

Antecedents

The contingency factors, which

impact the adoption and

implementation of data governance;

internal and external antecedents.

Data principles

Data scope Pertains to the data asset an

organization needs to

govern;traditional data and big data

Cloud vs

noncloud

Organizational

scope

Determines the organizational

expansiveness of data governance and

roughly corresponds to the unit of

analysis; intra-organizational and the

inter-organizational scope.

Domain scope Data decision domains, to which

governance mechanisms are applied:

data quality, data security, data

architecture, data lifecycle, meta data,

data storage and infrastructure

Data quality

Data access

Metadata

Governance

mechanisms

Core dimension of the framework and

encompass structural, procedural, and

relational mechanisms

Data lifecycle

Data access

Consequences The effects of data governance;

intermediate performance effects and

risk management

Data quality Data

access

This study had three categories of interview questions. First were

frameworkrelated questions that addressed why IT managers chose to implement DG

programs at the select participant organizations. Elements of these frameworks were

used to formulate interview questions for capturing insights into the principles,

antecedents, and life cycle of DG. The second category of interview questions addressed

the critical success factors (CSFs) that contributed to successful DG programs at the

selected institutions. Alhassan et al.’s (2019) CSF matrix was used for this purpose. The

third category of interview questions addressed the challenges that lead to DG

implementation failures, including but not limited to lack of executive support, costs,

data quality, complexity of business cases, IT focus, lack of clearly defined roles and

responsibilities, and technical skills gap (AlRuithe et al., 2019; Alsousi & Shah, 2022;

Janssen et al,, 2020; Rasheed et al., 2022). It was important for the current study to

capture this third category for lessons learned that may inform IT managers wanting to

implement DG programs.

Framework-Related Questions

1. What data principles (i.e., contingent factors, for example, data quality,

regulatory compliance) caused your organization to decide to implement a

data governance program?

2. What is the nature of data assets (traditional or big data) that are the main

target for data governance?

3. A data governance program is often primarily concerned with internal

organization data assets. Can you please comment on any third-party data

assets that your organization uses and needs governed?

4. Data quality, data access, or metadata management capability are often needs

or challenges for many organizations like yours. Can you please comment on

data quality, data access, or metadata management capability to your

organization and their classification as needs or challenges?

5. Some organizations establish a data life cycle process prior to implementing

the data governance program. What was the case for your organization?

6. A company culture that places data as a corporate asset often facilitates the

adoption and acceptance of data governance. Is data treated as an asset within

your organization’s culture? Can you please comment on the culture and

attitudes toward data governance within your organization?

7. Does your organization treat data governance as a company-wide initiative

backed up by the executive leadership, or an effort driven solely by certain

departments?

CSF-Related Questions

1. Prior to starting the data governance programs, did your organization have in

place a clear data strategy, including data management processes, procedures,

and policies? If so, what were they? Are they organization wide or

departmental?

2. Can you please summarize the various data management activities within

your organization?

3. What individuals are responsible for the data-related activities in your

organization? Who defines data policies and processes? What are data roles

in your organization?

4. What specific tools, applications, or technologies does your organization use

to manage data assets?

Challenges

1. What is the nature of challenges that your organization faced during the

implementation of the data governance program?

2. How did the data governance help (or not) address those challenges?

3. What strategies does your organization understake to address those

challenges?

4. As an organization, if you had to do it again, what would you do differently

and where would you start?

5. Any advice you would give to other companies wanting to build a data

governance program?

Conceptual Framework

DG plays a crucial role in optimizing the management of data within an

organization. DG is one of the seven complementary areas of the data management

framework, which also includes data architecture, metadata, data quality, data life cycle,

analytics, and data privacy. During the 1980s, a significant shift occurred in the field of

IT, moving from sequential storage methods such as punch cards and magnetic tapes to

random access storage through hard disks. This transformation revolutionized computing

and gave rise to the concept of data management (DM; Dagnaw & Tsigie, 2019). DM

encompasses an administrative process that spans the entire life cycle of data, including

its creation, ingestion, processing, storage, utilization, and disposal. The objective of DM

is to ensure the accessibility, reliability, and timely delivery of data to users. By

implementing effective DM practices, organizations can minimize risks and costs

associated with data loss, regulatory noncompliance, misuse, and data breaches.

However, within the broader framework of DM, DG focuses on establishing and

enforcing policies, standards, and procedures related to DM. DG encompasses the rules,

roles, responsibilities, and processes that guide the proper handling, protection, and

utilization of data assets. DG aims to address various aspects of data management

(Abraham et al. 2019):

•Data ownership: Clearly defining who is accountable for data assets within

the organization, ensuring that the right people have the authority and

responsibility to make decisions about data.

•Data dtewardship: Assigning data stewards who are responsible for managing

and maintaining data quality, integrity, and security throughout its life cycle.

•Data policies and standards: Developing and enforcing policies and standards

that govern data management practices, ensuring consistency and adherence

to regulatory requirements.

•Data quality management: Implementing processes and controls to monitor,

assess, and improve the quality of data, ensuring its accuracy, completeness,

consistency, and relevance.

•Data security and privacy: Establishing measures to protect sensitive data

from unauthorized access, ensuring compliance with relevant data protection

regulations and safeguarding individual privacy.

•Data compliance and risk management: Ensuring adherence to regulatory and

legal requirements related to data management and mitigating risks associated

with data breaches, noncompliance, and other potential issues.

•Data integration and interoperability: Facilitating the seamless integration and

sharing of data across different systems, applications, and departments within

the organization.

By optimizing DG practices, organizations can foster a data-driven culture,

enhance decision-making processes, improve operational efficiency, and gain a

competitive edge in the rapidly evolving digital landscape. The current study included

two conceptual frameworks: Abraham, et al.’s (2019) DG conceptual framework and

Khatri and Brown’s (2010) unified framework for DG. The DG conceptual framework

Abraham et al. offers a comprehensive approach to comprehend the implementation of

DG, including challenges and successes, through six dimensions of analysis: governance

mechanisms, organizational scope, data scope, domain scope, antecedents, and

consequences of data governance. The logical connection between this framework and

the nature of my study was using the six criteria domains to gain insights into the

motivations and strategies of IT managers when implementing DG programs.

Khatri and Brown’s (2010) unified framework for DG provides a comprehensive

guide for researchers and practitioners seeking to develop effective data governance

programs. This framework encompasses five key decision domains that serve as the

foundation for successful implementation:

•Data principles: This domain focuses on aligning data governance with the

business needs and establishing the expected behaviors of information

systems professionals and users. By defining clear principles, organizations

can ensure that DG efforts are in line with strategic objectives.

•Data quality: Within this domain, the emphasis is assessing the value of data

assets based on dimensions such as accuracy, timeliness, completeness, and

credibility. Understanding and maintaining data quality is crucial for reliable

decision making and efficient operations.

•Metadata: The metadata domain involves capturing and managing data

descriptions. Metadata provides essential context and understanding of data

elements, facilitating effective DG practices and enabling efficient data

discovery and utilization.

•Data access: This domain deals with defining levels of data access, such as

confidential or restricted access, to meet internal and regulatory requirements.

Establishing appropriate access controls ensures data security and compliance

with relevant policies and regulations.

•Data life cycle: The final decision domain focuses on managing the entire life

cycle of data, including its creation, storage, use, archiving, and disposal. By

establishing comprehensive data life cycle processes, organizations can

ensure data integrity, accessibility, and compliance throughout its lifespan.

To investigate strategies employed by IT managers for implementing DG

programs, I employed these five decision domains as a basis for designing interview

questions. By exploring these domains in-depth, I was able to provide insights into the

approaches, challenges, and successes of DG initiatives within different organizational

contexts. The Abraham et al. (2019) DG conceptual framework (see Figure 1) and Khatri

and Brown’s (2010) unified framework for DG (see Figure 2) share common elements,

particularly in terms of data quality and data principles. Table 1 displays the mapping

between Abraham et al.’s framework and Khatri and Brown’s framework. A single

framework, such as Abraham et al.’s (2019), would have sufficed. However, I decided to

incorporate the framework by Khatri and Brown as well because it offered a more

detailed analysis compared to the former. The framework presented by Abraham et al.

focuses on the cause (antecedents), process, and effect (consequences) of a DG program,

while Khatri and Brown’s framework emphasize attributes.

Figure 1

Khatri and Brown (2010) Unified Framework for Data Governance

Figure 2

Abraham, R., Schneider, J., & Vom Brocke, J. (2019) Conceptual Framework Data

Governance

Furthermore, Alhassan et al.’s (2019) CSFs (see Table 2) provided additional

perspectives to gain deeper insights into the effective implementation of data governance

by IT managers. Theories, frameworks, primitives, constructs, and design patterns are

valuable guides for design and implementation. However, the industry’s practical and

tested knowledge, in the form of best practices, provided the most valuable insights.

These best practices are instrumental in assessing the success or failure of an

implementation. Similarly, CSFs played a vital role in this evaluation. To gauge IT

managers’ adherence to industry best practices when implementing DG programs, I

employed CSFs as a metric to measure their response.

Table 2

Appplication of Abraham et al. (2019) Conceptual Framework for Data Governance and

Khatri and Brown (2010) Data Governance Framework to Research

Reference Title Khatri and Brown

(2010)

Abraham et al.

(2019)

Abraham et al. (2019) Data governance: A conceptual

framework, structured review, and research

agenda

Al-Badi et al. (2018) Exploring big data governance frameworks x

Alhassan et al. (2017) Data governance activities: A comparison

between scientific and practice-oriented

literature

Alhassan et al. (2019) Critical success factors for data governance:

A theory building approach

Alsousi & Shah (2022) Data governance for SME: Systematic

literature review

Bento et al. (2022) How data governance frameworks can

leverage data-driven decision making

D’Hauwers et al. (2022) Business models value and control in data

ecosystems x

Derakhshannia et al. (2020) Data lake governance: Towards a systemic

and natural ecosystem analogy x

Fernandes (2022) Trajectories and challenges for data

governance implementation in the public

sector: A systematic review

Janssen et al. (2020) Data governance: Organizing data for

trustworthy artificial intelligence

x x

Jimenez et al. (2019) Oververview of data governance in business

contexts

Juddoo et al (2018) Data governance in the health

industry:Investigating data quality

dimensions within a big data context

Karkošková (2022) Data governance model to enhance data

quality in financial institutions

x x

Machado et al. (2022) Sustainable data governance: A systematic

review and a conceptual framework x

Mao et al. (2022) Government data governance framework

based on a data middle platform

x x

Nadal et al. (2022) Operationalizing and automating data

governance

Nielsen (2017) A comprehensive review of data

governance literature

Vilminko-Heikkinen & Pekkola

(2019)

Changes in roles, responsibilities and

ownership in organizing master data

management

Vojvodic & Hitz (2022) Relation of data governance, customer

centricity and data processing compliance

Walsh et al. (2022) Grounding data governance motivations: A

review of the literature

Yebenes & Zorilla (2019) Towards a data governance framework for

third generation platforms

Definition of Terms

Antecedents of data governance: Internal (e.g. organization strategy, IT strategy,

corporate decision-making authority, IT infrastructure) or external (e.g. legal and regulatory)

factors that influence the adoption and implementation of data governance (Abraham et al.,

2019; Schlackl et al., 2022).

Big data: Data sets with the following characteristics (volume, variety, velocity,

veracity, value): very large volume, complexity, variety to include structured and

nonstructured data, streaming, flat files; this mix of files is usually stored in a data lake

cloud storage (Janssen et al., 2020; Ramachandran et al., 2022; Yebenes & Zorrilla,

2019).

Big data governance: DG as applied to big data.

Critical success factors (CSFs): Factors that impact the success of DG

implementation; ranked in order of importance they are (a) employee data competencies,

(b) clear data process and procedures, (c) flexible data tools and technologies, (d)

standardized easy to follow data policies, (e) established data roles and responsibilities,

(f) clear inclusive data requirements, and (g) focused and tangible data strategies

(Alhassan et al., 2019).

Data governance (DG): Policies and procedures to maintain authority and control

over the management of an organization’s data assets for the purpose of increasing the

value of data and minimizing data-related cost and risk (Abraham et al., 2019;

Hikmawati et al., 2021). The four pillars of data governance are data quality, data

stewardship, data protection and compliance, and data management.

Data life cycle: Processes for master data management, data production, storage,

retention, retirement, traceability (Khatri & Brown, 2010).

Data quality (DQ): Value of data assets along dimensions such as accuracy,

timeliness, completeness, credibility, reliability, relevancy (Khatri & Brown, 2010;

Hikmawati et al., 2021).

Data principles: DG objectives intended to fulfill business needs including

definining behaviors of information professionals and users; common data principles are

value add, compliance, decision rights, future proofing, performance, process (Khatri &

Brown, 2010).

Decision domains: Five core DG areas of the Khatri and Brown (2010) unified

framework for DG: (a) data principles (to link data governance to the business needs, and

define behaviors of IS professionals and users), (b) data quality (value of data assets

along dimensions such as accuracy, timeliness, completeness, credibility), (c) metadata

(data descriptions), (d) data access (levels, e.g. confidential, restricted for internal and

regulatory purposes), and (e) data life cycle.

Information governance: A decision and accountability framework that dictates

allowed behaviors for the information life cycle (creation, valuation, use, sharing,

storage, archiving, and deletion); information governance includes the policies,

standards, processes, metrics, and roles governing the efficient and effective use of

information in alignment with the organization’s strategy and objectives (Al-Ruithe et

al., 2018; Mikalef et al., 2020).

Master Data Management (MDM): Core mission is to provide method and

processes for maintaining, integrating, and harmonizing master data to ensure consistent

system information across the organization; MDM controls master data by keeping it

consistent, accurate, current, relevant, and contextual to serve business needs

companywide; master data (customer, supplier, and material data, emloyee data) is the

version of the truth and reference data across the company’s business functions (sales,

services, order management, purchasing, manufacturing, billing, accounts receivable,

and accounts payable) to be used by business intelligence and applications (Haneem et

al., 2019;

Hikmawati et al., 2021).

Assumptions, Limitations, and Delimitations

Assumptions

As immutable requirements, assumptions play a crucial role in any academic

research, and the current study was no exception. As the researcher, I needed to state

these assumptions to ensure they were understood. Misguided assumptions can have a

detrimental impact on the analysis of findings and the results of a study. One assumption

was that IT managers interviewed were knowledgeable in DG, had implemented a DG

program, and were able to speak to the implementation of such. Additionally, I assumeed

a working DG program in lieu of a successful DG program. Success means different

things depending on audience and criteria. To declare a DG program successful implies a

prior evaluation. Another assumption was that organization documentation about DG

processes was up to date with the technology and industry.

Limitations

Due to purposeful sampling and a smaller sample size, qualitative research often

faces challenges in terms of reliability and validity, which are crucial for ensuring the

quality of the study (Malakar, 2022; Small, 2021). To address these challenges,

researchers often employ triangulation to enhance the reliability and validity of

qualitative results. However, qualitative studies also have inherent limitations, such as

difficulties in accessing secondary data and the potential for inadequately designed

questionnaires, that may not capture sufficient information to address the research

question. To mitigate these challenges, I used triangulation techniques and designed

interview questions rigorously. Unlike quantitative studies, in which the sample size can

be large, qualitative studies, including case studies, generally recommend a sample size

of 10 to 20 participants, at which point data saturation is typically observed (Malakar,

2022). Considering budget constraints, time limitations, and the availability of IT

managers willing to participate in interviews, I strived to limit my sample size as much

as possible, ideally targeting 12 participants.

This study was based on two conceptual frameworks, Abraham et al.’s (2019)

DG conceptual framework and Khatri and Brown’s (2010) unified framework for DG,

which have been widely referenced in the literature on DG. However, these frameworks

have not been validated through practical implementation, which presented an inherit

weaknesses associated with them. Abraham et al. acknowledged limitations of their

framework, including differences in applicability based on data scope (e.g., traditional

data vs. big data) and the lack of validation against real-world use cases by practitioners.

Delimitations

The current study aimed to address specific research questions. Therefore, it was

essential for me to determine and communicate the scope of the study. Delimitations

play a crucial role in defining the scope by establishing boundaries that determine what

is included and excluded from the study (Malakar, 2022). In the current study, there were

three main delimitations: program delimitation, nature of the study delimitation, and

framework delimitation.

First, the focus was on applying existing research to bridge gaps in knowledge

rather than creating new theories, inventions, or technical solutions. The study did not

propose a novel DG solution, architecture, theory, or framework. Instead, it sought to

understand the strategies employed by IT managers in implementing DG, with the aim of

sharing lessons learned and best practices with other practitioners undertaking similar

programs.

The nature of the study aligned with the characteristics of a pragmatic study.

Qualitative research designs often employ purposeful sampling, such as criterion

sampling, to identify information-rich cases for the case study (Takahashi & Araujo,

2020). For the current study, the participant population was limited to IT managers in

select private sector companies in the United States, with specific selection criteria

outlined in the Research Method and Design section and aligned with the conceptual

frameworks used.

This study included two conceptual frameworks to understand the

implementation of DG: Abraham et al.’s (2019) DG conceptual framework and Khatri

and Brown’s (2010) unified framework for DG. Within Abraham et al.’s framework, the

data scope dimension differentiated between traditional data and big data. Traditional

data is stored and managed on a centralized architecture such as a database or

datawarehouse; big data, on the other hand, is used on distributed architecture and

storage such as a data lake. Although the literature review included big data to ensure

completeness and relevance to the current data economy, the current study’s focus, data

collection, data analysis, and conclusions were limited to traditional data rather than big

data. This distinction was made due to the increased interest in big data management and

governance, which presented a significant research question of its own. By clearly

establishing these delimitations, I could narrow the study’s focus, provide valuable

insights within its defined scope, and contribute to the understanding and application of

DG strategies for IT managers in the private sector.

Significance of the Study

Contribution to IT Practice

The primary objective of IT is to facilitate business functions, thereby leading to

the achievement of an organization’s strategic objectives and profitability. DG plays a

crucial role in maintaining authority and control over an organization’s data assets,

aiming to increase their value while minimizing costs and risks (Abraham et al., 2019;

Hikmawati et al., 2021). The success or failure of DG impacts IT practices.

The findings of the current study have both theoretical and practical implications

that may benefit companies and data governance practitioners. First, the study may

provide valuable insights into the factors and challenges associated with implementing

DG, thereby enabling organizations to make informed decisions. Second, the study may

highlight emerging trends in the DG field, including aspects such as data quality and

trustworthiness, which are essential for enabling advanced analytics and AI in the

context of Industry 4.0 (see Yebenes & Zorrilla, 2019).

Successful DG is a comprehensive initiative that involves the collaboration of

multiple stakeholders encompassing various roles and business units within the

company. However, many organizations tend to conduct DG in isolated silos, often

driven solely by IT (Khatri & Brown, 2010). By recognizing and addressing this issue,

companies can foster synergy between strategy and technology, thereby enhancing their

DG practices.

The positive outcomes of effective DG implementation include improved data

quality, enhanced data security, increased trustworthiness, and overall value. These

attributes amplify the value of data assets, promoting innovation and generating revenue

for the enterprise. Additionally, robust DG helps mitigate risks such as mishandling

sensitive customer data, data breaches, and noncompliance with regulations, which can

cost companies millions of dollars annually (Schlackl et al., 2022). By maximizing the

value and minimizing the risks associated with data assets through robust data

governance practices, companies can achieve a healthy bottom line. This translates to

their ability to sustain jobs by retaining current employees, hiring new talent, and

avoiding significant layoffs during adverse economic situations.

Implications for Social Change

Technological innovation should prioritize the best interests of society, fostering

positive social change. This is achieved through improved data quality, security, and privacy,

as well as the democratization of DG. Effective DG enhances the quality, trustworthiness and

exchange of data (Hikmawati et al., 2021), enabling companies to leverage advanced

analytics and AI to unlock value from big data. Consequently, job creation, technological

development, and economic growth benefit society as a whole. However, the implementation

of DG is often costly and prohibitive for smaller companies, impeding its democratization.

To address this issue, the commoditization and self-serve nature of DG would significantly

reduce costs, enabling smaller companies and individuals to access higher data quality and

security. Currently, DG applications incur significant expenses, ranging from tens to

hundreds of thousands of dollars annually. Such costs are often unaffordable for small and

medium-size companies. Future trends indicate that independent software vendors such as

Tableau and Microsoft plan to integrate DG capabilities into their platforms and applications,

providing businesses with accessible solutions. This becomes even more crucial in the era of

big data, in which DG remains a considerable challenge. The availability of self-serve DG

will cater to the needs of the masses, extending beyond companies alone.

A Review of the Professional and Academic Literature

To conduct a study effectively, researchers should devise a strategy for selecting

academic literature that aligns with the topic and research questions and draws from the

existing body of knowledge. In the current study, the literature review aimed to equip me

and the reader with a theoretical foundation on the DG topic (definition, conceptual

frameworks, challenges, best practices, research topics). Such a premise was critical for

gaining insights into the strategies employed by IT managers to implementing DG

programs, which was the focus of this study. The structure of the literature review must

align not only with the research topic but also with the conceptual frameworks identified

and main themes from the literature.

This literature review focuses on DG and the chosen conceptual frameworks,

aiming to align with the topic. The review follows a structured approach based on the

themes derived from two conceptual frameworks (see Abraham et al., 2019; Khatri &

Brown, 2010). The literature review is structured into three sections. In the initial

section, I explore the use of conceptual frameworks in previous studies and research

contexts. The subsequent section centers on the application of these frameworks in the

context of the current study. The third section addresses alternative DG frameworks,

identifies existing gaps, and outlines potential future research avenues in the field of DG.

The review gives an overview of big data governance and identifies potential future

research topics in DG. The comprehensive understanding of challenges, success factors,

antecedents, and governance mechanisms gained from this review aided me in

formulating interview questions that elicited key insights regarding strategies and

motivating factors that drive companies to invest in DG programs.

The relevance of big data governance in this study was justified for several

reasons. First, it was crucial to ensure the comprehensiveness of the study by not focusing

solely on traditional DG. Second, big data has become pervasive in cloud computing and

Industry 4.0, making it imperative to discuss. Failing to address big data governance

would have been a significant omission in this study. Moreover, as more companies

migrate their IT and data infrastructures to the cloud, the subject of big data and its

governance becomes increasingly relevant. When interviewing IT managers regarding

their DG strategies, I needed to be aware that some have already experienced or were

planning to undergo a complete cloud migration for their companies. Therefore, big data

governance could serve as a potential topic for future research.

A methodological search strategy is fundamental to a good literature review. For

this literature review, I gathered 100 peer-reviewed papers published within the last 5

years (2018 onward) from reliable sources such as Google Scholar, Walden University

Library, and UC Santa Barbara Ulrich Periodicals Directory. To assemble a

comprehensive list of relevant articles, I initially used keywords such as “data

governance,” “big data governance,” “data,” and “governance.” I then filtered the list to

exclude articles unrelated to DG. I narrowed the results further by using the keyword

“framework” to identify frameworks suitable for my study. Two frameworks surfaced

that were comprehensive enough to be used as a foundation for my study: Abraham et

al.’s (2019) conceptual framework for DG and Khatri and Brown’s (2010) DG

framework. Using the dimensions of DG from these frameworks, I further refined my

search by employing a combination query of “data governance” AND keywords such as

“antecedents,” “data principles,” “data scope,” “domain scope,” “governance

mechanisms,” “data access,” “data lifecycle,” and “metadata.” By following this

systematic approach, I expected to provide a solid foundation for the study. Furthermore,

the literature review enabled me to develop a research question related to the core

definition for DG and the existing body of knowledge on the topic.

Application of Khatri and Brown’s (2010) and Abraham et al.’s (2019) DG

Frameworks

The aim of this section is to explore the practical applications of two key data

governance frameworks, the Abraham et al. (2019) conceptual framework for DG and

the Khatri and Brown (2010) DG framework, within the broader academic literature. To

achieve this, I conducted a thorough examination of references to these frameworks

within the literature review, as presented in Table 3. This analysis involved assessing

how these frameworks have been employed in various studies, identifying their

applications, pinpointing any gaps or limitations in their use, and drawing relevant

conclusions. Khatri and Brown (2010) played a foundational role in establishing a

reference point for DG frameworks, serving as a precursor tp subsequent research in this

domain. The current study highlighted a significant gap in the field of DG: the absence

of a guiding framework for practitioners and researchers. To address this issue, I

included five key decision domains that are essential for comprehending DG: data

principles, data quality, data access, data lifecycle, and metadata.

Table 3

Unified Critical Success Factors for Data Governance (Alhassan et al., 2019; Bento et

al., 2022; Yatya et al., 2021; Daneshmandnia, 2019)

Author Category

Employee data

competencies

Data governance activities that involve human

action. Directly impact the defining, implementing,

and monitoring of data processes and procedures,

as well as data policies and data requirements.

[1],[2] skills

Clear data processes

and procedures

Data policies in place which should be detailed

and operationalized during activities related to data

processes and procedures.

[1],[2] management

Flexible data tools and

technologies

All the activities related to software and hardware

that affect the data in an organization, including

the presentation and storage of data.

[1],[2] technology

Standardized easy-

tofollow data policies

Data policies are short statements that provide the

high-level guidelines and rules necessary for

dealing with data. These relate to data principles.

[1],[2],[4] management

Established data roles and

responsibilities

Individual(s) responsible for the data-related

activities in the organization, who defines the

policies and processes for the data as well as

assigning the duties for the actions related todata.

[1],[2],[5] management

Clear inclusive data

requirements

Define all aspects of data implementation, suchas

data flows and integration.

[1],[2] management

Focused and tangible data

strategies

Include planning fordata governance in order to

achieve its goals, as well as the main activities

related to considering data as assets.

[1],[2] management

Phased implementation

Start with small steps [5] management

Support from senior

management

Get buy-in and support from top leadership;

stakeholder support.

[2],[3],[5] organization

DG implementation and

maturity assessment

Measure goals with metrics [4] management

Communication Prioritize early communication, communicate

vision,strategy, and trust across the organization to

socialize the value of DG, and a culture that

embraces DG and its benefits. Strategic

communications.

[2], [4],[3],[5] management

Data quality Maintain high standards of data quality [4] technology

Team work Collaboration [4] organization

Critical success factor Detail

[1] Alhassan et al. (2019)

[2] Bento et al. (2022)

[3] Daneshmandnia (2019)

[4] Ibrahim et al. (2021)

[5] Yatya et al. (2021)

In building on the work of Khatri and Brown (2010), Abraham et al. (2019)

expanded the existing conceptual framework. This expansion included dimensions that

enhance the understanding of DG motivations and programs. Notably, these dimensions

align with the five decision domains proposed by Khatri and Brown (2010). To provide

practical context, the following sections offer a concise overview of how this study’s

conceptual frameworks have been applied to selected references from Table 3.

Motivations for DG Programmes

This study sought to gain insights into the strategies employed by IT managers

when implementing DG within different organizations in the United States. I aimed to

explore the motivations driving these strategies and the challenges encountered during

the process. This aligned with the broader scope of research in the field, which has

focused on unraveling the underlying motivations behind the adoption of DG practices.

Traditionally, the four pillars of DG have been data quality, data stewardship, data

protection and compliance, and data management (see Figure 3). In a related study

conducted by Walsh et al. (2022), the primary objective was to understand the rationale

behind companies initiating DG programs. Walsh et al. employed a structured

framework developed by Khatri and Brown (2010), encompassing five key decision

domains: data principles, data quality, data access, metadata, and data lifecycle. The

findings were organized within these domains. Notably, the conclusions drawn from the

analysis of data principles, data access, and data quality domains pointed toward

motivations rooted in operational and technological aspects. However, the study included

limitations in revealing motivations within the metadata and data life cycle domains.

This gap in understanding underscored the need for further research in these areas.

Figure 3

Four Pillars of Data Governance: Data Quality, Data Stewardship, Data Management,

Data Protection and Compliance

One significant contribution of the current study was its recognition of the Khatri

and Brown (2010) framework, which serves as a model and a template for constructing

and evaluating DG programs. The framework’s comprehensive attributes within each

domain offer valuable insights for researchers and practitioners. The framework’s

attributes provide a deeper understanding of the motivations driving DG initiatives, as

D Quality

D Protection

well as the organizational and technical factors influencing them, while addressing the

associated challenges. This framework emerged as a valuable tool for advancing the field

of DG.

Sustainability of DG Programs

DG’s primary goal is to enhance the value of data assets while minimizing

organizational risks. To achieve this, companies must ensure that such initiatives are

sustainable and ongoing. Bento et al. (2022) emphasized the significance of DG in

governing data assets, deriving insights, defining roles and accountabilities, and

formulating strategies. Although their research extended the foundations laid by Khatri

and Brown (2010) and DG frameworks proposed by Abraham et al. (2019) and others,

Bento et al. revealed the deficiency in addressing the sustainability aspect necessary for

continual program review, improvement, and value assessment within organizations.

Bento et al. proposed a framework of 12 CSFs for evaluating the sustainability of a DG

program. These factors include employee data competencies, well-defined data processes

and procedures, adaptable data tools and technologies, dtandardized user-friendly data

policies, established data roles and responsibilities, clear and inclusive data requirements,

focused and tangible data strategies, securing stakeholder buy-in, effective and strategic

communications, assessment of data governance status, and definition of sustaining

requirements. In conjunction with the conceptual frameworks, these CSFs offered

valuable perspectives for understanding IT managers’ strategies when implementing DG

programs. Alhassan et al. (2019) also proposed their set of CSFs for DG, further

contributing to the field’s knowledge base.

To address a gap in the field of DG, Alhassan et al. (2019) conducted a

comprehensive investigation using a single case study approach and incorporated

semistructured interviews with Al Rajhi Bank in Saudi Arabia. The primary objective

was to pinpoint the CSFs that underpin effective DG. Findings indicated that positioning

data as a valuable asset serves as the primary motivation for implementing DG. This

foundation drew upon the framework of Khatri and Brown (2010), encompassing five

key decision domains: data principles, data quality, data access, and data lifecycle. The

study uncovered seven core categories of CSFs, each prioritized by its level of

significance: (a) employee data competencies, (b) clear data process and procedures, (c)

flexible data tools and technologies, (d) standardized easy to follow data policies, (e)

established data roles and responsibilities, (f) clear inclusive data requirements, and (g)

focused and tangible data strategies. Notably, the success of DG programs hinged on

employee data competencies and established data roles and responsibilities, making them

critical factors to consider. Although this study’s findings may not be broadly applicable

(i.e., not easily generalized), they provide valuable insights that informed the current

study, particularly leveraging these CSFs in the context of conducting interviews with IT

managers.

Machado et al. (2022) conducted a systematic literature review to investigate the

potential role of DG in sustainable development. Their systematic literature review

commenced with the definition of DG as a set of controls over data management aimed

at maximizing data value and minimizing associated risks. Sustainable development,

conversely, encompasses economic, environmental, and social sustainability components

and strives to balance present consumption with future needs. The study revealed a gap

in existing DG frameworks, which focus on addressing regulatory antecedents (e.g., the

general data protection regulation [GDPR]) to ensure data consistency, trustworthiness,

decision-making accountability, and user privacy during implementation. Building on

Khatri and Brown’s (2010) data life cycle domain, Machado et al. proposed a

conceptual framework for sustainable DG, positioning DG within the product life cycle.

However, the proposed framework was conceptual and relied on a limited number of

articles. Future research is necessary to expand the existing body of literature, develop

models, and apply them in real-world use cases.

Data Ecosystem Model of DG

D’Hauwers et al. (2022) conducted three comprehensive case studies on Smart

Cities, focusing on the Rotterdam Digital Twin, the Helsinki Digital Twin, and the Smart

Retail Dashboard. Their primary objective was to develop a robust business framework

model that facilitates the efficient exchange of data assets among various stakeholders,

be they private or public entities, within or engaged with a data ecosystem. This endeavor

resulted in the formulation of the Data Ecosystem Model Framework, which hinges on

two pivotal factors: (1) the value of data assets (creation, revenue and cost model), and

(2) control (value network for relationships, and data governance) to meet the needs of

organizations. The significance of this Data Ecosystem Model Framework to our current

study lies in its ability to augment our understanding of strategies employed by IT

managers for data governance. Unlike conventional approaches that primarily address

data management within an organization, this framework broadens the scope to

encompass data exchanges that transpire across different organizations. Furthermore, it is

worth noting that the Data Ecosystem Model Framework effectively aligns with the

fundamental domains outlined in the Khatri and Brown (2010) framework, namely: data

principle, data quality, and data lifecycle.

Application of Khatri and Brown (2010) and Abraham et al. (2019) Governance

Frameworks to the Current Study

The aim of this section is to demonstrate how the two conceptual frameworks can

be applied as lenses to understand the IT managers motivations, challenges, and

strategies to implementing data governance. The following sections are included: data

governance definition and scope, historical overview, challenges, success factors,

frameworks, antecedents and data principles, governance scope (data and domain),

governance mechanisms, consequences, Big Data governance, and future research topics

in data governance. Firstly, the review begins by establishing a shared understanding of

data governance by defining its types and setting a common foundation for IT managers’

strategies in implementing data governance. It acknowledges that the term “data

governance” can have different meanings for different audiences. Secondly, a historical

overview of data governance is provided to offer insights into its evolution within the

academic body of knowledge. This section aims to equip readers with a comprehensive

understanding of how the subject has developed over time, up to the present day.

Thirdly, the challenges associated with data governance are addressed. This section

delves into the difficulties companies face when implementing data governance

programs, providing valuable background information. Fourthly, the review discusses

success factors, which serve as lessons learned and best practices in the implementation

of data governance programs. It recognizes that practitioners often combine industry best

practices to theoretical knowledge and design patterns to achieve successful outcomes.

Fifthly, data governance frameworks in the existing literature are examined. This section

explores how researchers approach the design, implementation, and evaluation of data

governance, offering insights into different theoretical perspectives. Sixthly, the review

explores the antecedents and data principles that justify why organizations engage in

implementing data governance in the first place. It aims to shed light on the reasons and

motivating factors behind organizations’ decision to invest in data governance initiatives.

Seventhly, governance domains are discussed. These domains encompass areas such as

data quality, data security, and compliance, where data governance is applied. This

section provides justifications for implementing data governance and highlights the

specific areas where it is crucial. Eighthly, data governance mechanisms and their

consequences are explored. Organizational aspects, roles and responsibilities, and

processes and procedures related to governance implementation are examined.

Additionally, the outcomes and benefits of data governance, such as improved data

quality and compliance, are discussed.

Data Governance Definition and Scope

As an organization asset data needs to be governed and properly managed to

boost competitiveness and help achieve strategic objectives. In recent years since 2010,

data has emerged as a valuable asset in the modern information economy. As both

private and public sector organizations have generated and consumed vast amounts of

data, the exponential growth of data and the increasing use of data-driven decision-

making

through business intelligence, companies have begun to recognize the importance of

managing and governing data as a valuable asset (Al-Ruithe et al., 2018). Data

governance (DG) plays a crucial role in managing data assets, alongside IT governance

and cloud governance, as part of the broader framework of IT information governance

and corporate governance (see Figure 4). As this study focuses on Digital Governance

(DG) as its primary area of investigation, it is crucial to establish a well-defined

framework for the literature review by clearly defining and outlining the scope of DG.

The Data Governance Institute (DGI) defines data governance as a system of

decision rights and accountabilities for managing data assets, determining who can

access and manipulate data and for what purposes (Al-Ruithe et al., 2019). Most scholars

agree that data governance involves the management and control of data assets (Alsousi

Shah, 2022; Mao et al., 2022; Al-Ruithe et al., 2018; Janssen et al., 2020). However,

Alsousi and Shah (2022) and Fernandes (2022) propose an expanded definition of data

governance that includes value creation and risk mitigation as additional dimensions.

This broader perspective implies that data governance should not only enhance the value

of data assets, such as driving innovation, but also reduce costs associated with data

assets, such as compliance failures and risks related to data handling (e.g., security

breaches, privacy violations, data leaks, and theft). These dimensions align with the

compliance, data quality, and data access aspects of Khatri and Brown’s (2010)

framework, as well as the antecedents and domain scope of Abraham, Schneider, and

Vom Brocke’s (2019) framework. Hikmawati et al. (2021) emphasized an often

overlooked but critical dimension of data governance, which is the need for

organizationwide initiatives to establish standardized processes and procedures for

managing and controlling data assets. This standardization and control enable alignment

between data management and business strategy, including regulatory compliance and

risk mitigation. In the context of the medical field, McCaig and Rezania (2021) defined

data governance as encompassing all policies, processes, and principles related to the

management, use, and security of data. This definition underscores the significance of

security considerations within data governance programs, similar to Khatri and Brown’s

(2010) data access dimension or Abraham et al.’s (2019) data domain dimension. Data

governance, as defined by various authors, revolves around several key themes that can

drive organizational success. It aims to enhance the value of data through improved

quality and trustworthiness, while also mitigating risks and reducing associated costs.

Compliance with laws and regulations is emphasized to avoid legal complications.

Importantly, data governance should be treated as an organization-wide initiative,

ensuring standardization of practices and promoting collaboration. Lastly, a key focus

lies in ensuring security and privacy for end users and data consumers, fostering trust and

loyalty.

In addition to the definition, it is crucial to distinguish between different

taxonomies of data governance. Abraham et al. (2019) proposed a framework that

includes a Data Scope dimension, which pertains to the type of data asset an organization

needs to govern: traditional data and big data. Similarly, Al-Ruithe et al. (2018) provided

a taxonomy of data governance, differentiating between traditional data governance for

non-cloud computing and cloud data governance for cloud computing. Traditional data

governance primarily focuses on data stored on-premises in structured data stores, such

as databases and data warehouses. It encompasses areas like Master Data Management

(Hikmawati et al., 2021; Haneem et al., 2019), which establish high-quality and trusted

reference data across various business functions within a company, including sales,

services, order management, purchasing, manufacturing, billing, accounts receivable, and

accounts payable. On the other hand, cloud data governance deals with Big Data stored

in flexible cloud stores, such as data lakes (Ramachandran et al., 2022; Yebenes &

Zorrilla,

2019). Traditional data governance is often associated with siloed implementation and an

IT focus, making it challenging to trust data outside of the organization (Al-Ruithe et al.,

2019; Kariotis et al., 2020). In contrast, cloud data governance faces complexities and

challenges inherent to the nature of Big Data, including volume, velocity, variety, and

veracity (Al-Sai et al., 2019). This study specifically concentrates on traditional DG

rather than Big Data DG. Consequently, the main emphasis of the literature review is on

traditional DG. The following paragraph offers a concise history of DG, providing

readers with a contextual understanding of the evolution of this topic within the realm of

research literature.

Figure 4

Interrelations Between Governance Domains (Al-Rhuite et al., 2018)

Brief History of Data Governance

The evolution of DG as a research topic mirrors that of data technologies starting

from relational databases in the 1970s, then datawarehousing and business intelligence in

the 1990s and 2000s, an finally the data explosion with Big Data and advanced analytis

since 2010. Data governance traces its early origins back to the emergence of relational

database management systems (RDBMS) in the 1970s (Al-Ruithe et al., 2019). The need

for structured data management practices became apparent with the introduction of data

models, dictionaries, and integrity enforcement rules. The mid-1990s saw a rise of

regulatory compliance requirements with federal data protection laws (Mulliganet al.,

2019) in regulated sectors such as healthcare, finance, and telecommunications. This

trend led organizations to develop data policies and procedures to ensure data privacy,

security, and reporting (Al-Ruithe et al., 2019). The next period is that of

datawarehousing and business intelligence.

During the late 1990s, as data warehousing and business intelligence gained

prominence, organizations began recognizing the strategic value of data. Consequently,

there arose a need to govern data assets, including master data, to ensure the quality

required for effective decision-making, such as accuracy, consistency, and reliability

(Khatri & Brown, 2010; Abraham et al., 2019). In the early 2000s, data governance

emerged as a distinct discipline within organizations, marked by the establishment of the

Data Governance Institute (DGI) in 2004. The DGI aimed to promote best practices and

frameworks like the Data Management Body of Knowledge (DMBOK), which offered

authoritative guidance in the field (Aisyah & Ruldeviyani, 2018). The digital economy’s

present phase is characterized by an exponential surge in data starting from 2010,

necessitating enhanced data asset governance.

The period since 2010 is marked by an unprecedented data growth worldwide

which is accompanied by increased concerns for data quality, privacy, and security and

justifying the development of additional data governance frameworks. The frameworks

proposed by Khatri and Brown (2010) and a decade later by Abraham et al. (2019) have

been a response to the gap and need. Moreover, the adoption and maturation of cloud

computing, along with the proliferation of Big Data generated by social media and the

Internet of Things (IoT), coupled with the automation and exploitation capabilities of AI,

have presented new challenges for data governance (Al-Sai & Abdullah, 2019). To

address these challenges and safeguard data quality, privacy, and security, organizations

must embrace Big Data and AI governance and research must fill the gap with Big Data

governance frameworks (Yebenes & Zorrilla, 2019; Zorrilla & Yebenes, 2022). This

ensures that both innovation and customer interests are protected (Al-Badi et al., 2018;

Al-Ruithe et al., 2018; Al-Ruithe et al., 2019). Data governance continues to evolve to

keep pace with challenges stemming from advancements in technology, changes in

regulations, and the ever-expanding range of complex use cases.

Data Governance Challenges

The mission of Data Governance (DG) is to effectively manage and control data

assets to maximize their value while minimizing associated costs and risks. This entails

addressing organizational, process, technological, and human challenges inherent in

managing and controlling data assets. The problem statement which is the genesis for

this study is that many companies fail to implement working data governance programs

due to challenges impacting IT managers strategies. This study offers valuable insights

into data governance implementation, benefiting companies and data governance

practitioners in multiple ways. Firstly, it sheds light on the factors and challenges

associated with data governance, providing organizations with a deeper understanding of

the key elements involved. This knowledge empowers companies to make informed

decisions regarding data governance strategies and implementation. Furthermore, the

study highlights emerging trends in the data governance field, enabling organizations to

stay up-to-date with the latest developments. By being aware of these trends, companies

and practitioners can proactively adapt their data governance practices to align with

industry advancements and best practices.

This literature review uncovered four categories of challenges for traditional DG

faces several challenges, and will be discussed in the following sub-sections:

•Methodological challenges: There is a lack of a comprehensive framework to

guide the implementation of DG practices.

•Technical challenges: DG often suffers from siloed or inadequate

implementation, technology limitations, data quality issues, and concerns

related to data security and privacy.

•Organizational challenges: DG efforts tend to be IT-focused and may lack

clear roles and responsibilities, well-defined data policies, and alignment with

the organization’s strategies.

•Compliance and regulatory challenges: DG must navigate the complexities of

compliance and regulatory requirements, ensuring that data management

practices meet legal and industry standards.

Methodological Challenges. Traditional data governance (DG) was initially

suffering a gap in research with a lack of frameworks to guide both research and

practitioners. Although traditional data governance (DG) has reached a certain level of

maturity, there are still gaps in the existing research that warrant further investigation.

Khatri and Brown (2010) were early contributors to the field of data governance

research. They noticed a lack of a framework to assist with DG implementation and thus

proposed a unified conceptual framework (referred to as the Khatri and Brown (2010)

Data Governance Framework in this study, depicted in Figure 1) to guide researchers and

practitioners in developing effective programs. The Khatri and Brown DG framework

encompasses five decision domains crucial for understanding and guiding DG

implementation: data principles, data quality, data access, data lifecycle, and metadata.

Over time, this framework has become a point of reference for future research. Nearly a

decade later, Abraham et al. (2019) found that despite the maturity of DG practices in

various industries, including healthcare, there was still a lack of a comprehensive

framework for data governance that could guide practitioners and researchers. In

response, they developed a conceptual framework (referred to as the Abraham,

Schneider, and Vom Brocke (2019) Conceptual Framework for Data Governance in this

study, illustrated in Figure 2) to evaluate data governance implementation based on six

dimensions of analysis: governance mechanisms, organizational scope, data scope,

domain scope, antecedents, and consequences of data governance. In the quest to enable

data-driven decision making in the evolving information economy and the emerging

Industry 4.0 era, Bento et al. (2022) explored how DG frameworks could best facilitate

this process. Similarly the study identified research gaps in data governance frameworks,

including the absence of a holistic implementation guide as challenges that cause DGs to

be unsustainable. To address this, the authors proposed a matrix of critical success

factors (CSF) that could enhance the success of DG programs: employee data

competencies, clear data processes and procedures, flexible tools and technologies,

standardized data roles and responsibilities, clear and inclusive data requirements,

focused and tangible strategies, stakeholder buy-in, effective and strategic

communications, and assessment of data governance maturity.

Over the past two decades, the body of knowledge on DG (Data Governance) has

significantly expanded, and several conceptual frameworks have been adopted as

references. However, it is widely acknowledged among researchers that more research

on DG is necessary. Specifically, there is a need for authoritative reference architectures

and best practices to guide the implementation of DG programs for DG and IT

practitioners. Implementing DG solutions presents challenges due to their technology-

stack dependency, high costs, and variations across organizations based on their unique

antecedents. This literature review identified Critical Success Factors (Alhassan et al.,

2019; Bento et al., 2022; Ibrahim et al., 2021; Yatya et al., 2021) that hold the potential

to provide a solid foundation for implementing DG. To leverage these factors effectively,

further quantitative research is required to validate their applicability across various DG

practices, industries, and the broader community of researchers. Once validated, these

findings could empower practitioners with valuable insights for successful DG

implementation. Enhancing the effectiveness of DG requires a comprehensive approach

that not only incorporates methodological frameworks with guidance and best practices

but also emphasizes resolving the inherent technical and organizational challenges that

frequently lead to the failure of DG implementations.

Technical Challenges. Traditionally, organizations, both business and

nonbusiness, have relied on their IT departments to manage their computing, networking,

applications, and data infrastructures. This has led to a strong IT-focused approach to

data management and data governance, often implemented in silos. Khatri and Brown

(2010) differentiated between IT assets (computers, networking, database systems) and

information assets (data), where IT assets facilitate the management of information

assets, and governance being concerned with the efficient use of IT. Al-Ruithe et al.

(2018) emphasized the interdependence of data governance and IT governance within the

broader concept of Information Governance (see Figure 1). They also identified IT-focus

and siloed implementation as significant factors contributing to the failure of data

governance initiatives. In contrast to Khatri and Brown (2010), Al-Ruithe et al. (2019)

added the complexity of business use cases for data as another factor affecting data

governance implementation. In healthcare, Kariotis et al. (2020) found that data

governance is often implemented in silos by different institutions, making it challenging

to establish trust across organizations. Similarly, Ngesimani et al. (2022) identified

technological challenges in the healthcare sector in South Africa, such as data quality

issues and limited competencies of IT professionals in operating electronic medical

record systems (EHR) and healthcare information systems (HIS), hindering the success of

data governance programs. In the public sector of China, Mao et al. (2022) highlighted

data quality issues caused by a lack of effective control over the data lifecycle (data entry,

processing, and reporting) as a barrier to successful data governance. Scope et al. (2022)

discussed compliance challenges faced by data governance due to technological

limitations in storage systems to enforce data privacy and security throughout the entire

data lifecycle (creation, storage, usage, archival, destruction). They suggested the need

for further research and development to integrate data governance into platforms natively.

Cloud Service Providers like Microsoft are playing a role in democratizing data

governance by offering tools such as Azure Purview, which require less financial

investment compared to traditional enterprise data governance applications. However,

companies still face challenges in complying with data governance regulations, such as

GDPR, which necessitate a strong understanding of how data is stored, used, and

managed. These challenges arise from issues related to data quality (inaccuracy and

incompleteness), IT-siloed environments (fragmented enterprise architecture and legacy

systems), and compliance with regulations (Abraham et al., 2019).

To address the IT-focus and siloed implementation challenges, companies are

encouraged to make data governance a company-wide initiative, supported by top

leadership, which recognizes data as a valuable business asset (Yatya et al., 2021).

Furthermore, data-driven transformation (Smith & Heffernan, 2019) requires considering

data as a strategic asset within the company and fostering a culture that values high data

quality and trustworthiness. This approach maximizes the value of data and reduces costs

associated with risks, both of which are outcomes of effective data governance.

However, a practical challenge lies in transforming a company’s culture and processes to

prioritize data as an asset, as simply producing transactional data does not equate to

recognizing its value. It requires executive direction to embed this mindset throughout

the organization.

Implementing data governance involves establishing organizational standards and

processes, which can be subject to various organizational challenges, such as IT-focus

and siloed implementation, lack of clearly defined roles and responsibilities, inadequate

data policies, and misalignment with organization-wide business strategies. The issues of

IT-focus and siloed implementation were discussed in the previous paragraph.

Organizational Challenges. As a vital component of Information Governance,

Data Governance (DG) is not an isolated process. It is heavily influenced by various

factors within an organization, such as its culture, strategy, roles and responsibilities, and

policies. Some of these factors can hinder the implementation and success of DG. Yatya

et al. (2021) emphasized the increasing importance of DG in the digital economy and its

crucial role in establishing a foundation for data architecture, security, and quality. By

enhancing the value of data assets and mitigating associated risks, DG contributes

significantly to overall organizational success. However, the authors highlight cultural,

political, and organizational resistance as key factors that can negatively impact the

effectiveness of a DG program. To ensure successful DG implementation, several best

practices have been identified. These include a phased approach to implementation,

support from senior management, clearly defined roles and responsibilities, goal

measurement using metrics, and early communication prioritization. Yatya et al. (2021)

also suggest two additional best practices: providing non-monetary internal rewards,

such as career growth opportunities, to encourage participation in the DG program, and

periodic reminders from a Data Steward about policies and new guidelines.

Aisyah and Ruldeviyani (2018) conducted a study on DG design and

implementation for the Indonesia Deposit Insurance Corporation (IDIC). They identified

challenges such as siloed processes, scattered data assets across business units, and a lack

of defined roles for data management within the organization. Another study by Alsousi

and Shah (2022) focused on Small and Medium-sized Enterprises (SMEs) and evaluated

the value of DG for these organizations. The researchers found that SMEs face specific

challenges in implementing DG, including poorly implemented data governance

frameworks, unclear definitions of duties and responsibilities, a lack of leadership and

organizational support due to perceived cost-benefit concerns, and external factors

beyond SMEs’ control. Several other studies, including those by Alhassan et al. (2019),

Almeida et al. (2019), Alsousi and Shah (2022), and Bento et al. (2022), highlight critical

success factors (CSFs) for DG programs. These studies emphasize the need for DG to be

a company-wide initiative and identify common organizational challenges such as an

ITfocused approach, siloed implementation, lack of clear roles and responsibilities, and

insufficient executive and company-wide support. The discussion of these CSFs will be

further explored in the Data Governance Success Factors section. Companies still

encounter challenges in achieving compliance and effectively implementing DG

practices across industries including healthcare.

Regulatory Compliance Challenges. In regulated industries such as healthcare

and finance, Data Governance (DG) plays a crucial role in protecting patient or

consumer records from misuse and abuse. Tiffin et al. (2019) recognized the benefits of

digital health in developing countries, but also highlighted the prevalent security and

privacy challenges surrounding consumer health data in those regions. The healthcare

sector in developing countries often faces weak or non-existent privacy laws and limited

security implementation. To address this gap, Tiffin et al. (2019) proposed a

comprehensive DG framework with four dimensions (Figure 5): (1) Ethics and consent,

(2) Data access protection, (3) Documentation, backups, accessibility, (4) Legal

framework.

Mahanti (2022) emphasized compliance as a significant driver for data

governance, extending beyond the healthcare sector discussed by Tiffin et al. (2019).

Non-compliance to governance regulations comes at high cost;therefore, the essential

role of data governance in mitigating non-compliance risks. With the exponential growth

of data since 2010, the potential for malicious data breaches has also increased. In 2021

alone, businesses across industries were fined a total of $957.3 million for compliance

violations, with the largest fine of €746 million imposed on Amazon under GDPR.

Technological challenges, such as data breaches, contribute to compliance difficulties,

along with the cumbersome and costly requirements of consumer privacy protection

regulations like GDPR. Mahanti (2022) recommended that compliance should be a

collaborative effort involving all stakeholders, including business, IT, and other

departments with a vested interest in data compliance. Mulligan et al. (2019) also

examined consumer data protection and privacy, providing an overview of US federal

data protection laws and comparing them to the EU’s General Data Protection

Regulation (GDPR) and the California Consumer Privacy Act (CCPA) of 2018. These

regulations apply to companies involved in collecting, storing, distributing, or selling

information and grant consumers three main rights: the right to know, the right to opt

out, and the right to delete. While the CCPA mirrors some aspects of GDPR, the latter

provides more extensive protection for consumer personal data. The study also covered

federal data protection laws such as the Gramm-Leach-Bliley Act (GLBA) of 1999,

Consumer

Financial Protection Bureau (CFPB), Health Insurance Portability and Accountability

Act

(HIPAA), Fair Credit Reporting Act (FCRA), and Consumer Financial Protection Act

(CFPA). These laws establish governance standards for how consumer personal data

should be handled by both data processors and controllers.

Acetedenents and Data Principles

Motivating factors often lead organizations to invest in new information systems

(IS) and implement policies or procedures. These motivations are often driven by the

need to support business functions, ensure regulatory compliance, and meet industry

standards. Antecedents and data principles work the same way when it comes to DG.

The antecedents outlined in Abraham, Schneider, and Vom Brocke’s (2019) Conceptual

Framework for Data Governance align with the data principles in Khatri and Brown’s

Unified Framework for Data Governance (2010). Both frameworks identify the internal

and external factors that influence the adoption and implementation of data governance

(DG). Implementing DG programs has traditionally been motivated by regulatory

compliance and data-related issues such as poor data quality, limited access, and lack of

trustworthiness (Vojvodic & Hitz, C. 2022). Mahanti (2022) emphasized the cost of

noncompliance to governance regulations and highlights the crucial role of DG in

mitigating non-compliance risks. Since 2010, the exponential growth of data volumes

and velocity has increased the potential for malicious data breaches. In 2021 alone,

businesses across industries faced fines totaling $957.3 million for compliance

violations, including the record-breaking €746 million GDPR fine imposed on Amazon.

This study aims to understand IT managers strategies for implementing DG. Those

strategies stem from specific motivations, let it be regulatory compliance with respect to

data, data quality, efficiency, or other organizational needs. Additionally, this study

underscores that regulatory compliance is a significant driver and antecedent for DG,

necessitating the establishment of good data quality practices. Achieving compliance

should be a collaborative effort involving all stakeholders, including business, IT, and

other departments with a vested interest in data compliance. By fostering good

governance on specific domain scope and maintaining high data quality standards,

organizations can streamline the attainment of compliance objectives.

Domain Scope

Data Governance aims to effectively control and manage data assets in specific

domains. DG decisions are applied to various aspects of data assets, such as data quality

and security. Abraham, Schneider, and Vom Brocke (2019) proposed a conceptual

framework for Data Governance that includes three dimensions for data assets: Data

Scope, Organization Scope, and Domain Scope. Data Scope refers to the different types

of data assets, distinguishing between traditional data assets stored on-premises in

relational databases and data warehouses, and cloud-based data assets including

relational, non-relational, unstructured, semi-structured, and Big Data. Organization

Scope determines the unit of analysis for DG, distinguishing between intra-

organizational scope (within a single organization) and inter-organizational scope (across

multiple organizations). Domain Scope corresponds to specific domains of decision-

making that are subject to DG mechanisms, including data quality, data security, data

architecture, data lifecycle, metadata, data storage, and infrastructure. These decision

domains align with the dimensions of Data Quality, Data Access, and Metadata in the

Khatri and Brown (2010) Data Governance Framework. Data quality and data security

are essential pillars of DG. They are crucial for decision-making through data analytics,

as well as for regulatory compliance and risk mitigation. IT managers motivations to

implement DG may stem from the need to improve data quality, trustworthiness,

security, or data infrastructure and processes. Companies employ a combination of

governance mechanisms to control and manage activities within each decision domain of

DG. These mechanisms help ensure that data assets are effectively governed and utilized

in accordance with organizational goals and regulatory requirements.

Governance Mechanisms

Governance mechanisms refer to the formal structures, processes, and procedures

employed to manage a data governance program. Scholars have categorized governance

mechanisms into three groups: structural mechanisms, procedural mechanisms, and

relational mechanisms (Abraham et al., 2019; Khatri and Brown, 2010). These

mechanisms encompass various aspects such as data management, decision-making,

roles and responsibilities, monitoring, and collaboration among stakeholders. Structural

mechanisms encompass reporting structures, governance bodies, and defined roles and

responsibilities. Typical roles within a data governance program include executive

sponsor, data governance leader, data owner, data steward, data governance council, data

governance office, data producer, and data consumer (refer to Table 4). Procedural

mechanisms primarily focus on effective data management, including data acquisition,

security, and utilization. They involve developing data strategies, policies, standards,

processes, procedures, and handling data-related issues. Relational mechanisms aim to

foster collaboration among stakeholders throughout the daily operations of the data

governance program. These mechanisms include communication, training, and

coordination of decision-making. Communication plays a crucial role in raising

awareness of the program among stakeholders and the entire organization. Training

ensures that stakeholders possess the necessary skills and knowledge for successful

program implementation. Analyzing governance mechanisms in a specific DG

implementation offers valuable insights into both the strategies employed by IT

managers and the challenges they encounter during the process. By implementing and

effectively executing governance mechanisms, positive outcomes or consequences can

be achieved for data governance, such as improved data quality and controlled access to

data assets.

Table 4

Data Governance Roles and Responsibilities

Role Responsibility

Executive sponsor Strategic direction, business prioritization and alignment,

funding

Data governance

leader

Day to day management of the DG program, coordinates tasks

of data stewards, reports on program performance

Data owner Line-of-business executives accountable for data assets of

their respective business units

Data steward Business leaders or subject matter experts with detailed

knowledge of business and data requirements; translate

requirements into technical specifications

Data governance

council

Hierarchy with cross-functional body, establishes the strategic

vision of the entire governance program and aligns it with

organization goals and strategy

Data governance

office

Staff that supports governance and decision-making activities

as well as the data steward teams and governance councils

Data producer Creates data or aggregates data from third-party

Data consumer User of the data

Consequences of Data Governance

The primary outcomes sought out for DG are to add value to data assets while

reducing risks associated with their use. Abraham, Schneider, and Vom Brocke (2019)

proposed a Data Governance Conceptual Framework that includes various dimensions,

such as antecedents and consequences. On one hand antecedents, analogous to data

principles in Khatri and Brown’s Unified Framework for Data Governance (2010), refer

to the factors that influence the adoption of DG. Regulatory compliance and data-related

issues (e.g., poor quality, limited access, lack of trustworthiness) have traditionally been

key motivations for implementing DG programs. On the other hand, the consequences of

DG are the positive outcomes sought through its implementation, such as improved data

quality, enhanced data access, and strengthened data security (Yatya et al., 2021;

Hikmawati et al., 2021; Khatri & Brown, 2010). For instance, master data management

(MDM) serves as a traditional form of DG aimed at providing high-quality and

trustworthy reference data utilized across the organization (Ibrahim et al., 2021). Such

reference data mitigates the risks of errors in business intelligence and analytics,

ultimately leading to better decision-making. The primary rationale for data governance

(DG) is to establish policies and procedures that enable the exertion of authority and

control over data assets management. The ultimate goal is to maximize value while

minimizing risks. DG can be conceptualized using an input/output model, comprising

influential factors and desired outcomes. Understanding the consequences of Data

Governance (DG) is crucial in evaluating the success of its implementation and

determining if it aligns with the motivations that drove IT managers to adopt these

programs initially. By comprehending the outcomes of DG, one can gain insights into its

effectiveness and whether it fulfills the original objectives set by IT managers. Numerous

researchers in the field have identified success factors for DG, offering guidance to

practitioners during program implementation.

Data Governance Success Factors

Researchers and practitioners in the fields of Corporate Governance, Information

Governance, and IT Governance not only share a theoretical foundation but also

exchange best practices to guide the design and implementation of various governance

programs, including Data Governance. Alhassan et al. (2019) identified seven CSFs for

successful Data Governance programs, including employee data competencies, clear data

processes and procedures, flexible data tools and technologies, standardized data

policies, established data roles and responsibilities, clear data requirements, and focused

data strategies. Bento et al. (2022) found that organizational Data Governance programs

often lacked a holistic framework for sustainable implementation. As a solution, they

proposed a set of CSFs, which overlapped with the CSFs identified by Alhassan et al.

(2019) and added three additional factors: stakeholder buy-in, effective and strategic

communications, and regular assessment of data governance maturity. Daneshmandnia

(2019) focused on the influence of organizational culture on the effectiveness of

Information Governance in higher education, with Data Governance being a subset of it.

The study highlighted the positive impact of a culture of competition/result-orientation,

control/hierarchy, and trust, while information silos were found to have an adverse effect

on Information Governance. This aligns with the stakeholder support and strategic

communications aspects emphasized by Bento et al. (2022). Ibrahim et al. (2021)

conducted research on critical factors affecting master data quality, which serves as a

single reference for core data across an organization. They identified 19 factors grouped

into five categories: organizational, managerial, stakeholder, technological, and external.

Several of these factors overlapped with the CSFs identified by Bento et al. (2022) and

Alhassan et al. (2019).

In addition to Critical Success Factors (CSFs), researchers have explored various

best practices that significantly contribute to the success of DG (Data Governance)

programs.Yatya et al. (2021) conducted literature reviews on best practices for

implementing successful Data Governance programs. They emphasized the importance

of full stakeholder support, organization-wide cooperation, and adapting strategies to suit

different organizational cultures. The study also acknowledged internal factors such as

cultural, political, and organizational resistance that can hinder the success of a Data

Governance program. The five main best practices they identified were: starting with

small steps, acquiring support from senior management, defining clear roles and

responsibilities, measuring goals with metrics, and prioritizing early communication.

This aligns with the findings of Daneshmandnia (2019), Bento et al. (2022), and

Alhassan et al. (2019) regarding internal organizational factors, senior management

support, clear roles and responsibilities, maturity assessment, and communication.

Considering the overlap in the CSFs identified by different studies, practitioners

including IT managers can leverage this body of literature to implement effective Data

Governance programs. Conversely, the Critical Success Factors (CSFs) could aid in

comprehending DG strategies and challenges. Table 5 provides a unified synthesis of the

CSFs, which can serve as a practical guide for practitioners. The next section will focus

on discussing Big

Data Governance based on the selected literature from this study.

Table 5

Alhassan et al. (2019) Critical Success Factors for Data Governance

Critical success factor

Employee data

competencies

Data governance activities that involve human

action. Directly impact the defining, implementing,

and monitoring of data processes and procedures,

as well as data policies and data requirements.

Clear data processes and

procedures

Data policies in place which should be detailed and

operationalized during activities related to data

processes and procedures.

Flexible data tools and

technologies

All the activities related to software and hardware

that affect the data in an organization, including the

presentation and storage of data.

Standardized easy-tofollow

data policies

Data policies are short statements that provide the

high-level guidelines and rules necessary for

dealing with data. These relate to data principles.

Established data roles and

responsibilities

Individual(s) responsible for the data-related

activities in the organization, who defines the

policies and processes for the data as well as

assigning the duties for the actions related todata.

Clear inclusive data

requirements

Define all aspects of data implementation, suchas

data flows and integration.

Focused and tangible data

strategies

Include planning fordata governance in order to

achieve its goals, as well as the main activities

related to considering data as assets.

Detail

Other Data Governance Models and Future Topics

Since 2010, DG research has undergone significant development, transitioning

from a significant gap to a mature body of knowledge. Numerous conceptual frameworks

have emerged in this field, providing valuable references for subsequent studies. Among

these frameworks, Khatri and Brown’s foundational framework (2010) has served as a

reference for scholars, including Abraham, R., Schneider, J., & Vom Brocke, J. (2019).

In my study, these two frameworks have been extensively discussed within the

Theoretical or Conceptual Framework section. Alsousi and Shah (2022) conducted a

comprehensive literature review to identify factors influencing the adoption of DG by

small and medium-sized businesses (SMEs). Their study proposed a conceptual

framework that encompasses six domains: Technology Context, Organizational Context,

External Task Environment Context, Technology Use, Data Quality, Organization Scope

and Domain Scope dimensions from the framework proposed by Abraham, R.,

Schneider, J., & Vom Brocke, J. (2019). Furthermore, Al-Ruithe et al. (2018) introduced

a conceptual framework aimed at establishing a unified perspective for comprehending

data governance in both traditional and cloud environments. They outlined a taxonomy

consisting of two dimensions: (1) traditional (non-cloud) data governance, which

encompasses people and organizational bodies, policy and processes, and technology;

and (2) cloud data governance, which encompasses data governance structure, policy and

processes, cloud deployment model, service delivery model, cloud actors, organizational

and technological aspects, service level agreements, monitoring matrices, and legal

considerations. Given the increasing prevalence of cloud computing and big data,

companies are progressively migrating their data assets, infrastructure, and processes to

cloud platforms, thereby elevating the significance of cloud DG as a highly compelling

topic. Overall, these conceptual frameworks, along with the insights gained from

research conducted by Alsousi and Shah (2022) and Al-Ruithe et al. (2018), contribute to

a comprehensive understanding of data governance in various contexts, including SMEs,

traditional environments, and cloud-based settings across industries and sectors.

Several DG frameworks have been proposed in various domains, such as

government (Girard, 2020; Mao et al., 2022) and master data management (Hikmawati et

al., 2021; Haneem et al., 2019). Jimenez and Duarte (2019) focused on Data Governance

implementations for business organizations, identifying the a crucial role of DG in

managing data assets to increase profitability, reduce costs, and mitigate risks. The

complexity of data use cases necessitates effective data governance.

Focusing specifically on Canada, Girard (2020) proposed a data governance

framework to assist business leaders in aligning their organizations with digitization and

AI. This framework comprises twelve dimensions, which share similarities with Khatri

and Brown’s (2010) framework. The dimensions include data governance objectives,

scope (historical vs. archived), roles and accountabilities, data ownership rights, data

collection, data access and lifecycle, data analytics use cases, data residency, privacy,

ethics and trust, approval and implementation, as well as compliance and certification.

Other frameworks have also addressed data quality, particularly in the context of master

data management (MDM).

The primary objective of Master Data Management (MDM) is to establish a

centralized repository for master data, encompassing customer, supplier, material, and

employee data, which can be utilized by all organizational processes, including Business

Intelligence. Hikmawati et al. (2021) conducted a literature review to explore the role of

MDM in enhancing organizational data quality (DQ) and data governance (DG). MDM

is particularly sensitive to data quality, as data quality issues are inherent within

organizations due to the diverse sources and varying standards through which data is

generated and managed. Poor data quality can lead to detrimental consequences such as

duplication, inaccuracy, and inconsistency of information. The MDM process primarily

involves data profiling, standardization, consolidating master data into a unified

repository, integrating it with existing applications, and finally synchronizing it with

business processes and applications. DG aims to establish ownership, define roles and

responsibilities, and establish rules for data management. The ultimate objective is to

ensure that data is reliable, trustworthy, and delivered in a timely manner. MDM and DG

are inter-related as MDM facilitates enhanced data management and quality through the

roles and responsibilities defined by DG.

MDM use cases for government sector call for more research. Haneem et al.

(2019) focused on identifying the key factors influencing the adoption of MDM by local

governments in Malaysia. Utilizing the Technology-Organization-Environment (TOE)

framework, they developed a conceptual model that highlighted factors such as

complexity, top management support, technological competence, and citizen demand as

influential in the adoption of MDM. Additionally, Mao et al. (2022) recognized the

increasing need for DG within China’s public sector and the lack of effective controls

over data entry, processing, and reporting in data asset management. To address these

issues, they proposed a robust data governance framework (DGF) tailored for the public

sector. The government DGF encompasses technical aspects of data governance as well

as organizational challenges such as defining roles and responsibilities, establishing data

policies, and aligning with organizational strategies. Using a design science approach,

the study proposed a tiered structure for the government DGF, consisting of public

services departments as the business frontend, the administrative department as the

backend, and a data middle platform as a bridge between the two, enabling reusable data

governance.

Implementing a government DGF ensures control and expansion of core organizational

data assets associated with administrative processes and systems. In the following

paragraphs I am discussing other DG models such as ecosystem data governance,

sustainable data governance, self-governance through intermediaries, trust-based data

governance, and security and privacy-focused data governance.

Ecosystem Data Governance Model

With the rapid expansion of data, its significance as a valuable business asset has

grown, resulting in the necessity for effective data governance. Lis and Otto (2021)

aimed to investigate the key characteristics and attributes of data governance within

ecosystem settings. They defined data ecosystems as independent organizations that

engage in data sharing to drive business innovation and implement use cases. As data

ecosystems continue to proliferate, it becomes crucial to establish data governance

frameworks to bridge existing gaps. However, implementing data governance in such

ecosystems can be challenging due to varying practices and standards across different

parties. The study proposed a three-layer taxonomy consisting of eight dimensions: the

data layer (data ownership, decision rights), the governance layer (configuration,

structure, mechanism), and the interaction layer (purpose, scope, phase). Limitations of

the study are notably the scarcity of research in this area and the absence of a universally

accepted definition of a data ecosystem.

Sustainable Data Governance Model

The explosive growth of data, commonly referred to as Big Data, presents

significant challenges in terms of acquisition, processing, storage, and secure

management, while also ensuring user privacy. In a recent study by Machado et al.

(2022), a systematic literature review was conducted to explore the role of Data

Governance in sustainable development. Data governance was defined as the

implementation of controls over data management to maximize its value and mitigate

associated risks. Sustainable development, on the other hand, aims to promote current

consumption without compromising future utilization, encompassing economic,

environmental, and social sustainability. The literature review revealed a gap in existing

data governance frameworks, which primarily focus on addressing regulatory aspects

such as GDPR to ensure data consistency, trustworthiness, decision-making

accountability, and user privacy. To bridge this gap, the study proposed a conceptual

Data Governance framework for Sustainable Data Governance, integrating data

governance practices within the product lifecycle. It is important to note that the

proposed framework is conceptual and relies on a limited number of articles,

highlighting the need for future research to expand the existing body of literature.

Furthermore, the development of models and their application in practical use cases

should be considered to enhance the framework’s effectiveness and applicability.

Self-Governance Through Intermediaries Model

Traditional DG within an organization is primarily self-governance whereby the

organization establishes its own standards and controls over data to achieve given

compliance requirements. According to Khatri and Brown (2020), data quality is

essential and can be ensured through core requirements such as accuracy, credibility, fit

for purpose, quantifiability, and relevancy. Medzini and Levi-Faur (2023) investigated

the Self-governance approach to Data Governance (DG) by focusing on credibility, using

a proxy in three use cases. The findings of their research indicate that credibility can be

established through third-party data validation or certification, which serves as a robust

foundation for regulatory compliance.

Trust-Based Data Governance Models

Most traditional data governance models in the existing body of research are

control-based as opposed to trust-based. Van der Sloot and Keymolen (2022) extensively

explored the concept of trust-based data governance models as a viable alternative to

control-based models. They examined two main existing models of data regulation

present in the literature: the control model and the legal standards model. The control

model grants individuals full rights to their own data, but its enforcement is challenging,

and it can impede business operations and hinder innovation. On the other hand, the legal

standards model relies on government enforcement of data regulations, with an

independent organization responsible for administering sanctions and enforcing rules.

However, this model places a significant burden on the government and often focuses

primarily on larger data processing companies.

In contrast to these two models, the authors proposed a third model based on

trust, which involves a trustor and a trustee. In this trust-based model, a third-party entity

is entrusted with assisting individuals in making decisions about their data. Three

trustbased data management models were outlined: information fiduciaries, data curators

and stewards, and data trusts. Each of these models has its own advantages and

disadvantages.

In the case of information fiduciaries, organizations are entrusted with managing

individuals’ data, and they must earn and maintain loyalty and trustworthiness. However,

a challenge arises in determining who monitors the fiduciary and ensures that they act in

the best interests of the citizens. The data curator, data custodian, and data steward model

involve assigning specific roles within an organization to enforce and monitor data

quality and privacy. For example, a data protection officer (DPO) may be employed by

the data processing organization, which acts as the trustee. While this model offers some

benefits, it also raises questions about potential conflicts of interest and independence.

Lastly, the data trust model entails the establishment of a dedicated institution

responsible for governing data on behalf of customers and acting in their best interests.

This model aligns well with a data-driven culture, but its implementation requires careful

consideration and planning. Most countries currently employ a combination of regulatory

models for data governance, typically involving informed consent and governmental

regulation. However, the trust-based models proposed in their study offer an alternative

approach, with the data trust model being particularly well-suited to a data-driven

culture.

Consumer Health Data Security and Privacy

Digital health benefits in developing countries have been well-documented in the

health literature; however, it is important to acknowledge the associated risks. Tiffin et

al. (2019) conducted a study focusing on the challenges of consumer health data security

and privacy in developing countries’ healthcare sectors. These challenges arise due to the

nascent stage of security implementation and the absence of robust privacy laws in these

regions. To address the existing data security and privacy gap, the authors proposed a

comprehensive DG framework, as illustrated in Figure 5. The framework is built upon

four essential pillars:

•Ethics and Consent: This pillar emphasizes the need to prioritize ethics and

consent in digital health initiatives. It involves identifying vulnerable

populations, obtaining informed consent, and transparently disclosing the

procedures involved.

•Data Access Protection: Procedural oversight and structural control are

critical aspects of safeguarding data access. This pillar focuses on

implementing mechanisms to protect data from unauthorized access. It

encompasses establishing oversight processes and maintaining structural

controls to ensure data privacy and security.

•Documentation, Backups, Accessibility: This pillar highlights the importance

of maintaining comprehensive documentation, backups, and ensuring

accessibility to health data. Proper documentation facilitates accountability,

while backups ensure data resilience. Additionally, ensuring accessibility

enables effective utilization of health data for improved healthcare outcomes.

•Legal Framework: The final pillar emphasizes the need for a robust legal

framework to protect individuals’ right to privacy and ensure data security. It

also addresses the oversight of third-party access to health data, promoting

responsible data handling practices.

By incorporating these four pillars, the proposed DG framework aims to bridge

the existing gaps in data security and privacy in developing countries’ healthcare sectors.

Implementation of this framework can help enhance the benefits of digital health while

mitigating potential risks and ensuring the protection of individuals’ privacy and

security.

Figure 5

Pillars of Digital Data Governance to Ensure Protection of Individuals (Tiffin et al.,

2019)

Big Data Governance

Although the primary focus of this study is traditional data governance, the

inclusion of Big Data governance in the literature search is essential for several reasons:

1) ensuring completeness, 2) acknowledging the growing use of Big Data assets across

industries, and 3) recognizing the expanding body of knowledge on Big Data and its

need of governance. These aspects informed the literature search strategy for this study.

In the subsequent sections, I will present a concise overview of Big Data, encompassing

its definition, challenges, opportunities, and governance frameworks.

Big Data Definition. Over the past decade, there has been an unprecedented data

explosion driven by social media, the Internet of Things (IoT), and transactional business

processes giving rise to Big Data. This surge in data has necessitated the development of

specialized frameworks for Big Data governance. Big Data refers to extremely large and

complex data sets that require unique management, processing, and storage capabilities

(Jansen et al., 2020). These data sets are characterized by three key attributes: Volume

( vast amounts of information, often measured in terabytes, petabytes, or even exabytes),

velocity (generated at high speeds, e.g., streaming data, requiring real-time or near-

realtime processing to enable continuous processing and storage), variety ( various

formats and types, including structured data, semi-structured data, and unstructured data.

The literature review revealed a lag in Big Data governance research in general

and the lack of data governance frameworks in particular. Yebenes and Zorilla (2019)

identified a lack of conceptual frameworks to guide Big Data governance research and

implementation. To effectively leverage Big Data for advanced analytics, such as

business intelligence, machine learning, and AI, it is crucial that the data is timely,

consistent, reliable, trustworthy, and of high quality. Therefore, the field of Big Data

requires governance frameworks to guide researchers and practitioners. Although Big

Data is being applied across various sectors, including healthcare, retail, education, and

government, there is still a research gap in the area of Big Data governance, primarily

due to the unique nature of these data sets. Yebenes and Zorilla (2019) proposed a

conceptual framework for Big Data governance, and a reference framework to guide Big

Data governance practice (Zorilla & Yebenes, 2022). However, challenges related to

privacy and security persist in protecting the consumers who generate this data and

contribute it to various platforms. In the following section, we will explore the benefits

and challenges associated with Big Data.

Big Data Benefits and Challenges. By its nature and complexity, Big Data sets

cannot be managed, process, or stored by traditional data processing systems. Numerous

authors covered in this literature review discussed opportunities and challenges of Big

Data including governance. Big Data brings the following benefits:

•Improved decision-making: Big Data analytics can help organizations extract

valuable insights to take informed and data-driven decisions, leading to

improved efficiency and competitiveness (Maniam & Singh, 2020; Micheli et

al., 2020; Smith & Heffernan, 2019; Bento et al., 2022; Donald et al., 2023;

Hummel et al., 2021).

•Enhanced customer understanding: Big Data analytics processing for

example wearable IoT data or social media data can gain a better

understanding of customer behavior and preferences, therefore, better target

marketing strategies and product/services sales.

•Innovation and new opportunities: Big Data opens up innovation for example

in healthcare industry, and AI.

•Operational efficiency and cost reduction: Big Data analytics and applied AI

can detect inefficiencies in processes and product defects to enhance both

efficiency and productivity.

•Fraud detection and risk mitigation: Big Data analytics can play a good role

in detecting and preventing fraud. For example, analyzing massive volumes

of network traffic data using AI can help detect fraudulent patterns, data

breaches, and cyber attacks.

This literature review revealed that the the biggest challenges of Big Data are 1)

the lack of data governance frameworks and best practices, 2) data quality (accuracy,

trustworthiness, reliability), 3) data security and privacy. Big Data use cases are ever

growing and penetrating many industries, health care (Chen 2020; Juddoo et al., 2018),

data lakes (Derakhshannia et. Al, 2020; Hlupić et al., 2022; Ramachandran et al., 2022),

corporate sectors like retail and telecommunications (Kastouni et al., 2022; Rasheed et

al., 2022), public sector and education (Fernandes, 2020). Big Data governance

challenges range from a variety of aspects:

•Data governance challenges: lack of governance frameworks and research to

guide practitionners (Yebenes & Zorrilla, 2019; Zorrilla & Yebenes, 2022).

•Data privacy and security: collecting and processing vast amounts of data,

from users, on social media, IoT (e.g. health wearables) or retail platforms

raise concerns for out privacy and security. Also protecting against sharing

sensitive and unauthorized access, breaches, and potential misuse are not

bullet proof for Big Data sets stored on a data lake, unless strong governance

and security are put in place (Achar et al., 2022; Al-Badi et al, 2018; Almeida

Teixeira et al., 2019; Chen, 2020; Cui, 2019; Donaldet al,2023; Hirsch et al.,

2020; Kariotis, et al, 2020; Machado et al., 2022; Maniam & Singh, 2020;

Mulligan, et al., 2019; Tiffin et al., 2019; van der Sloot & Keymolen, 2022).

•Data quality and reliability: Big Data is complex and if not processed

correctly can result in poor quality (incomplete, inaccurate, or inconsistent

data). Ensuring data quality and reliability is challenging and must be

addressed (Hikmawati et al., 2021; Hirsch et al., 2020; Ibrahim et al., 2021;

Juddoo et al., 2018; Karkošková, 2023; Khatri & Brown, 2010; Mahanti,

2022; Mao et al.,2022; Ngesimaniet al., 2022; Smith & Heffernan, 2019; van

der Sloot & Keymolen, 2022; Vojvodic & Hitz, 2022; Walsh et al., 2022).

•Data integration and compatibility: integration of data from various sources

and formats is challenging and requires savvy data engineering and

datascience skills.

•Skill gap and talent shortage: need for skills and expertise in data science,

statistics, programming, domain knowledge, and AI to process Big Data, and

extract godlen data.

•Ethical considerations: ethical questions arise from the collection of massive

user data without their consent and without transparency. Business decisions

made from biased data impact users.

Big Data Governance Frameworks. While the body of literature on Big Data

Governance is growing from non-existent, there is still a research gap in terms of data

governance frameworks that can effectively address challenges related to data quality,

security, and privacy. While several conceptual dimensions have been proposed to tackle

these issues, such as those suggested by Al-Badi et al. (2018), Al-Sai et al. (2019),

Derakhshannia et al. (2020), Janssen et al. (2020), Juddoo et al. (2018), Kastouni et al.

(2022), Maniam & Singh (2020), Osu & Navarra (2022), and Rasheed et al. (2022), two

comprehensive frameworks have been put forth by Yebenes and Zorrilla (2019) and

Zorrilla and Yebenes (2022). Yebenes and Zorrilla conducted a systematic literature

review to address the lack of a Data Governance framework for Industry 4.0. In the

context of Industry 4.0, which represents the fourth industrial revolution, data serves as a

valuable business asset and a fundamental component of software architecture with

dataas-a-service (DaaS) capabilities. Their literature review focused on third-generation

platforms (3GP) and provided them with a model to propose a conceptual framework of

Data Governance for Industry 4.0. In the context of 3GP, Data Governance is defined as

“a companywide framework for assigning decision-related rights and duties to

adequately handle data as a company asset.” While traditional Data Governance

primarily deals with structured data managed and stored on-premises within IT

structures, Big Data Governance framework must account for data-centric architecture

and the requirements of the Big Data era, including cloud platforms and Big Data

technologies.

The proposed Data Governance framework by Yebenes and Zorrilla aligns with

the Industry 4.0 environment and 3GP and consists of four pillars (see Figure 6): (1)

planning, which includes strategies, objectives, principles, policies, standards, and rules,

similar to traditional Data Governance; (2) organization, encompassing decision-making

bodies, decision rights, authority, and responsibility; (3) implementation or operation,

covering processes, people, and technologies; and (4) monitoring. Building on their

previous proposal of a Data Governance conceptual framework for Industry 4.0 based on

3GP (Yebenes & Zorrilla, 2019), the authors provide a detailed design and definition of a

Data Governance reference framework for implementing Data Governance systems in

the context of Industry 4.0 (see Figure 7). This reference framework is based on TOGAF,

ISO-42010, and RAS standards, reinforcing the principles of treating data as a business

asset that requires proper governance. The new Data Governance reference framework

includes a maturity model, a reference architecture, a standards Information Base, and an

architecture development method. Yebenes and Zorrilla’s (2022) reference framework

provides a detailed foundation for implementing Big Data Governance in the context of

Industry 4.0 for enterprises.

Figure 6

Data Governance Framework for Industry 4.0 (Yebenes & Zorrilla, 2019)

Figure 7

Data Governance Reference Framework for Industry 4.0 (Zorrilla & Yebenes, 2020)

Future Research Topics in Data Governance

This study focuses on investigating the strategies utilized by IT managers in

implementing traditional data governance (DG) and emphasizes the significance of Big

Data governance as a complementary aspect. The literature review places primary

emphasis on traditional DG while also acknowledging the growing importance of Big

Data governance due to the widespread adoption of Big Data in various industries and

the emergence of the Industry 4.0 era (Yebenes & Zorilla, 2019; Al-Badi et al., 2018;

Janssen et al., 2020). Therefore, understanding IT managers strategies for Big Data

governance would be a logical extension to the current study in the field of data

governance. This literature review has identified several emerging topics and potential

areas for future research. These include datafication, data markets, data sovereignty, data

cooperatives, data sharing pools, public data trusts, data ecosystems, governance of

Artificial Intelligence, and digital platform ecosystems governance. The subsequent sub-

sections offer an overview of these topics.

Datafication. The current era of the digital economy and the upcoming Industry

4.0 are characterized by a massive influx of data and the widespread adoption of Big

Data technologies across various industries, including healthcare and retail. In their

survey, Donald et al. (2023) explored the phenomenon of datafication, referring to the

explosion of Big Data and the increasing emphasis on data-driven transformation in the

digital economy and Industry 4.0. Datafication presents both opportunities and

challenges. On the one hand, it enables innovation through AI and advanced analytics,

leading to improved revenues and profitability for companies, as well as creating new

use cases for customers. On the other hand, it brings forth challenges such as the need for

effective data governance. Without proper data governance, the potential of data-driven

technologies like AI, machine learning, and the Internet of Things (IoT) can negatively

impact user privacy and security, accentuate biases, or result in job losses due to AI

automation. Building data exchanges and data markets are key components of data-

driven transformation. These frameworks facilitate the exchange of data between

different parties, enabling them to share, collaborate, and derive insights from diverse

datasets.

Data Markets. In the data economy, data is considered a valuable asset that

exists within a market ecosystem consisting of producers, consumers, and intermediaries.

Driessen et al. (2022) conducted a comprehensive literature review to address three

primary research questions. Firstly, they aimed to provide an overview of the existing

research on the design of data markets. Secondly, they sought to identify the application

domains where data markets can be utilized. Lastly, they aimed to establish a taxonomy

of use cases for data markets along with best practices. The ultimate goal of their study

was to contribute knowledge that can be utilized by software architects and engineers in

designing data markets. A data market, in essence, is a data platform that encompasses

both infrastructure and services to facilitate the exchange of data between providers and

consumers operating in diverse environments. The study identified five primary types of

data markets that are commonly implemented in various industries: generalist, specialist,

industry exchange, enabler, and aggregator. Governance and sovereignty are key

considerations when it comes to the exchange and utilization of data produced by an

organization as an asset, especially when involving third parties and end users.

Data Sovereignty. In the digital era of unprecedented data growth, increased data

exchanges among third parties, and the rise of data-driven technologies, addressing the

question of data sovereignty becomes crucial. Hummel et al. (2021) conducted a

literature review with the primary research goal of defining data sovereignty and its

characteristics. Data sovereignty pertains to the control of data within a specific

jurisdiction. It encompasses aspects such as control, ownership, and various claims on

data assets and infrastructures. Agents can include individual consumers, organizations,

entire societies, and even countries. Data storage and processing technologies such as

data mesh have an impact on the practical aspect of data sovereignty. As a decentralized

approach to emphasize domain-oriented ownership and governance (as opposed to

traditional centralized data systems like a data lake), a data mesh has the potential to

impact data sovereignty; however, it does not inherently determine who should be

responsible for the overall governance of the data shared in the mesh. Hummel et al.

(2021) stressed the need for further research to establish a comprehensive understanding

of data sovereignty, especially in relation to emerging technologies and changing data

storage paradigms. Similar to a data marketplace, a data ecosystem offers an optimized

framework for facilitating controlled data exchanges between multiple parties.

Data Ecosystem. Building upon the exponential growth of data, this research

emphasizes the significance of data as an asset and the necessity of data governance to

effectively manage data as an ecosystem asset. Data governance involves managing data

assets within an organization and facilitating data exchanges between organizations in an

inter-organizational context. Lis and Otto (2021) aimed to identify the key characteristics

and attributes of ecosystem data governance through their research. They defined data

ecosystems as autonomous organizations that engage in data sharing to implement

business use cases and drive innovation. With the increasing volume of data and data

exchanges, data ecosystems are emerging as a prominent phenomenon. Traditional data

governance practices have predominantly focused on intra-organizational aspects. The

rise of data ecosystems necessitates the establishment of data governance frameworks to

bridge this gap. Implementing data governance in such ecosystems poses challenges due

to the inherent uncertainty arising from varying data governance practices and standards

across different parties. The study proposed a three-layer taxonomy with eight

dimensions: the data layer (including data ownership and decision rights), the

governance layer (comprising configuration, structure, and mechanism), and the

interaction layer (encompassing purpose, scope, and phase).

D’Hauwers et al. (2022) also focused on the data ecosystem model but for a

specific use case, Smart Cities, to establish a data governance framework specifically

tailored to producers and consumers of data assets within that environment. The

researchers proposed the Data Ecosystem Model Framework, which revolves around two

key factors: (1) the value of data assets, including aspects of creation, revenue, and cost

models, and (2) control, encompassing the value network for relationships and data

governance to cater to the organization’ needs. While a data ecosystem framework

encourages the perception of data as a product, it does not necessarily establish a

foundation of trust among the participating entities.

Data Cooperatives, Data Sharing Pools, Public Data Trusts. Data governance

is a significant challenge in the age of datafication, and further research is required to

address this issue. In a literature review conducted by Micheli et al. (2020), four

emerging models of data governance were identified to regulate the access, control,

sharing, and utilization of data on various platforms:

1. Data sharing pools (DSPs) where data is treated as a commodity and shared

among multiple parties to promote data-driven innovation, create new

services, and provide value-added benefits to all stakeholders.

2. Data cooperatives (DCs) where data and associated rights are distributed

among various actors, similar to DSPs.

3. Public data trusts (PDTs) involves a public entity that accesses, aggregates,

and utilizes data about its citizens, including data held by commercial entities.

PDTs establish a relationship of trust with these entities to access the

necessary data.

4. Personal data sovereignty (PDS) provides data subjects with greater control

over their data in terms of privacy management and data portability compared

to the prevailing dominant model. In PDS, data subjects are considered key

stakeholders alongside digital service providers.

The existing data governance model is primarily controlled by large high-tech

corporations. Just like the General Data Protection Regulation (GDPR) (Mulligan et al.,

2019), the Personal Data Store (PDS) model enables individuals to have authority over

the utilization, sharing, and administration of their data. This model could serve as a

foundation for addressing the governance of artificial intelligence (AI) concerning its

exploitation of end-user data.

Governing Artificial Intelligence. In the era of a maturing data economy and the

advent of Industry 4.0, companies are focusing on harnessing the potential of Big Data

and AI innovation. However, the lack of governance in this field is becoming

increasingly evident. Taeihagh’s (2021) article addressed the pressing need and

challenges of governing Artificial Intelligence (AI). While AI holds the promises of

economic efficiency and improved quality, it also poses risks that can negatively impact

individuals and society as a whole. Concerns such as privacy, bias, unemployment

resulting from automation, and the allocation of responsibility and legal liability (in the

context of human control) being transferred to robots and AI systems, are among the key

issues at hand.

Digital Platform Ecosystems Governance. In today’s information age, the

remarkable rapid rise of digital social media platforms, such as Facebook, YouTube, and

Twitter, has enabled hundreds of millions of users worldwide to produce and exchange

content. However, these platforms lack systematic data governance, leading to

significant security and privacy concerns for their billions of users. While these

platforms effectively connect users and facilitate content creation, exchange, and storage,

their exponential growth has resulted in a lack of clear governance mechanisms. To

address the data governance gap, Hein et al. (2020) proposed a definition of a digital

platform ecosystem consisting of three components: the platform owner, an ecosystem of

autonomous consumers and commentators, and value creation mechanisms. Within this

framework, Une Lee et al. (2017) identified governance challenges related to data

platforms, such as unclear content ownership between the platform owner and end users,

insufficient transparency of data flow, and ethical considerations. The field of research

on data governance for digital platform ecosystems closely aligns with other areas like

Big Data governance (Janssen et al., 2020) and cloud governance (Al-Ruithe et al.,

2018). Une Lee et al. (2017) proposed a decentralized model based on blockchain

technologies as a new approach to designing data governance for platform ecosystems.

This decentralized data governance model has the potential to address the existing

challenges and provide enhanced security, transparency, and user control.

Transition and Summary

The Foundation of the Study section laid the ground for this study by specifying

the problem statement and main research question, identifying conceptual frameworks to

guide the research design, data collection, analysis, and interpretation, and finally a

review of academic literature to support the study. Two conceptual frameworks were

retained for this study: Abraham, R., Schneider, J., & Vom Brocke, J. (2019) Data

Governance Conceptual Framework, and Khatri and Brown (2010) Unified Framework

for Data Governance. The literature review has uncovered several factors influencing the

adoption and implementation of data governance. These include considerations related to

data quality, data security, data validity, data availability, regulatory compliance,

associated costs, and the necessary skillset. Additionally, it also explored various aspects

of governance strategies such as standards, best practices, and policies.The next section,

the Project, details the research design including population and sampling criteria, data

collection and analysis, and a discussion of reliability and validy of the study.

Section 2: The Project

Purpose Statement

The purpose of this qualitative pragmatic study was to explore strategies used by

IT managers to implement DG. The study aimed at understanding the factors that help or

hinder the implementation of DG by companies. The population included IT managers in

the United States at select companies that had successfully implemented a company-wide

DG process. The implications for positive social change are improved data quality and

data security for businesses and individuals, and the democratization of DG because of

the availability of applications at a low cost.

Role of the Researcher

To ensure the integrity of this study, I needed to acknowledge and address

potential sources of personal bias that could have influenced the research process. My

role in shaping the study’s design, data collection, and data analysis was paramount. Two

crucial considerations for researchers were addressed. First, I adhered to ethical

principles outlined in the Belmont Report Protocol (see Nagai et al., 2022; U.S.

Department of Health and Human Services, n.d.) to ensure the responsible conduct of the

study. Second, I identified and mitigated personal biases to prevent their negative

influence on the research process and conclusions.

Abiding by the Belmont Report Ethical Principles

In this study, I adhered to the ethical guidelines outlined in the Belmont Report.

The primary goal of this study was to expand knowledge and address important issues.

However, research can carry risks for individuals and society if researchers do not follow

essential ethical principles. The Belmont Report, which emerged from the 1976 Belmont

Conference, established three fundamental ethical principles that all researchers must

adhere to to safeguard individuals and society from any adverse consequences of research

(Nagai et al., 2022): beneficence, respect of autonomy, and justice. The beneficence

principle emphasizes the commitment to conducting research with kindness and without

causing harm. The respect for autonomy principle ensures that participants are free from

coercion and have the autonomy to choose to participate in or withdraw from research.

Finally, the justice principle upholds the principles of justice by guaranteeing equal

protection, opportunities, and burdens for all research participants.

In the current study, I was committed to upholding ethical principles. The

primary goal of this study was to gain valuable insights into the strategies used by IT

managers when implementing DG. The overarching aim was to provide valuable

knowledge to professionals in the field of DG. This commitment aligned with the ethical

principle of benevolence guiding this study. To honor the autonomy principle, I

integrated a consent form for each participant, granting them the autonomy to freely

engage in interviews and the freedom to withdraw from the study at any point. The

concept of informed consent in research is legal and regulatory in nature (Bazzano et al.,

2021). A consent form is a requirement for any study that engages subjects.

The principle of justice was upheld in the current study through meticulous

design measures, which prevented any form of discrimination among participants. In the

United States, federal and state antidiscrimination laws prohibit any discrimination based

on race, gender, religion, age, occupation (FTC.Gov, n.d.); therefore, avoidance of

discrimination is a legal and regulatory requirement for researchers. Responsible

research with human subjects must remain ethical by design and be conducted in such a

way that it avoids discrimination of participants. These measures ensure equal

opportunities for all individuals to participate in a study. Furthermore, the principle of

justice demands a biasfree study, one that does not adversely affect either the participants

or the study’s conclusions. Survey research is particularly vulnerable to researcher bias,

stemming from the framing of questions and interpretation of responses (Story & Tait,

2019). Interviewbased research shares the same researcher bias potential. I needed to

identify personal bias early and devise strategies to address it prior to and during the

study. I identify potential personal biases in the following sections and discuss strategies

aimed at mitigating their influence throughout the research process, including the

formulation of findings and conclusions.

Addressing My Personal Bias

Personal bias, stemming from my unique perspective, was identified and

mitigated to safeguard the quality and objectivity of the study’s outcomes, particularly in

the context of qualitative research in which I was involved as the primary data collection

instrument (see Jager et al., 2020, Mazhar et al., 2021). My individual background,

encompassing cultural, experimental, and personal dimensions, included biases, values,

and perceptions that may have impacted the interpretation of collected data. In this

pragmatic study, data collection took the form of interviews conducted with IT managers

to gain insights into their strategies for implementing DG. Throughout this process,

several potential sources of personal bias needed to be considered and managed. First,

my professional background had the potential to influence my perspective. For instance,

a technical background might have inclined me to view DG as a technical endeavor

confined to the realm of IT, rather than recognizing it as an organization-wide initiative

with carefully crafted strategies. Second, my personal familiarity with the topic of DG

may have introduced bias. Preconceived notions or a deep understanding of the subject

matter may have inadvertently shaped the interview protocol and data collection process,

potentially skewing the findings.

To address these potential biases, I implemented mitigation strategies. These

strategies included transparency, reflexivity, and methodological rigor. Transparency

involved openly acknowledging my background and potential sources of bias.

Reflexivity encouraged me to continually self-assess my role and biases throughout the

research process, making adjustments as needed to maintain objectivity (see Dodgson,

2019). Methodological rigor ensured that the research design and data collection

techniques were well grounded, reducing the influence of personal bias. Recognizing and

addressing personal bias was fundamental to the success and validity of this study. By

implementing robust mitigation strategies, I was able to navigate the challenges posed by

their unique perspective, ensuring that the study yielded unbiased and reliable

conclusions.

Background Bias

A researcher’s background, such as being a senior data engineer/architect with

over a decade of experience delivering data platform solutions to Fortune 500

companies, can shape their perspective. In this study, my extensive IT background might

have led me to view DG primarily through a technical lens, focusing on technical

implementation using existing software applications or custom solutions. To address and

mitigate this potential bias, I developed a multifaceted strategy relying on adherence to

conceptual frameworks, interview protocol, member checking, and data triangulation.

First, to prevent undue technical bias, I adhered to established conceptual

frameworks: Abraham et al.’s (2019) DG conceptual framework and Khatri and Brown’s

(2010) unified framework for DG. These frameworks provided a structured and

comprehensive approach to studying DG. Second, recognizing that I played a central role

in data collection, steps were taken to minimize interviewer bias during interviews. The

interview protocol was designed to be neutral and impartial. It was designed not to lead

interviewees toward technical aspects but instead encourage responses that addressed

business and organizational dimensions.

Third, to validate data collection, I employed member-checking interviews.

Member checking, a crucial step in qualitative research, involves verifying

participantsupplied data (from interviews or surveys) to ensure the accuracy of

transcriptions (Motulsky, 2021). Lastly, to enhance the credibility of the findings,

multiple data sources and methods were used. Triangulation (combining data from

interviews, surveys, and document analysis) provided a more comprehensive and

balanced view of the research topic. Relying solely on the interpretation of interview

data may have introduced researcher bias due to the data collection protocol and the

application of conceptual frameworks (see Jager et al., 2020). To enhance the credibility

of the current study findings, I supplemented the analysis with external sources such as

public or industry reports. These additional sources offered valuable validation from the

perspectives of DG practitioners and researchers. By implementing these strategies, I

aimed to conduct a rigorous and unbiased study on DG, ensuring that my IT background

did not influence the research outcomes.

Interview Protocol

I gathered data through interviews with participants, using well-designed

questions centered on the research question and two conceptual frameworks. Bias could

have been introduced through the interview questions and my manner. According to

Jager et al. (2020), many types of biases can be introduced during data collection, which

can have an adverse impact on the internal validy of the study, including selection bias

(sampling bias, evidence prevalence bias) and information bias (interviewer bias,

observer bias, lead time bias). To mitigate bias, I adhered to best practices in question

construction and interview delivery by using a well-established and vetted interview

protocol. The primary strategy was to employ an open-minded approach during

interviews, allowing IT managers to freely express their DG strategies without

influencing their responses. First, I crafted questions that aligned with the research

question and conceptual frameworks. Second, I avoided closed-ended questions that

would have yielded yes/no answers. Finally, I refrained from using leading questions.

This strategy was encapsulated in a rigorous interview protocol that I applied uniformly

to all participants (see Appendix B).

Researcher’s World View of the Topic of DG

My perspective may have shaped my approach to the research question and the

use of existing conceptual frameworks from the body of literature. I demonstrated a

strong understanding of IT managers’ strategies for DG and the chosen conceptual

frameworks. However, there was a potential bias in favor of emphasizing the data quality

dimension of the Khatri and Brown (2010) unified framework for DG, potentially

neglecting the broader framework. To address this potential bias, I incorporated

interview questions that covered all six data governance domains and underlying

attributes within the Khatri and

Brown (2010) framework, as well as those within the Abraham et al. (2019) framework.

This approach ensured a more holistic exploration of data governance. Furthermore, I

was mindful not to oversimplify traditional DG by conflating it with MDM. Although

MDM and DG contribute to improved data quality, MDM’s primary focus is on creating

high-quality and reliable reference data for the entire organization, whereas effective DG

may encompass MDM practices in certain instances but not vice versa. To ensure I

remained objective and free from potential researcher bias, I conducted the data

collection and analysis phases with meticulous rigor. This approach helped me maintain

the integrity of the study and its findings.

Data Collection and Analysis

The interview protocol served as the tool enabling me to gather input from IT

managers regarding their DG strategies. During the data collection and analysis phase, I

realized that researcher bias may have adversely impacted the internal validity of the

study and the study’s outcomes (see Jager et al., 2020). To mitigate this, I employed

various strategies such as participant member checking, transcript validation, and data

saturation checks. The trustworthiness, auditability, credibility, and transferability

(TACT) framework for rigor in qualitative research provides guidance on how to boost

trustworthiness, auditability, credibility and transferability of a study (Daniel, 2019).

Member checking (participant or respondent validation) involved transcript validation

and review in which I sent participants their transcript and allowed them to confirm or

correct it (see Motulsky, 2021). This method is highly effective in preventing

misinterpretation or inaccurate recording during interviews. However, it may extend the

overall time required, and participants might be unwilling to allocate extra time.

Additionally, biases or errors may arise during coding and analysis. To maintain the

integrity of the study, I ensured that coded data aligned with the original raw data

obtained from the interviews. In pragmatic studies relying on interviews, participants

play a pivotal role as the primary sources of interview response data.

Participants

Effective participant selection is vital in preparing for data collection in a

qualitative study employing an interview protocol. Poor selection of participants can

introduce selection bias that will negatively impacts the internal validity of the study and

conclusions (Jager et al., 2020). I prevented selection bias by establishing well-defined

eligibility criteria for participant inclusion and outlining strategies for recruiting

participants and nurturing meaningful relationships with them.

Eligibility Criteria

This study focused on IT managers at Fortune 500 companies in the United States

with an established DG program lasting 18 months or more. To qualify, IT managers

needed to have a minimum of 1 year of employment at the company, possess expertise in

DG, and be familiar with the company’s DG program. This ensured participants’

thorough understanding of the program’s implementation within the organization.

Strategies for Gaining Access to Participants

Gaining access to IT managers within companies can be challenging, but my

strategy was designed to efficiently connect with potential participants. The approach

consisted of two key steps:

1. Identify prospects: First, I focused on identifying potential participants within

professional networks and associations. This involved searching platforms

such as LinkedIn and relevant DG groups. The keywords I used for searches

primarily related to “data governance” in job titles or job functions.

2. Solicitation outreach: Once I identified these networks and potential

participants, I initiated contact by sending a well-crafted solicitation message

explaining the purpose of the study and outlining the selection criteria. This

outreach continued until an initial sample of 36 prospects was obtained.

The strategy in this study is purposive sampling selection. For the initial phase,

the goal is not to limit the sample to a specific number but rather to collect as many

potential candidates as possible. It should be understood that the sample size might vary,

and I remained flexible in this regard. The approach is be pragmatic, and I continued

gathering candidates until a point where data saturation was achieved.

In summary, my strategy involved identifying potential participants through

professional networks and associations, utilizing relevant keywords, and initiating

outreach to build a diverse initial sample. The precise sample size was determined

moving along, with a focus on achieving data saturation or a manageable sample size,

possibly around 12 participants.

Strategies for Establishing a Working Relationship With Participants

Before commencing the interview process, it is essential to establish a productive

rapport with potential participants. To accomplish this, I employed effective networking

techniques using platforms such as LinkedIn or send an email containing a concise

overview of the study, participant requirements, eligibility criteria, a request for

informed consent, and a commitment to upholding ethical standards (Johnson et al.,

2020) to ensure the well-being of both participants and their respective organizations.

Following the initial outreach, I sent an email to individuals who had expressed their

intent to participate in the study, along with the consent form. Once the prospective

participant signed and returned the consent form, indicating their willingness to

participate in the research, I collaborated with them to schedule interview dates and

times. Subsequently, I sent a scheduling invitation and awaited confirmation, while also

providing additional follow-ups and reminders as necessary. In cases where a

prospective participant did not responded, I gently remind them with one to three

reminders. When there was still no response, I discontinued contact. These follow-ups

were conducted via email or through LinkedIn chat. Another strategy to maintain a

working relationship with the participant was to extend a connection invitation on

LinkedIn and engage in discussions related to data governance to establish a shared

rapport on the subject. This interaction continued until after the member-check follow-up

interview was complete.

Research Method and Design

Method

In the “Nature of the Study” section, I chose a qualitative research method,

specifically a pragmatic inquiry, as the most suitable approach for this study. Unlike

quantitative research, which relies on systematic empirical methods to examine

phenomena through numerical data collection and analysis, qualitative research takes an

inductive approach, offering the advantage of delving deeply into the subject matter,

thereby facilitating the generation of fresh insights and theories (Lima &

NewellMcLymont, 2021). This research method is exceptionally well-suited for gaining

a comprehensive understanding of the data governance strategies utilized by IT

managers. In this study, we have opted for a qualitative approach with an exploratory

nature to investigate the “what,” “why,” and “how” behind the data governance strategies

implemented by IT managers.

While mixed-method research, which combines both quantitative and qualitative

elements, could potentially provide quantitative validation for the findings obtained

through qualitative inquiry (Lo et al., 2020), it was not considered in this study due to

time and resource constraints within the scope of our research. Quantitative research,

however, could be explored in a different context with a distinct research question. For

instance, it could be utilized to validate conclusions derived from a prior qualitative

study on data governance strategies. Such a quantitative study would be amenable to

generalization based on statistical sampling and analytical methods.

Research Design

Under the pragmatic inquiry approach, an exploratory multiple case study design

has been adopted, which facilitates a thorough investigation using participant interviews

as data collection method. Importantly, it provides the flexibility to study one or multiple

bounded cases without external interference, as highlighted by Takahashi et al (2020)

study. In this study, each company’s implementation of data governance constitutes an

individual case study, with one IT manager participating in the interview for each case.

The pragmatic inquiry design has been selected as the most appropriate approach

for this study due to its emphasis on gaining a profound understanding of the data

governance strategies employed by IT managers. This design, operating within the

pragmatic framework, facilitates a comprehensive investigation through various data

collection methods such as observations, interviews, and secondary data analysis. It

grants the freedom to explore specific cases or multiple instances without external

interference, as outlined by Takahashi et al. (2020). In this study, the objective is to

comprehend data governance strategies by engaging in interviews with IT managers

from various organizations, making it inherently a multiple case study.

Alternative research designs, such as ethnographic and phenomenological

approaches, were not chosen because they do not align with the primary research focus

on IT managers’ strategies in the realm of data engineering. Ethnographic research

excels in uncovering shared patterns within cultural groups (Gherardi, 2019), a

perspective that does not apply to the specific goals of this study. Likewise, the

phenomenological approach concentrates on individual life experiences (Greening,

2019), making it unsuitable for investigating data engineering topics like data

governance.

Data saturation is a pivotal concept in qualitative research, particularly in

methodologies such as grounded theory, ethnography, content analysis, and multiple case

studies. It signifies the point at which gathering additional data no longer yields novel

insights or information pertinent to the research questions or objectives (Shaheen &

Pradhan, 2019; Fusch & Ness, 2015; Sebele-Mpofu, 2020). This indicates that I have

acquired a sufficient amount of data to comprehensively comprehend the phenomenon

under investigation. The process can be summarized as follows, as depicted in Figure 8:

1. Initial Participant Selection: Commence with a pool of 36 prospective

participants.

2. Interview and Transcription: Select one participant and schedule a live

interview. Following the interview, transcribe it meticulously.

3. Member-Checking Review: Arrange a member-checking review with the

interviewee after transcribing the interview.

4. Theme Extraction and Comparison: Analyze the transcript to identify and

extract themes. Compare these themes with the existing list of themes in the

research topic.

5. Decision-Making Point: Make a decision based on the comparison. If new

themes are identified, repeat the process by selecting another participant from

the list. If no new themes are discovered and the number of processed

candidates is less than 12, continue the process. If data saturation is achieved

(i.e., no new themes emerge, and at least 12 candidates have been processed),

terminate the data collection process.

Figure 8

Data Saturation Process

The decision to involve a minimum of 12 participants for a multiple case study

was made to ensure an adequately sized sample for comprehensive analysis. Detailed

considerations regarding data saturation are elaborated upon in the Population and

Sampling section of this study.

Population and Sampling

Population

This study focused on IT managers within Fortune 500 companies in the United

States. We aim to assemble a final sample with a minimum of 12 participants, each

representing a different company. To be eligible, IT managers were required to possess a

minimum one-year tenure with their respective companies, demonstrate expertise in data

governance, and have a deep understanding of their company’s data governance

program. Additionally, a pre-requisite for the company is that it must have a well-

established data governance program in place for at least 18 months. In the Participants

section, we outlined the approach to identify potential participants, who were primarily

sourced from professional networks like LinkedIn and Data Governance groups. We

initiated contact by sending out a solicitation, and upon their response, followed up with

a formal email containing a consent form. This form was essential, as it ensured their

voluntary participation in our data collection interviews. To build a rapport with

participants, I employed strategies such as establishing a LinkedIn connection or

engaging in professional discussions related to data governance topics. These

interactions helped create a comfortable working relationship. Subsequently, I

coordinated with each participant to arrange a live interview at a mutually convenient

time. These interviews were conducted using platforms Microsoft Teams for a seamless

experience. After the interview sessions, I transcribed the content and scheduled a

member-checking interview with each participant. This step aims was to validate the

accuracy of the transcripts against their initial responses, further enhancing the overall

data quality (Motulsky, 2021) Sampling

The study employed purposive sampling with a size of 12, as recommended for

qualitative research or until data saturation is achieved (Shaheen & Pradhan, 2019).

Purposive sampling enhances data validity and trustworthiness. Nevertheless, it is

essential to mitigate sampling bias (Johnson et al., 2020) and ensure sample diversity to

adequately address the research question and achieve data saturation. I interviewed 11

participants, each with over ten years of experience in data governance across various

industries, including retail, pharmaceutical, travel and hospitality, healthcare, and finance

(see Table 6).

Table 6

Participants and Industries

Participant DG experience in industries

Participant 1 IT, health

Participant 2 IT, pharmaceutical

Participant 3 Banking, financial services

Participant 4 IT

Participant 5 Oil and gas, tech real-estate marketplace, health insurance

Participant 6 Pharmaceuticals

Participant 7 IT

Participant 8 Pharmaceuticals

Participant 9 Manufacturing

Participant 10 Financial services and banking, manufacturing, insurance, retail

Participant 11 Travel technology, tech real-estate marketplace

I initially chose purposive sampling in my research design to select target

participants from LinkedIn, ensuring a good mix of industries. However, enlisting

volunteers proved challenging as potential participants did not commit. To address this, I

switched to snowball sampling, starting with one or two participants, interviewing them,

and obtaining additional participants through their referrals. Snowball sampling is a

nonprobability technique often used in qualitative research, particularly with hard-to-

reach or hidden populations, relying on referrals to generate new subjects. This method

was highly effective. Once a participant accepted the invitation, I had them agree to a

consent form sent via email. Subsequently, I conducted semi-structured interviews over

Microsoft Teams to gather data on their experiences implementing Data Governance

programs, following the interview protocol I designed. To enhance data reliability, I

followed the interviews with member-checking. Data collection was followed by data

analysis using the classic Data Analysis Spiral strategy in qualitative research.

I employed the Data Analysis Spiral strategy for data analysis, which involved

the following steps: (1) Managing and organizing data, (2) Reading in-depth and

memoing emerging ideas, (3) Describing and classifying codes into themes, (4)

Developing and assessing interpretations, (5) Representing and visualizing data, (6)

Providing an account of the findings.

Data Saturation

The objective of data saturation in qualitative research is to ensure the precision

and validity of the data, analogous to the statistical validity of a sample in quantitative

research. Given the pragmatic nature of our study (multiple case study), I opted for

purposive sampling, as previously discussed. It is worth noting that although an ideal

sample size of no more than 12 is recommended for this type of study, caution must be

exercised to avoid both excessively small and overly large samples. Due to the absence

of clear guidelines for sampling and generalizing findings in qualitative research, the

concept of data saturation has emerged as a widely accepted tool among researchers to

enhance the rigor, quality, and trustworthiness of their findings.

As data saturation can be assessed in various ways, Sebele-Mpofu (2020)

suggests that researchers incorporate additional criteria into their study to confirm the

attainment of data saturation. Fusch and Ness (2015) have presented indicators for

verifying data saturation, including the absence of new data, absence of new themes,

absence of new coding, and the ability to replicate the study with new participants from

the same population and time frame. In qualitative research, the quality of data is of

paramount importance compared to the quantity of data, underscoring the need for rich

and insightful information. To ensure data quality, our strategy involves targeting an

initial population of 36 companies (three times the recommended sample size of 12).

Subsequently, i conducted interviews until the optimal sample size of 12 participants or

data saturation was achieved, whichever came first. In the data collection, data saturation

occurred after 11 participants were interviewed. This approach was designed to enhance

the overall quality and reliability of the findings of this research.

Ethical Research

Responsible academic research including a doctoral study must obey to ethical

principles. Estalblished in 1976, the Belmont protocol constitutes a reference for ethical

requirements that must govern research in order to protect individuals and society from

adverse consequences (Nagai et al., 2022). The Belmont protocol rests on three core

principles: beneficence, respect of autonomy, and justice. These standards encompass

principles such as voluntary participation, the right to withdraw, informed consent, as

well as the protection of privacy and confidentiality for all participants. This study

conforms to those principles.

Firstly, Ethical research requires prioritizing benevolence, ensuring that the

study’s design and execution do not harm or bias participants. Stringent ethical standards

are essential to safeguard the well-being of individuals and organizations. The study

employed a robust data collection strategy to guarantee the confidentiality and

anonymity of participants and their affiliated companies (Saunders et al., 2015).

Interview transcripts featured coded participant names and their respective companies for

enhanced organization and confidentiality. Company identities and any confidential or

proprietary information remains undisclosed, except when such information is already in

the public domain, as in reports and industry publications.

Ensuring the ethical protection of research participants is a paramount

responsibility for any researcher. In this study, I ensured the appropriate ethical

protection of participants by following a rigorous process. First and foremost, I sought

ethical approval from the Walden Institutional Review Board (IRB) before commencing

any research involving human participants. The IRB thoroughly reviewed my research

proposal to ensure it aligns with ethical standards and granted an approval (03-19-

241059331). This is a mandatory step for doctoral student researchers before conducting

interviews or other data collection activities with participants. Furthermore, I relied on

my dissertation chair and committee members to meticulously assess my study design.

They helped me identify and address any gaps related to the ethical protection of

participants in my research. Lastly, the informed consent form used in this study

emphasized the voluntary nature of participation and the right to withdraw at any time.

This not only respects participants’ autonomy but also contributes to the ethical

protection of participants in the research. In summary, safeguarding the ethical protection

of research participants is a multi-step process that involves obtaining IRB approval,

seeking input from advisors, and emphasizing voluntary participation in the consent

form.

Secondly, the principle of autonomy was satisfied in that, for this study, each

participating IT manager was presented with a consent form before the interview with

me. In qualitative research, a consent form serves as the essential instrument that

empowers participants by granting them the freedom to willingly engage in a study or

discontinue their participation at any point without facing any form of pressure or

negative repercussions (Nagai et al., 2022). This consent form is designed to ensure the

safeguarding of the rights and well-being of not only me and participant but also Walden

University. All participants were required to sign the Participant Consent Form. The

consent form also indicated that participation to the interviews is voluntary and

participants could withdraw anytime they wished. To initiate withdrawal from the study,

participants would follow these steps. First, when during the interview, participants

could simply notify me of their decision to withdraw. In response, I would promptly halt

the interview and proceed to delete all recorded data. Otherwise, if a participant wished

to withdraw after the live interview, they would have to submit a written withdrawal

request to me. Once I received this request, I would take the necessary steps to delete the

participant’s interview raw data and records from the transcript. Furthermore, no

coercion was exerted in participants in any shape or form to participate in the study.

Participants received no form of incentive, whether financial or otherwise, for their

participation in the study. The Interview Protocol (see Appendix B) reinforced some of

these aspects.

Thirdly, to respond to the justice principle, ethical considerations were upheld

throughout the study, particularly in participant selection (Johnson et al., 2020). To

ensure ethical standards were met, the following steps were taken:

1. Privacy and Confidentiality: The study strictly adhered to company privacy

and confidentiality regulations, as well as respect individual privacy and

freedoms. Consent was obtained from each participant prior to conducting

interviews (see Appendix B). Also the interview was confidential, that is,

data collected was not shared with third party, and the identities of

participants and their companies were revealed in the analysis and

conclusions of the study. Ensuring privacy and confidentiality was imperative

throughout the entire research data lifecycle. Consequently, I explicitly

apprised interview participants and informed the eventual readers of the study

that all raw interview data will be securely stored for a duration of five years,

employing either a cloud drive or the Walden file server.

2. Avoidance of Sensitive Information: The survey carefully avoided delving

into sensitive or confidential company information. I provided a guarantee

that both company and individual privacy were to be rigorously maintained.

Participant and company names were coded and omitted from the transcripts

3. Non-Discrimination: Participant selection did not discriminate based on

gender, religious affiliation, or age, ensuring a fair and unbiased

representation of the workforce. From a legal and regulatory perspective, the

United States requires non-discrimination as mandated by both state and

federal governments (FTC.gov, n.d.).

4. Minimum Tenure Requirement: Participants were required to have a

minimum one-year tenure at the company. This ensures that participants

possessed sufficient knowledge about the DG programs implemented within

the company.

5. Relationship Building: Prior to conducting interviews, an effort was made to

establish a positive and respectful rapport with the selected participants.

Building a strong researcher-participant relationship fosters trust and open

communication (Knott et al., 2022).

By implementing these ethical guidelines, the study upheld at best the highest

ethical standards and ensured the rights and privacy of all participants were respected

throughout the research process.

Data Collection

Effective research relies on solid data as evidence to address primary research

inquiries. The data collection phase is a pivotal step in the research process. During this

phase, researchers employ suitable instruments, strategies, and techniques to gather data

while also implementing methods to organize the collected information. The credibility

of qualitative research hinges on several key factors: a well-defined research question

supported by a conceptual framework, the choice of an appropriate research

methodology, my self-awareness and consideration of their own biases, and

meticulousness in both data collection and analysis (Johnson et al., 2020). Boosting the

trustworthiness of a study involves adhering to best practices, utilizing computer

software for data collection and analysis, engaging in peer reviews, conducting audits,

implementing triangulation, and exploring alternative case analyses. The following

sections delve into a detailed discussion of instruments, collection techniques, and data

organization strategies.

Instruments

The aim of this research was to comprehensively comprehend the strategies

employed by IT managers in implementing data governance within institutions that have

already adopted this process. This exploration delved deep into these strategies, with the

overarching objective of recognizing both challenges and opportunities in this context. In

terms of data collection instruments, it was essential that they align closely with the

chosen research method and design. Accordingly, for this study, semi-structured

interviews were selected as the primary data collection instrument (please refer to the

Interview/Survey Questions section). I developed an interview protocol document (refer

to Appendix B) that served as the consistent framework for interviewing all participants.

I refered to this protocol while conducting interviews with participants. This approach

ensured uniformity in wording and constructs, effectively mitigating any potential bias or

variations in the interview administration.

Semi-structured interviews are well-suited for open-ended inquiries and are

intended to probe participants’ opinions, beliefs, and insights related to the phenomenon

under investigation (Adeoye‐Olatunde & Olenik, 2021). The interview questions were

thoughtfully crafted to align with the core research topic and incorporate elements from

both Khatri and Brown’s (2010) Unified Framework for Data Governance and Abraham,

R., Schneider, J., & Vom Brocke’s (2019) Data Governance Conceptual Framework. By

embedding dimensions from these frameworks into the interview questions, this study

aimed to shed light on IT managers’ motivations and strategies for implementing data

governance through the lens of these comprehensive frameworks.

To enhance data reliability and accuracy, this study incorporated a

memberchecking phase. During this phase, I held a live meeting with the participant to

carefully review interview transcripts to confirm their fidelity to the original interviews

(Motulsky, 2021). This approach mitigates potential researcher bias and minimizes the

risk of misinformation introduced during the transcription of audio recordings into

summarized

text.

Data Collection Technique

For this study, I conducted semi-structured interviews with experienced

individuals responsible for data governance programs in US-based Fortune 500

companies, each with over ten years of relevant expertise. Semi-structured interviews

serve as a data collection method that bridges the gap between structured and

unstructured interviews. They offer the benefit of adaptability and in-depth exploration

of topics, but are accompanied by challenges pertaining to time, training, analysis, and

potential bias. One notable advantage is their flexibility in questioning (Ruslin et al.,

2022). Interviewers can tailor the interview protocol questions based on the participant’s

responses and adjust the format to suit different participants. This flexibility fosters a

more comfortable environment for participants, encouraging richer and more

comprehensive responses that yield valuable insights into the research phenomenon.

Additionally, semi-structured interviews provide an abundance of rich data. Participants

can thoroughly explore their responses, offering a profound insight into complex

socialbehavioral research inquiries. This rich dataset can be further enhanced by

triangulating it with other data sources, such as surveys, to gain a more comprehensive

understanding of the research topic. However, despite these advantages, semi-structured

interviews do have their drawbacks. One common concern is the potential for

interviewer bias, which is inherent in most qualitative research involving human

participants. Furthermore, they can be time and resource-intensive both in preparation

and delivery. Their open-ended nature can result in response variability, posing

challenges in drawing clear and consistent conclusions. Finally, conducting a

comprehensive analysis of semi-structured interview data can be intricate, necessitating

expertise in identifying themes and extracting meaningful insights to ensure the

credibility and validity of the findings. In the following paragraph, I provide a concise

overview of my approach to conducting semi-structured interviews with specific

participants.

Authorization from the Walden University Institutional Review Board (IRB) was

a prerequisite for conducting these interviews. The interviews were conducted using the

Microsoft Teams video-conferencing platform, allowing both audio and video recording

capabilities. My initial interaction involved introducing myself and explaining the

interview’s purpose, emphasizing the importance of safeguarding participants’

information. Although I intended to pre-screen participants before the interviews, I

reconfirmed their seniority within the company and their experience in data governance

during the interview. These qualifying questions served as a foundation for exploring

core topics, including the factors driving data governance implementation, challenges,

emerging trends, and opportunities. To maintain research rigor and minimize potential

researcher bias, I formulated open-ended questions, fostering participant input without

leading or adding excessive context. To enhance data reliability, I employed clarifying

follow-up questions and take detailed notes during the interviews.

To improve data reliability and accuracy, this study included a member-checking

phase. In this phase, I conducted a live meeting with the participant to thoroughly

examine interview transcripts and verify their alignment with the original interviews

(Motulsky, 2021).

Data Organization Techniques

Efficiently organizing text data is a crucial step in the data collection process,

setting the foundation for effective data analysis. To uphold the principles of reliability,

validity, security, and traceability in collected data, it is imperative for researchers to

methodically catalog and store the information within a database (Johnson & Chauvin,

2020). This requirement aligns with the guidelines outlined in Walden University’s

Doctoral Study guidelines. My approach involved utilizing Turboscribe software, a

reliable online audio-to-text transcription software, to convert the Microsoft Teams audio

recording of the interview into a text file. Teams offered an audio-to-text transcription

feature, but I preferred Turboscribe. Subsequently, I meticulously reviewed and made

necessary edits to enhance the accuracy and clarity of the transcription. Furthernore, I

conducted a member-checking interview with participants to boost the data accuracy

(Motulsky, 2021). Finally, I ensured the safekeeping of all raw data ( the original

interview audio files and the edited transcripts) by storing them encrypted in my

Microsoft OneDrive secure cloud drive location for 5 years. This systematic approach

not only ensures the integrity of the data but also facilitates seamless access and analysis

when conducting my research.

Data Analysis Technique

Data analysis is a pivotal research stage where qualitative data is systematically

processed, transforming it into meaningful insights. Qualitative research poses unique

challenges in data analysis due to its open-ended nature. Thematic analysis is a

commonly preferred method for qualitative research. Castleberry and Nolen (2018) have

outlined a five-step methodology for conducting thematic analysis, which includes:

compiling data, disassembling data, reassembling themes,

interpreting themes, and concluding the analysis. These phases are

interdependent and collectively serve as the backbone of qualitative research, facilitating

the extraction of valuable insights and understanding from the collected data. This study

employed Castleberry and Nolen’s

(2018) methodology for conducting a thematic analysis of the collected data.

Data Compilation

The initial phase involved the compilation of qualitative data collected. This

process can be likened to piecing together a puzzle, where each data point, interview

transcript, or observation is meticulously organized and structured. This preliminary

organization serves as the cornerstone for subsequent analytical stages. During this

phase, I utilized Turboscribe audio-to-text transcription software to transcribe the raw

interview recordings into a readable text format. Following this transcription, I

thoroughly reviewed it, enabling me as a researcher to gain a deeper understanding of the

phrases and meanings extracted from the transcripts. This step remains incomplete until

the member-checking interview is conducted. The purpose of member-checking is to

enhance data quality (Motulsky, 2021), as it involves validating the transcribed responses

with the participant to ensure fidelity to the original conversation. The outcome of this

phase is a consistent and well-organized transcript, which served as the input for the next

phase of analysis.

Data Disassembly

Disassembling data involves utilizing coding techniques to discern and organize

meaningful groupings within the transcribed interview text by identifying themes,

concepts, and ideas. This coding process serves as a fundamental aspect of qualitative

research, characterized by its inductive approach, as opposed to the deductive approach

employed in quantitative research. In this qualitative endeavor, researchers permit

meaning to naturally emerge from the data. A crucial step in this process is the careful

consideration of a coding strategy before initiating the analysis (Castleberry & Nolen,

2018). These codes can often be drawn from existing research on the topic found within

the body of literature. For this particular study, I constructed codes based on the data

governance domains and dimensions outlined in both Khatri and Brown’s (2010) Unified

Framework for Data Governance and Abraham, R., Schneider, J., & Vom Brocke’s

(2019) Data Governance Conceptual Framework. This analysis can be done either

manually or with software. For this study, I used the manual approach, with in-depth

reading of transcripts, noting codes, and grouping into themes. The advantage of the

manual approach is that it does not require the cost of expensive software. The

disadvantage is that it is cumbersome

The alternative approach to facilitate this analysis, NVivo software can be used,

complemented by manual processes as needed. NVivo can assist in processing the text

file, which is derived from the the audio-to-text transcription of the interview (Allsop et

al., 2022). This software can aid in code generation and the identification of emerging

themes. An example of thematic coding with NVivo software is shown in Figure 9.

Codes are clustered together in the text data to form groups. The size of nodes represents

the frequency of each code, while colors represent code families that can be grouped into

themes. Subsequently, emerging themes cqna be organized into thematic groups.

Furthermore, it’s essential to perform manual adjustments to ensure that these themes are

in alignment with the primary research question, this study’s conceptual frameworks, and

the established academic knowledge on data governance.

Figure 9

Example of Thematic Analysis Coding With NVivo Software

Data Reassembly

Data reassembly, a pivotal phase in the research process, involves the

consolidation of coded data into cohesive themes and narratives. This step serves as the

foundation for the subsequent thematic interpretation. During this phase, I established

connections among coded segments to uncover overarching insights and derived

meaningful conclusions (Castleberry & Nolen, 2018). Furthermore, meticulous

documentation of codes, code-to-theme mappings, and theme descriptions was essential

at this stage.There are two primary methods for mapping code to themes: hierarchies and

matrices. Thematic hierarchies involve creating visual graphs by clustering codes,

generating hierarchies with varying levels of granularity. On the other hand, matrices are

constructed by organizing participant roles, themes, variables, and emerging concepts in

rows and columns.

Thematic hierarchies are the more commonly employed approach for mapping

codes to themes and establishing relationships among themes. However, it’s important to

note that reassembling these hierarchies requires analytical skills on my part to justify the

relationships and hierarchy of themes.

Eight major themes emerged from data analysis as follows, in no particular order

of importance:

1. Data Quality, Security/Privacy, and Regulatory Compliance: These are the

main drivers for Data Governance (DG) programs.

2. New Drivers: Data discoverability, data cataloging, and automation are

emerging as significant motivators for DG.

3. Third-Party Data Assets: These should be treated with the same rigor as

internal data to preserve security and privacy.

4. Data as an Asset: While most companies recognize the value of data, few

treat it as an asset unless it impacts the bottom line, either as a revenue

generator (e.g., data as a product, enabling AI innovation) or as a liability

(e.g., risk, compliance).

5. Executive Sponsorship: Vital for DG success. Without it, there is no

organizational alignment, authority, or budget support, leading to program

failure.

6. Value Justification: One of the biggest challenges is justifying the value of

DG upfront to both executive and lower management.

7. Data Competencies: DG success requires relevant competencies for all DG

roles.

8. Targeted DG Programs: DG programs should be targeted, use case-bound,

and antecedent-driven, avoiding cumbersome, large-scale initiatives

undertaken without a clear purpose.

Interpreting

Data interpretation is a crucial stage in the research process, where researchers

draw conclusions based on their analysis of codes and derived themes. To ensure a

highquality interpretation, there are five key qualities that researchers should aim to

achieve (Castleberry & Nolen, 2018):

1. Clarity: The interpretation should be presented in a clear and logical fashion,

demonstrating how it was derived from the data.

2. Reproducibility: The interpretations should be reproducible when using the

same dataset, ensuring that others can arrive at similar conclusions.

3. Fidelity to Raw Data: It is essential to maintain fidelity to the original raw

data, ensuring that interpretations are grounded in the data without distortion.

4. Knowledge Enhancement: Interpretations should contribute to a deeper

understanding of the subject matter, adding valuable insights and knowledge

to the research field.

5. External Validity and Trustworthiness: Interpretations should be robust

enough to withstand external scrutiny by other researchers, ensuring their

validity and trustworthiness.

Thematic interpretation provides the bridge between raw data and the research’s

ultimate objectives. The themes extracted in the previous step are now refined and

ranked, keeping those that best align with the research question at the forefront of this

study. Table 7 provides an illustration of the themes that have surfaced through code

analysis.

Table 7

Example of Themes Emerging From Keywords

Domain Keyword, phrase Theme (as

framework

dimension)

Topics in data

governance

Definition, architectures, objectives Domain scope

Industries Multiple industries: finance, healthcare,

retail, highly regulated industries

Domain scope

Business drivers efficiency, accuracy, transparency,

regulatory compliance, save time and

money, cleaner data, data security

Antecedents

Data principles

Strategies policies, data access policies and

control, who/what, use cases,

standards, classification (sensitive,

nonsensitive, PII), data documentation,

metadata management (ingestion,

certification) cross-functional, adjacent

to IT

Data access

Data Lifecycle

Governance

mechanisms

Organization Who is responsible, all, exec sponsor C-

level, stand-alone lead, crossfunctional,

strategic vision

Organizational

scope

Challenges Leading challenges

Organizational:

- convincing organizations that it is

worth it;

- strategy first then technology

Regulatory compliance

Costs

- high

- applications are prohibitively

expensive Technical

applications and solutions not

automated

Data quality Data

scope

Antecedents

Trends Key future trends of data governance

- regulatory compliance

- big data security and governance

commoditization of governance

capabilities: established platforms

packaging data governance within

their

own suite of capabilities

Consequences

As part of this interpretive process, I constructed a thematic map that provides a

detailed representation of the identified themes and their interrelationships. These themes

were linked to the main research topic and the dimensions of data governance within the

conceptual framework. This visual representation aids in conveying the complex

relationships between themes and their relevance to the broader research context.

For this study, triangulation was used to boost the validity of the interpretation

and findings of the study. Triangulation is a valuable technique in the qualitative data

analysis process, serving to enhance both external validity and the trustworthiness of

research results (Farquhar et al., 2020; Noble & Heale, 2019). It is widely recognized as

an effective approach to bolstering the validity of a study, whether it is qualitative,

quantitative, or a mixed-methods study (Farquhar et al., 2020). This is particularly

relevant in the context of case study research, such as the one conducted in this study,

where triangulation plays a crucial role in attaining a higher level of validity and

confidence in the findings. Triangulation in research involves the utilization of various

methods, theories, or data sources to gain a more comprehensive understanding or

explanation of the same phenomenon. The core concept underlying triangulation is

corroboration or convergence, which means that these different sources, methods, or

theories should lead to the same conclusions regarding the observed phenomenon.

Consequently, triangulation serves to enhance both internal and external validity.

There are tree types of triangulation approaches. Firstly, data source triangulation

is a commonly employed technique in case study research (Farquhar et al., 2020). It

involves gathering data from various sources, such as conducting interviews with

multiple participants, observing the phenomenon at different times, and incorporating a

combination of primary and secondary sources derived from publicly available

publications and other research materials. Secondly, researcher triangulation is another

valuable method in case study research. In this approach, two or more researchers

independently investigate the same phenomenon to arrive at similar interpretations and

conclusions, thus enhancing the reliability of the findings. Lastly, theoretical

triangulation is employed to enrich understanding. This technique involves examining

the same set of data from different theoretical perspectives, which can provide valuable

insights and a more comprehensive understanding of the subject under investigation.

Concluding

The interpretive phase involves extracting themes and creating thematic maps

from the coded data. These themes are then thoroughly examined to derive meaningful

interpretations. The ultimate step in the data analysis process is drawing conclusions,

where these interpretations are formulated to address the primary research question,

considering the conceptual frameworks underpinning the study. In this particular study,

the conclusions were tailored to provide insights into the strategies employed by IT

managers for the implementation of data governance programs. These conclusions were

validated for alignment with the data governance domains and dimensions specified in

both Khatri and Brown’s (2010) Unified Framework for Data Governance and Abraham,

R., Schneider, J., & Vom Brocke’s (2019) Data Governance Conceptual Framework.

Reliability and Validity

In qualitative research, ensuring the reliability and validity of data is crucial to

establish trustworthiness and to generate meaningful findings that contribute to the

research body and address the main research question. Daniel (2019) has introduced a

framework consisting of four key dimensions known as Trustworthiness, Auditability,

Credibility, and Transferability (TACT) to guide student researchers in achieving these

goals. In the subsequent sections, I will delve into the reliability and validity aspects of

this research, along with the strategies I employed to ensure these crucial elements were

safeguarded.

Reliability

Reliability is a critical metric evaluating the resilience and robustness of both the

research instrument and the overall study. Essentially, it examines the consistency and

stability of research findings (Cichy & Rass, 2019). In quantitative studies, assessing

reliability is relatively straightforward as it measures the repeatability of the study’s

findings. In other words, it gauges whether the same results would hold when applied to

a different population sample with similar traits. However, in the realm of qualitative

research, determining study reliability can be more complex and, at times, less

meaningful (Cichy & Rass, 2019). This challenge is especially pronounced in research

methodologies like case studies and ethnographic studies, which are tailored to specific

populations with unique characteristics. Consequently, expecting identical outcomes

when these methodologies are applied to different populations may not be realistic.

For a qualitative study, the concept of reliability takes on several dimensions.

First is internal consistency, that is the consistency of the measurement methods.

Secondly, inter-coder consistency in the interpretation of results when more than one

researcher is involved. Third, the audit-trail to maintain an organized record of data

collection, and analayis processes. Finally, triangulation by using multiple sources to

enhance reliability of the study.

Consistent Data Collection

In this study, prioritizing the reliability of our instruments was paramount, and I

rigorously upheld this during the interview process. This involved a standardization of

data collection procedures and protocols with a meticulous design of interview questions

and a comprehensive interview protocol (see Appendix B), as detailed in the Data

Collection section. Achieving a consistent data collection also involves confirmability,

that is, to achieve objectivity and neutrality of research findings. As discussed in the

Role of the Researcher section, reflexivity is a common strategy, that I have used in this

study for reducing researcher bias in data collection (Dodgson, 2019).

Member Checking

Additionally, this study emphasizes the importance of fidelity in transcribing

interview responses to guarantee the accuracy of the participant’s original statements. To

ensure this accuracy, I validated the interview transcripts with each participant in an

interview follow up with a member-checking interview (Motulsky, 2021). This was an

opportunity for participants to validate the accuracy of their contributions to ensure their

perspectives and represented without a researcher bias. This was discussed in the Data

Collection section.

Dependability and Credibility

This aspect measures the reliability of the data as it pertains to both the internal

validity and external validity. For this purpose, the study employed data triangulation

(Farquhar et al., 2020). I employed multiple data sources, including participant interview

transcripts, public domain literature and white papers and methods to cross-verify

findings, reducing the risk of bias and increasing the dependability of results. The

following six secondary sources were considered for data triangulation: Atlanta Regional

Commission (2019), Biggenden (2022), ESG (2022), Nephos Technologies (2022),

OvalEdge Team (2022).

Validity

Validity is a critical gauge of a research or study’s credibility as it measures the

accuracy and truthfulness of research findings. Validity encompasses two key

dimensions: internal and external validity (Daniel, 2019). Internal validity ensures that

the study’s outcomes consistently address the primary research questions or hypotheses.

Meanwhile, external validity ensures that the study’s results align with established

knowledge in the field, findings from analogous studies conducted by other researchers,

and industry norms.

In this study, I assessed external validity by utilizing conceptual frameworks and

critical success factors for data governance (Alhassan et al., 2019; Almeida et al., 2019;

Alsousi & Shah, 2022; Bento et al., 2022). Specifically, the study aimed to determine if

the data governance strategies and motivations of IT managers align with the decision

domains outlined by Khatri and Brown (2010) as well as Abraham et al. (2019) in their

conceptual frameworks for data governance. The focus was on ensuring that the

conclusions drawn from this study harmonize with the antecedents decision domain of

the conceptual frameworks when it comes to motivations. Additionally, the researcher

intended to align the findings on challenges and strategies with the organizational

dimensions of these conceptual frameworks. Furthermore, conclusions regarding

strategies should be consistent with the established success factors and best practices for

data governance (Alhassan et al., 2019; Almeida et al., 2019; Alsousi & Shah, 2022;

Bento et al., 2022), as documented in our literature review. It’s essential to acknowledge

that this study takes an inductive approach rather than a deductive one, implying that it

does not anticipate a perfect alignment with the existing frameworks. Instead, the

primary objective was to uncover insights into the data governance strategies employed

by IT managers and contribute new knowledge to the existing body of literature in this

field.

In this study, internal validity was assessed by evaluating how effectively the

findings address the primary research question. This involved uncovering plausible

motivations (antecedents) behind IT managers’ decisions to implement data governance

in their respective companies, along with the associated strategies. Additionally, internal

validity was gauged by confirming that data saturation (Sebele-Mpofu, 2020; Fusch &

Ness, 2015) is achieved within the retained participant sample. To enhance the accuracy

and fidelity of data collection, member-checking (Motulsky, 2021), was conducted

whereby each participant confirmed that the transcript was faithful to their original

responses to the interview with me. Finally the TACT framework (Daniel, 2019) was

leveraged to gauge credibility and transferability this study.

Confirmability

The rigor of qualitative research hinges on four essential criteria of the Lincoln

and Guba’s Framework for trustworthiness research (Devakirubai, 2020): credibility,

confirmability, transferability, and dependability. Confirmability assesses the extent to

which the conclusions of a qualitative study are firmly grounded in the collected data and

the subsequent analysis, as opposed to being influenced by the researcher’s personal

biases, values, or preconceptions. In this study, I ensured confirmability by maintaining

transparency throughout the research process. This will was achieved by meticulously

documenting every step, including data collection, analysis, and data archiving. The

appendices contain comprehensive documentation of these procedures, encompassing the

participant interview protocol, member-checking protocol, participant consent form, and

records of the code and themes extraction.

Transferability

Transferability in qualitative research assesses the applicability of study insights

and conclusions to similar contexts, populations, or settings. Unlike quantitative

research, qualitative research, such as multiple case studies, is inherently context-specific

and typically lacks generalizability (Johnson et al., 2020; Devakirubai, 2020; Dodgson,

2019). To enhance the transferability of this study, my approach was based on three key

pillars. Firstly, by anchoring it in the particular domains and dimensions of established

theoretical frameworks, such as Abraham, R., Schneider, J., & Vom Brocke’s (2019)

Data Governance Conceptual Framework and Khatri and Brown’s (2010) Unified

Framework for Data Governance, this research employs concepts that serve as a

foundation for making our findings more applicable to broader contexts. Secondly, to

enhance the study’s transferability, I employed reflexivity to acknowledge and mitigate

my personal biases as a researcher (Dodgson, 2019). Lastly, my data collection approach

involved a diverse range of participants from various companies, spanning different sizes

and cultures. This approach ensured that the findings were not confined to a single

company or location in the United States. Furthermore, meticulous documentation of the

data collection and analysis promotes transferability through transparency and

confirmability.

Transition and Summary

As the study design, Section 2 offered a comprehensive overview of the overall

research design. It encompassed the selection of the research method and specific design,

participant criteria, sampling methods, data saturation, data collection and analysis

procedures, as well as an exploration of validity and reliability. The study employed an

exploratory multiple case study approach with a pragmatic design and purposive

sampling of participants. These participants consisted of IT managers with expertise in

data governance, having a minimum tenure of one year, and employed at Fortune 500

companies in the United States with a functioning data governance program in place for

at least 18 months. Ethical considerations are of utmost importance, and I trictly adhered

to the Belmont Protocol to ensure the protection of both participants and their respective

institutions. I also addressed reflexivity by identifying personal bias and discussed

strategies on how to mitigate their adverse impact on the internal validity and

conclusions of the study.

Data collection was executed through carefully structured online interviews with

participants, following a consent form process. Additionally, member-checking

interviews were conducted as a follow-up to enhance the quality of the collected data.

Raw data collected from interviews was be kept encrypted, secure and confidential using

my Microsoft OneDrive cloud storage. Qualitative data analysis methods were applied,

including the compilation of interview data, transcription of interviews, and the

incorporation of member-checking to bolster data quality. This process culminated in the

identification of key themes that formed the basis for drawing conclusions. To maintain

the study’s reliability and validity, it adhered closely to the TACT Framework (Daniel,

2019).

Section 3 explores the presentation of study findings derived from the research

design, data collection, and analysis outlined in Section 2. It involves a comprehensive

discussion of the study’s findings, accompanied by recommendations that elucidate their

relevance to the broader field of data governance. The conclusions drawn from this

research endeavor pinpoint the strategies discerned for the implementation of data

governance programs by IT managers within American companies. This discussion also

encompasses the challenges encountered in this process. Furthermore, Section 3 delves

into the implications of these findings for the professional practice of data governance. It

also explores the potential ramifications for driving social change through improved data

governance practices. Additionally, this section identifies limitations within the study,

gaps in existing knowledge, and suggests areas for future research that can further

enhance the insights gained from this study.

Section 3: Application to Professional Practice and Implications for Change

This study focused on exploring strategies employed by IT managers to

implement successful DG programs. In this section, I present the findings from data

collection and analysis. Additionally, I discuss how these findings contribute to the

existing body of knowledge in the DG field and their implications for social change.

Finally, I suggest potential future research directions and conclude with my personal

reflections as the researcher.

Overview of Study

This qualitative pragmatic study aimed to explore strategies IT managers use to

implement DG programs, focusing on factors that facilitate or hinder these

implementations. Primary data were collected through semistructured interviews with

mid-level IT managers, including those in director-level roles, who had experience in

DG across various industries. Secondary data from organizational documents and

publicly available data from governance reports were also used for triangulation. This

study revealed several key findings and relationships among them that influence and

guide the strategies to implement DG:

1. Driving factors for DG: Data quality, security, privacy, and compliance are

the top needs driving data governance. Additionally, data discoverability,

data cataloging, and data governance automation have emerged as new

drivers.

2. Data as an asset: Companies acknowledge the importance of data, but its

value is best demonstrated when it impacts the bottom line as a revenue

generator, such as through innovation.

3. Third-party data assets: These should be governed with the same rigor as

internal data sources but with established standards for quality and privacy

protection.

4. Executive support: Executive support and sponsorship are crucial for the

success of DG programs. However, justifying the value of these programs to

executives and middle management remains a significant challenge.

5. New CSFs: The study identified two new success factors for data governance

programs:

•Ensuring data competencies for all roles driving the DG program.

•DG programs should have a clear strategy, be targeted, use case bound,

and be antecedent driven to achieve success (see Figure 10).

Figure 10

Major Themes and Relationships

Presentation of the Findings

The research question of this study was the following: What strategies do IT

managers use to implement data governance programs in practice? The aim was to

uncover insights that would contribute to the body of knowledge on DG and provide

valuable information for practitioners in the field. This section presents the findings from

my research, organized into eight identified themes.

Theme 1: Effective Data Governance Strategies Must Address the Challenges of

Data Quality, Security/Privacy, and Regulatory Compliance

One theme emerging from data analysis was that data quality, security/privacy,

and regulatory compliance are the main drivers for a successful DG implementation.

Most interviews with participants highlighted that these factors represent both essential

needs and significant challenges for organizations, directly affecting the overall

effectiveness of DG. As a result, IT managers must develop effective strategies to

address these challenges and meet these needs.

Data quality encompasses characteristics that enhance the value and trust

worthiness of data; these include accuracy, completeness, reliability, and relevance.

Companies and organizations need high-quality data to make informed business

decisions and often mission-critical ones. For regulated industries such as health care and

financial, high data quality is necessary for compliance with laws and standards. Poor

data quality can lead to of high costs and risks that adversely impact business

performance and the bottom line. Data security is concerned with protecting data from

unauthorized access, corruption, and theft throughtout the life cycle; this includes aspects

of data access, privacy, authorization, and authentication. Protection of sensitive data

such as personally identifiable information (PII), company financials, and intellectual

property falls in this category. Data breaches can result in significant financial losses due

to fines, legal fees, and the cost of remediation. Data privacy, on the other hand, is

concerned with proper handling of data (collection, storage, and sharing) ensuring that it

is used in a manner that complies with laws and allowing individuals to control how their

personal data is used and shared. Finally, businesses and organizations are required by

law to meet regulatory compliance or face financial penalties. Strict regulations exist that

govern certain industries as is the case of HIPAA for the health care industry, EU GDPR

and the California Consumer Privacy Act (CCPA) for consumer data privacy in retail

and other industries.

All 11 interview partitipants indicated either data quality, security/privacy, or

regulatory compliance as the main drivers (or antecedents) for DG programs.

Participants shared views from experiences in various industries including regulated

(health care, financial), retail, oil and gas, transportation, hospitality, and real estate. Data

quality driver was a common denominator among all participants. Participant 1 shared

data quality concerns that most businesses have with inaccurate data impacting usability

and trustworthiness. This participant also spoke of the importance of data security and

privacy to comply with GDPR and CCPA regulations; this view was shared by

Participant 5 with regard to HIPAA regulation for the health care industry. Participant 4

and Participant 7 shared a similar view that governing the data is usually data quality,

data security, and data privacy.

Furthermore, in the era of AI, companies are honing in on innovation. Participant

7 mentioned the need for high data quality and trustworthiness to feed into the large

language models used by Gen AI and other AI tools. Participant 2, Participant 6, and

Participant 8, from pharmaceuticals, also shared data quality issues as the main concern,

but added that businesses need to discover where data resides and know the content of

data assets and lineage sources to provide an intimate knowledge of the data. All this is

needed to pass the various compliance audits. Developing a data strategy to understand

where the organization wants to go and how the organization wants to use data becomes

critical. For the financial and banking industries, Participant 3 and Participant 10 insisted

on regulatory compliance as the main driver for data governance besides data quality,

data security, and privacy. The Financial Industry Regulatory Authority has a mandatory

audit of financial institutions to show the lineage of data for all transactions. Data quality

is important for accuracy and trust of the transactions. Security is paramaount to avoid

data breaches, which could allow identity theft leading to litigation and regulatory

penalties.

Interview participants suggested various strategies to enhance data governance,

focusing on data quality, security and privacy, and regulatory compliance. For data

quality, Participant 1 proposed a governance strategy centered on ranking the business

importance of data assets. This involves cross-functional engagement to identify which

data is most critical to the business. Participant 8 also emphasized improving data quality

by governing access, monitoring usage, and democratizing data across the company,

ensuring that data is made available in a centralized location. Data lineage was

highlighted by Participant 3, particularly in the context of regulatory audits within the

banking sector. Participant 3 stressed the importance of understanding where data

originates because this is foundational to any further data quality initiatives. Establishing

data lineage is crucial before addressing other aspects of data governance. For

information security, the strategy involves classifying data to determine its level of

sensitivity, as mentioned by Participant 4. This includes setting access controls and

adhering to data privacy regulations, which are especially stringent in regulated

industries such as financial services, health care, and life sciences. Participant 4 also

noted the importance of these strategies in avoiding costly fines and ensuring compliance

with regulations such as HIPAA, Federal Deposit Insurance Corporation, and Securities

Exchange Commission. Finally, the overarching strategy for regulatory compliance

involves understanding the organization’s data management goals and setting measurable

targets for data quality, security, and utilization. This approach ensures that data

governance aligns with the organization’s broader objectives and regulatory

requirements.

Methodological triangulation was achieved with organizational documents and

reports fully supporting this theme with a 100% consistency across all four sources (see

Table 8). In an effort to enhance transit data interoperability, quality, sharing, and

costeffectiveness, the Atlanta Regional Commission (2019) conducted research to

identify barriers to data governance. The study identified two primary challenges that a

robust data governance framework could resolve. First, inconsistencies in data structures,

formats, and semantics among regional transit authorities hindered data quality and

standardization, making data integration and comprehension difficult across systems.

Second, issues related to data security and privacy arose due to a lack of clear data

ownership and rights, complicating access and sharing. These challenges underscored the

need for implementing data governance to address the issues faced by the Atlanta

Regional Commission.

Table 8

Frequency of First Major Theme

Source of data collection n

Interviews 11

Documents 4

Note. Theme 1: Effective data governance strategies must address the challenges of data

quality, security/privacy, and regulatory compliance; n = frequency.

The Enterprise Strategy Group (ESG, 2022) surveyed 220 business and IT

professionals in data governance roles to understand how organizations implement DG,

focusing on the key drivers and challenges. The research identified data quality as the

leading factor in several areas: (a) how organizations define data governance (47%), (b)

the primary challenge in maximizing the return on DG efforts (45%), and (c) the top

driver for DG programs (41%). Recognizing the increasing adoption of DG by

enterprises, Nephos Technologies (2022) conducted a study to identify the needs and

challenges driving DG programs. The findings align with Theme 1, highlighting the

following key needs: (a) improved data quality, (b) enhanced data analytics, (c)

strengthened data security and privacy, (d) regulatory compliance, (e) minimized

financial risk, (f) increased efficiencies and cost savings, and (g) faster access to relevant

data. Needs A and B correspond with the data quality driver, Need C with data security

and privacy, and Need D with regulatory compliance. Needs E, F, and G are typically

achieved as positive outcomes when DG effectively addresses the top three drivers.

OvalEdge Team (2022) identified five key applications of DG across industries:

data privacy compliance, data discovery and literacy enhancement, centralized data

access management, creation of a standardized business term repository, and

collaborative analytics for new data product development. The data privacy use case

aligns with the security and privacy drivers discussed in Theme 1. The report noted that

regulations such as GDPR, CCPA, and International Association of Privacy

Professionals aim to protect consumer PII but have also imposed significant costs and

litigation risks on organizations. However, Gartner (2022) predicted that by 2024, 75%

of global user data will be protected by privacy regulations, with large organizations

budgeting $2.5 million or more to comply.

The extensive literature on DG reinforces Theme 1’s findings, identifying data

quality, data security and privacy, and regulatory compliance as the primary drivers for

data governance programs. Walsh et al. (2022) identified the data principles, data access,

and data quality components of Khatri and Brown’s (2010) DG framework as key

organizational motivations for implementing DG. For data principles, the primary

motivation was to enhance compliance with external regulations and internal business

and data processes. The goal of implementing data quality was to establish data

credibility and trustworthiness by ensuring accuracy, relevance, and accessibility for

users. The data access domain focused on security and privacy, aiming to restrict data

access to authorized users, monitor access to prevent unauthorized use, and protect data

assets, especially in cloud environments. Mahanti (2022) emphasized that improved data

quality and governance, particularly in terms of accuracy and credibility, play a crucial

role in helping companies achieve regulatory compliance and avoid accounting scandals

and financial disasters. Hikmawati et al. (2021) similarly emphasized the importance of

enhancing data quality, governance, and access through the implementation of MDM.

This Theme 1 aligns with the DG theories outlined by the conceptual frameworks

used in this study: Khatri and Brown’s (2010) unified Framework for DG and Abraham

et al.’s (2019)DG conceptual Framework. Khatri and Brown’s framework identifies five

decision domains critical for successful data governance: data principles, data quality,

metadata, data access, and data life cycle. Data principles, which include value addition,

compliance, performance, and decision rights, represent the business motivations for data

governance. Abraham et al.’s antecedents dimension aligns with these data principles.

The data quality domain focuses on the accuracy, timeliness, completeness,

trustworthiness, and credibility of data, which are essential for reliable decision making

and efficient operations. The data access domain addresses data protection and access

control, including defining access levels to meet internal and external compliance

requirements, essentially covering data security, privacy, and regulatory compliance.

From this analysis, the findings of Theme 1—specifically data quality, data

security, privacy, and regulatory compliance as key drivers of data governance programs

—align closely with the Data Principles, Data Quality, and Data Access domains in

Khatri and Brown’s (2010) framework. It is important to note that while the conceptual

framework positions these domains as justifications and guides for implementing data

governance programs, it does not necessarily rank them as top drivers. To ensure

effective data governance, a comprehensive program should address all relevant data

within an organization, focusing on the key domains of Data Quality, Data

Access, Metadata, and Data Lifecycle. A critical challenge and necessity in this process is

the ability to discover and catalog all data sets, ensuring they have a unified meaning

across the organization.

Theme 2: Data Discoverability, Data Cataloging, and Automation Are Emerging as

Significant Motivators for Successful DG Implementation

Another theme that emerged from the data analysis was data discoverability, data

cataloguing, and data governance automation. Data governance strategies must address

these needs. Data discoverability is the ease with which users can find, access, and

understand the data they need. This aspect of data management is crucial, particularly in

large organizations or systems with vast amounts of data in various formats and

sensitivity levels. In such situations, there’s a high likelihood of dark data—information

that organizations collect, process, and store without using or analyzing. This data holds

no value and can even pose a compliance risk. Effective data discoverability enables

quick access to the right data, leading to better decision-making, more efficient

operations, and improved outcomes. Achieving data discoverability requires high data

quality and robust metadata management to provide essential information about the data,

such as its type, format, source, and sensitivity. Automation tools are also necessary to

search, index, and organize data, allowing users to locate and interpret it effectively. This

process often results in the creation of a data catalog—a centralized repository that helps

users discover and understand available data within an organization, including metadata,

data lineage, and other relevant details. However, challenges such as data silos, large

data volumes, complexity, and the need to balance security and privacy—especially in

highly regulated industries like healthcare and finance—can complicate data

discoverability.

Improved discoverability not only facilitates better data management but also supports

regulatory compliance by ensuring data traceability and enabling usage audits.

Improving data discoverability involves a combination of technology, processes, and

governance practices, all aimed at making data easier to find, access, and use.

All participants emphasized the critical need for data discoverability and the

implementation of a data catalog to enhance data management, governance, and

democratization across organizations. Participants from the pharmaceutical industry

(Participants 2 and 8) highlighted the growing challenge of locating data stored in

various repositories, which is essential for operationalizing that data. They noted that

data discoverability and cataloging have become vital capabilities for companies

managing and governing their data assets. The participants stressed that, beyond just

discovering data, organizations must establish a company-wide data catalog. This catalog

would serve as a comprehensive data dictionary, providing meaning and, more

importantly, ensuring data lineage necessary for compliance and audits in regulated

industries. Participants 6 and 8 further emphasized that a data catalog, coupled with

governance automation, enhances data democratization. Participants from the financial

and banking sectors (Participants 3 and 10) underscored the importance of traceability—

tracking data lineage to show its sources, transformations, and final destinations. They

also noted that data discovery aids in classifying sensitive and PII data. Participants from

IT, manufacturing, and hospitality industries (Participants 1, 4, 7, 9, and 11) agreed that

the need for data discoverability is growing across companies, with data catalogs

becoming a priority.

They particularly stressed the importance of data dictionaries and, above all, data

lineage.

Participants 4 and 11 emphasized the need for automating data governance rules

and processes. As organizations manage increasing volumes and complexity of data

across various systems and platforms, automation becomes crucial (Nadal et al, 2022).

Manual governance methods are inadequate for scaling due to several challenges and

opportunities. First, the sheer volume and complexity of data, sourced from databases,

cloud services, and IoT devices, make it difficult to enforce consistent governance

policies manually. Secondly, regulatory compliance demands, such as GDPR and CCPA,

require organizations to maintain strict data privacy and security standards. Automation

ensures these policies are consistently applied, helping organizations stay compliant. In

highly regulated industries, like banking and finance, audits are a necessity. Automated

systems enhance tracking and reporting capabilities, making it easier for organizations to

demonstrate compliance during audits.

Most interviews with participants revealed that companies struggle with dark data

and a lack of centralized data definitions. These challenges underscore the need for data

discoverability and a data catalog to enhance data governance. In the financial sector,

data lineage is a regulatory requirement for auditing financial data. Implementing a data

catalog with lineage capabilities is a strategy that not only strengthens data governance

but also ensures regulatory compliance. Dark data, when left unused, can pose a liability

from both regulatory and security perspectives. Furthermore, the absence of a data

catalog prevents users across the company from accessing a unified understanding of

data, hindering the democratization of data and data skills. Consequently, IT managers

must develop effective strategies to implement data discoverability and a comprehensive

data catalog with lineage capabilities. Participants 1 and 3 emphasize a strategy focused

on implementing data discovery, data lineage, data classification, and a company-wide

data catalog as part of the data management lifecycle. They underscore the importance of

establishing a business glossary or data dictionary that spans from the business level to

individual applications, extending up to the enterprise level. Although essential, this

aspect of data management has often been deprioritized in their experiences. They argue

that understanding what data you have, where it resides, and its nature is crucial, as it

enables other functions like data lineage and classification to be effective. Data lineage,

in particular, is vital in regulated industries such as healthcare and finance. Data

classification, which includes categorizing information as highly classified, internal, or

public, is typically handled by information security groups. Participant 4 also supports

the implementation of data lineage and a data catalog but highlights the significant

challenges companies face in executing these strategies. The lack of automation, coupled

with the sheer volume of data and various drivers, makes data governance complex and,

at times, constrained. To overcome these challenges, Participant 4 suggests adopting the

right tools and technologies to facilitate cataloging, lineage tracking, metadata

management, and policy enforcement. Automation plays a critical role in reducing

manual effort, which, in turn, contributes to the success of data governance initiatives.

Additionally, transparent reporting on data governance metrics is crucial to

demonstrating the value of data quality efforts.

Table 9

Frequency of Second Major Theme

Source of data collection n

Interviews 11

Documents 4

Note. Theme 2: Data discoverability, data cataloging, and automation are emerging as

significant motivators for successful DG implementation; n = frequency.

Methodological triangulation was achieved for Theme 2, with all (4 out of 4)

articles and reports affirming the necessity of data discoverability and data catalogs for

companies. The Atlanta Regional Commission (2019) highlighted the lack of data

discovery—specifically, knowing who possesses what data—as a significant challenge

for agencies when sharing data with stakeholders. To address this, a data governance

framework was proposed to establish protocols and rules that enhance data sharing

efficiency by adopting data discovery services for cross-organizational data searches.

ESG (2022) identified key challenges in data management and governance, including

visibility and data quality issues. The study revealed that 42% of respondents reported at

least half of their data as “dark,” while 46% expressed a lack of confidence in the quality

of source data, hindering effective data utilization across their value chains. Additionally,

35% of respondents cited the lack of automation as a barrier to implementing data

governance. Overall challenges summed up to cataloging for data discovery (26%),

access control (24%), and visibility (23%). The OvalEdge Team (2022) identified that

the absence of data discovery tools hindered companies from identifying critical

information within large datasets, impeding the development of new data products.

Additionally, the lack of a centralized data catalog containing standardized business

terms made it difficult for non-technical users to analyze data, contributing to low data

literacy across the organization, data business terms, made it difficult for non-technical

people to analyze data, contribution to low data literacy across the company.

This Theme 2 aligns with the Metadata and Data Lifecycle dimensions of the

Data Governance theories outlined by Khatri and Brown’s (2010) Unified Framework

for Data Governance used in this study. The metadata domain involves capturing and

managing data descriptions. Metadata provides essential context and understanding of

data elements, facilitating effective data governance practices and enabling efficient data

discovery and utilization. Data catalog and lineage maps to the Master Data Management

and traceability aspects of the Data Lifecycle domain. Furthermore, the concept of data

discoverability aligns closely with Khatri and Brown’s (2010) emphasis on data quality

and stewardship. Ensuring that data is easily discoverable within an organization

enhances its quality by enabling better data usage and decision-making. Data catalogs

can be seen as a practical implementation of the Unified Framework’s principles,

particularly in relation to decision rights and data stewardship. By cataloging data assets,

organizations can better manage who has access to what data, enforce accountability, and

maintain data quality standards.

Theme 3: Successful DG Strategies Treat Third-Party Data Assets With the Same

Rigor as Internal Data Assets to Preserve Security and Privacy

The analysis revealed a key theme: from a governance standpoint, third-party

data should be managed with the same rigor as internal data assets, with additional

checks to ensure security and privacy. Businesses and organizations generate data

internally but also rely on external sources and providers. Third-party data, collected by

entities without direct relationships with the individuals or businesses to whom the data

pertains, is typically aggregated from various external sources. This data is widely used

in digital marketing, advertising, and analytics to enhance targeting, audience

segmentation, and personalization efforts. Other sectors, such as healthcare,

pharmaceuticals, and finance, also contribute to third-party data sources. Therefore,

third-party data should be included in an organization’s Data Governance programs and

processes. From a data management perspective, it’s essential to establish quality

standards and ensure privacy protection through Governance, Risk, and Compliance

(GRC) practices, with a particular focus on maintaining security and privacy.

All interview participants agree that third-party data sets should be treated as

another data source from a data governance perspective, with additional checks to ensure

contractual security and privacy. Participant 1 emphasized the importance of

understanding the standards, obligations, and applicable regulations when onboarding

third-party data, particularly regarding contractual usage rights. Many third-party data

sets, especially in regulated industries like healthcare and finance, come certified for

quality and completeness, such as SOC 1 (financial controls) and SOC 2 (availability,

security, processing integrity, confidentiality, and privacy) standards in the banking

sector. Participant 2 highlighted the need for thorough data governance quality checks to

ensure accuracy, cleanliness, and completeness. Participant 3, drawing on banking sector

experience, agreed with Participant 2 and stressed that the acquiring company is

responsible for the governance of any data it uses, whether internal or third-party.

Therefore, additional checks, including data quality and data movement checks, are

necessary. Participant 4 echoed Participant 3’s views, emphasizing that once third-party

data enters an ecosystem, the company assumes full liability for any breaches, security

issues, privacy concerns, and compliance. Participant 6 stressed the importance of

including checks to remove redundancy, while Participant 7 noted that third-party data

often contains sensitive information, such as PII, and that proper anonymization

techniques must be established. For HIPAA-regulated data, data protection and handling

standards must be strictly followed to ensure privacy and confidentiality. Participant 9

stated that “third-party governance is an integral part of data security and privacy

governance.” Participant 11 further discussed the importance of a structured process for

handling third-party data, emphasizing the need for close collaboration with other

stakeholders who have expertise in legal, privacy, and IT aspects. This process involves

three layers: legal (understanding legal implications and contractual obligations), privacy

(ensuring compliance with GDPR, CCPA, etc.), and IT (handling data engineering and

analytics based on expectations set by the legal and privacy teams).

Methodological triangulation for Theme 3 was not achieved, as none of the

articles or reports specifically addressed third-party data sets in the context of data

governance.

Table 10

Frequency of Third Major Theme

Source of data collection n

Interviews 11

Documents 0

Note. Theme 3: Successful DG strategies treat third-party data assets with the same rigor

as internal data assets to preserve security and privacy; n = frequency.

Theme 3 focused on the governance of third-party data, which requires

heightened checks on security and privacy, more so than data quality. In Theme 1, I

extensively discussed the three key drivers for data governance. Similar to Theme 1,

findings from Theme 3—particularly regarding data quality, security, and privacy—

closely align with the Data Principles, Data Quality, and Data Access domains in Khatri

and Brown’s (2010) framework. Additionally, there is a process element concerning how

organizations manage third-party data to meet contractual obligations, involving

collaboration with data owners, stewards, and legal, privacy, and IT departments. This

process aspect corresponds with the Data Policies component of the Governance

Mechanisms dimension in Abraham, Schneider, and Vom Brocke’s (2019) Conceptual

Framework for Data Governance, which includes formal guidelines for managing,

accessing, and utilizing data across the organization, with specific rules on privacy,

security, and quality.

Theme 4: Successful DG Strategies Tie the Value of Data as an Asset to Its Impact

on the Bottom Line

Most companies recognize the value of data, but traditionally, they only treat it as

an asset when it impacts the bottom line—either as a revenue generator (e.g., data

products or AI-driven innovation) or as a liability (e.g., risk and compliance). However,

with the rise of AI, companies are increasingly leveraging data to drive innovation and

enhance value. Governance strategfies in the AI era focus on innovation with two

aspects. First, companies focusing on AI innovation understand the critical need for data

quality, security, and privacy, and are increasingly recognizing the importance of data

governance (DG). Secondly, the growing significance of DG is particularly evident in

the context of big data.

The participants collectively recognize the growing importance of treating data as

a strategic asset within organizations, yet they highlight different perspectives and

challenges in achieving this. The participants agree on the importance of data as an asset,

but they emphasize the need for robust governance, cultural shifts, and a holistic

approach to fully realize its value in driving business transformation and innovation.

Participant 1 emphasized the evolving recognition of data as a valuable asset, with

companies increasingly exploring ways to monetize it. Three main approaches include

cost reduction through internal improvements, selling or licensing data, and enhancing

products through AI-driven insights. Participant 2 pointed out that while many

companies acknowledge the value of data, cultural and governance challenges often

prevent them from treating it as a true asset. Companies that directly monetize data

naturally view it as an asset, but others struggle due to fragmented data silos and

ineffective governance, often leading to missed opportunities.

Participant 3 highlighted that data governance is a relatively recent field and

underscores the necessity of understanding and ensuring data quality to genuinely

consider it an asset. The value of data as an asset is closely tied to its quality and

relevance. Participant 4 distinguished between organizations that merely claim data as an

asset and those that actively leverage it to transform their business. A truly data-driven

organization uses data to enhance efficiency, drive growth, and improve products and

services. Participant 5 noted that historically, data was not treated as an asset, but the rise

of AI has shifted this perception. The AI boom has made companies more aware of

data’s value, although many still treat data as an afterthought, addressing issues

reactively rather than strategically. Participant 6 reinforced this shift in perception,

noting that companies are now recognizing data as a valuable asset due to factors like

regulatory fines, advanced analytics, and AI advancements. Participant 11 emphasized

the need to understand data as a 360-degree asset, considering not only its production but

also its usage and consumption. Effective data governance requires a comprehensive

approach, recognizing the interconnectedness of data within an organization, rather than

viewing it in isolated parts.

Methodological triangulation was achieved with a single organizational

documents and reports fully supporting this theme. Data is a critical asset for any

business and is essential to a wide range of users across the enterprise, from executives

and line-of-business teams to IT decision-makers and staff (ESG, 2022). Many

organizations face challenges with data quality, including issues related to accessibility

and visibility, and often have large amounts of unutilized “dark” data. This underscores

the need for robust data governance solutions that offer transparency across the data

landscape, integrated with business context.

Table 11

Frequency of Fourth Major Theme

Source of data collection n

Interviews 11

Documents 2

Note. Theme 4: Successful DG strategies tie the value of data as an asset to its impact on

the bottom line; n = frequency.

The extensive literature on Data Governance reinforces Theme 4’s findings,

which position data as a strategic organizational asset. Khatri and Brown (2010) define

data as a strategic enterprise asset, emphasizing the need for governance to establish

accountability and regulate access and processes related to data (Walsh et al., 2022).

Similarly, Abraham et al. (2019) view data as a strategic asset, advocating for a

crossfunctional governance framework that specifies decision rights, formalizes policies,

and monitors compliance. Both frameworks underscore the necessity of applying

governance to maximize the value of data. Nielsen et al. (2017) reviewed 62 peer-

reviewed publications and conference proceedings, defining data governance as

company-wide processes that align decision-making rights and responsibilities with

organizational goals for effective data management. Satar (2021) highlights the critical

importance of Master Data due to its valuable organizational insights, ranking it as a top

priority for management. Yebeness and Zorilla (2019) also emphasize that effective

governance is essential for maximizing data’s value as a key organizational asset.

Theme 5: Executive Sponsorship Is Critical for Successful DG Implementation

Executive support, particularly an executive sponsor, is crucial for the success of

data governance (DG) programs. Without it, there is a lack of organizational alignment,

authority, and budget support, leading to program failure. Gaining executive support and

their buy in is one of the top challenges for DG programs. IT managers must identify an

executive sponsor and develop strategies to secure their backing. One effective strategy,

as revealed by the analysis, is to demonstrate the value of DG through use-case-driven

implementation. This can involve starting with small-scale programs or proof of concept

(POC) with modest budgets to showcase the impact and value of DG, particularly its

influence on AI innovation.

All participants agree that an executive sponsor is crucial for the success of a data

governance program. Without executive backing, even well-designed plans may falter

due to insufficient budget and resources (Participant 10). The Nephos Technologies

(2022) report also highlights the lack of executive support as a significant barrier to

adopting data governance programs. According to the report, while 98% of respondents

have ongoing governance initiatives, 33% feel they do not receive adequate executive

support. Participants discussed various strategies for gaining executive sponsor buy-in

for data governance:

1. Pilot Projects and Proof of Concepts: Starting with a pilot project or

smallscale proof of concept (POC) can effectively demonstrate the value of

data governance (DG) to executive leadership. Using key performance

indicators

(KPIs) for reporting can further illustrate the benefits and secure support

(Participants 1, 3).

2. Company-Wide Integration: Although data governance ideally should be a

company-wide initiative supported by executive leadership and embedded in

the company culture, it is often siloed with departments driving use cases

specific to their needs (Participant 2).

3. C-Suite Involvement: Securing buy-in from C-suite executives is crucial.

Data governance should not be solely IT-driven; it requires strong support

and involvement from top leadership. An executive sponsor or champion can

address challenges and advocate for data governance across the company

(Participants 4, 5).

4. Socialization and Incentivization: Having an executive champion can help

build belief in data governance through initiatives like roadshows,

newsletters, and gamification. These low-cost efforts aim to socialize and

incentivize data governance, making it more engaging for the business

(Participant 5).

5. Adoption Across Business Units: For data governance and related initiatives

(e.g., data quality, cataloging) to succeed, they need adoption across all

business units. Grassroots efforts alone are insufficient without executive

support (Participant 6).

6. Define clear Roles in Data Governance:

•Executive Sponsor : Champions the overall program

•Data Owner: Accountable for data.

•Data Steward: Manages data on a daily basis.

•Data Custodian: Handles technical aspects and system-specific

responsibilities.

•Data User: Anyone who interacts with data.

•Steering Committee: Provides oversight, similar to a legislative body.

•Data Governance Officer and Committee: Offers strategic guidance and

oversight (Participants 6, 9).

Table 12

Frequency of Fifth Major Theme

Source of data collection n

Interviews 11

Documents 2

The extensive literature on Data Governance reinforces Theme 5’s findings,

identifying the criticality of executive sponsor for the success of data governance

programs. Data governance requires the active involvement and commitment of all staff,

with robust support from both management and senior-level executives (Al-Ruithe et al.,

2019). Yebenes and Zorilla (2019) define the executive sponsor as the individual or

department responsible for driving, funding, supervising, and guiding the data

governance (DG) initiative. The executive sponsor also clarifies the initiative’s scope,

establishes milestones and goals, and ensures compliance with all relevant data laws and

regulations. Satar (2021) highlights top management support as a critical factor for the

success of Master Data Quality programs, emphasizing that top management must

recognize the importance of master data quality and endorse related management

activities. Abraham, Schneider, and Vom Brocke’s (2019) Conceptual Framework for

Data Governance categorizes the executive sponsor and other roles within the

Governance Mechanisms dimension. This framework identifies structural governance

mechanisms that define roles, responsibilities, and accountabilities for data governance.

Key roles and governance bodies include the executive sponsor, data governance leader,

data owner, data steward, data governance council, data governance office, data

producer, and data consumer. The executive sponsor plays a crucial role in providing

strategic direction, prioritizing business needs, and securing funding for data

management initiatives.

Theme 6: Effective Data Governance Strategies Must Address the Challenge of

Justifying the Value of DG Programs

The analysis highlighted a significant challenge for IT managers: justifying the

value of Data Governance (DG) programs to both executive leadership and business unit

managers. This challenge primarily stems from a lack of understanding among key

stakeholders. Executive leaders and lower management often struggle to see the

immediate ROI of DG, especially when it isn’t directly tied to compliance issues or risks

like data breaches. They’re concerned about the program’s length and potential to slow

down operations. Middle managers and end users are often even less familiar with DG.

To overcome this, IT managers must implement strategies to secure buy-in from both

executive leadership and mid-level management. Interview participants recommended

several approaches, including using agile methodologies for DG processes, focusing on

specific use cases rather than trying to address everything at once, building proofs of

concept before full implementation, and educating the entire organization on the benefits

of DG—such as improved data quality, security, and privacy. Additionally, employing

KPIs to measure and report on DG’s impact can effectively demonstrate its value. In

conclusion, reducing risks related to data security and privacy while ensuring regulatory

compliance is essential. Defining a KPI matrix to evaluate the value of data governance

(DG), along with smaller, targeted implementations, can help IT managers secure the

necessary support from both executive and middle management.

Five participants emphasized the challenge of justifying the value of DG to

organizations. Participant 4 noted that many companies treat DG as an afterthought to

their existing data strategy and transformation efforts, making it difficult to gain

executive and business buy-in. Successful DG implementation often requires extensive

collaboration with business stakeholders and IT, as well as ensuring projects stay within

budget and on schedule. A recommended strategy is to start small, perhaps with a proof

of concept, and demonstrate value through measurement and monitoring. Similarly,

Participant 6 pointed out that the challenge often lies in how DG is perceived. Many

view it as dull or overly regulatory, disconnected from tangible business outcomes. It can

be difficult to convey the value DG brings until stakeholders see concrete results. To

address this, IT managers should focus on using business language rather than DG

jargon. Instead of saying “let’s implement data governance,” they should frame it as

improving processes, saving time and money, and automating tasks. Additionally,

avoiding a textbook approach to DG and tailoring it to specific contexts and use cases

can further enhance its perceived value.

Participant 7 also noted that people often fail to understand the value of data

governance (DG) until they see quick wins. When presenting DG to executives,

discussions typically revolve around the practice’s involvement of people, processes, and

technologies. However, executives often ask for the vision and roadmap, and when told

that full implementation might take 12 months, it’s a hard sell—especially when

justifying the time, budget, and resources required. Executives usually evaluate DG’s

ROI against the company’s bottom line and are concerned about the length and potential

slowdowns associated with DG initiatives. To address this, the participant suggested

using an agile methodology rather than a waterfall approach. By delivering value

quickly, IT managers can gain executive buy-in. For instance, the first few sprints could

focus on establishing a data quality control and monitoring system. Breaking the process

down into milestones helps executives understand and appreciate the incremental value

being delivered. The participant also mentioned a selling strategy based on current

organizational pain points, such as poor data quality. For example, if poor data quality is

causing a significant financial loss, the case for DG becomes more compelling. By

framing DG as a solution to overcome these challenges, executives are more likely to see

the big picture and the long-term benefits of implementing a DG framework. Regulatory

compliance is another area where DG is easier to justify to executives. Failure to comply

with regulations directly impacts the bottom line, making it a more straightforward sell.

The participant suggested recommending regular audits of tools, technologies, processes,

and policies to prevent security breaches and ensure compliance. These strategies are key

to winning executive support for DG initiatives. Participant 10 also mentioned that

securing executive support for data governance (DG) is challenging and emphasized the

need to clearly demonstrate the program’s value to gain their backing. Similarly,

Participant 11 highlighted that the biggest hurdle is achieving buy-in, compounded by the

lack of standardized reference implementations for DG. Despite the existence of some

frameworks, there’s no universally accepted reference architecture, leading organizations

to develop their own approaches. Table 13

Frequency of Sixth Major Theme

Source of data collection n

Interviews 5

Documents 0

Note. Theme 6: Effective data governance strategies must address the challenge of

justifying the value of DG programs; n = frequency.

The challenge of justifying the value of data governance (DG) aligns with both

the Organizational Scope and Consequences dimensions of the Data Governance

Conceptual Framework by Abraham, Schneider, and Vom Brocke (2019). In terms of

Organizational Scope, the focus is on the intra-organizational level, where DG is

implemented within a single organization. This involves managing the quality, integrity,

security, privacy, and compliance of organizational data assets. Achieving this requires

collaboration with various stakeholders, business groups, and IT departments to address

their interests and secure their buy-in. However, this is often challenging due to the

diverse needs and interests across the organization. Business use cases vary, and

perceptions of ROI can differ significantly between mid-level management and

executives. The Consequences dimension addresses two key areas: intermediate

performance effects and the management of data-related risks. The performance effects

of DG refer to improvements in areas such as data quality, data access, and metadata

management, as outlined by Kathri and Brown (2019). Enhancements in data quality—

such as accuracy, availability, completeness, consistency, and timeliness—are tangible

and easier to recognize. The second aspect is the management of data-related risks,

particularly those arising from non-compliance with regulations, as well as security and

privacy breaches. Without concrete implementations and measurable outcomes, it can be

challenging to convince executives and IT managers of the positive impacts of DG

upfront. This underscores the core challenge highlighted in this theme.

Theme 7: Successful DG Implementation Requires Relevant Data Competencies for

All DG Roles

The success of data governance (DG) relies heavily on data competencies. Data

stewards and data owners need strong acumen in data management and a solid

understanding of data lifecycle components. IT managers implementing DG programs

should prioritize training participants in essential data skills, including data querying,

data classification, data analysis, and data reporting.

Two participants highlighted the importance of technical and data skills in their

responses, a point corroborated by two reports used for methodological triangulation.

Participant 7 emphasized the need for both technical skills to use data governance (DG)

tools and data skills to understand key aspects such as data quality, classification, and

security. For example, when IT integrates technology to implement the data lifecycle,

including governance, there is often a focus on the technology itself, sometimes at the

expense of the people side—change management and user adoption, which are crucial

for success. Participant 7 noted that while consultants may provide solutions for

challenges like data integration, quality, and lineage cataloging, these efforts can fail if

the people expected to own and maintain the solutions lack the necessary understanding.

Often, once the consultants leave, the organization struggles because those responsible

for the solution haven’t received proper training to maintain or enhance it. Participant 10

reinforced this point, suggesting that individuals hired to manage DG, whether on the

technical or process side, are not always experts in the field. DG is often mistakenly seen

as a project management task, but it requires deep technical knowledge. The right person

must have data skills, such as SQL proficiency and a solid understanding of data

platforms, to effectively run a DG program. Nephos Technologies (2022) further

supports this by identifying people as one of the top barriers to successful DG. Their

survey of DG professionals in the UK revealed that 52% of respondents felt they lacked

the necessary skills to deliver effective data governance. Similarly, ESG (2022) reported

that a skills shortage or gap represents 40% of the barriers to DG success.

Table 14

Frequency of Seventh Major Theme

Source of data collection n

Interviews 2

Documents 2

Note. Theme 7: Successful DG implementation requires relevant data competencies for

all DG roles; n = frequency.

Abraham et al. (2019) identified key aspects of data management that data

governance aims to address, including Data Ownership, Data Stewardship, Data Policies

and Standards, Data Quality Management, Data Security and Privacy, Data Compliance

and Risk Management, and Data Integration and Interoperability. These aspects align

with the Governance Mechanisms dimension of their framework, which includes

structural, procedural, and relational mechanisms—essentially the processes and policies

that guide data governance. Each of these aspects requires specific skills to ensure

successful implementation. Without the appropriate expertise, data governance efforts

are likely to struggle or even fail.

Theme 8: Successful DG Programs Should Have a Clear Strategy, Be Targeted, Use

Case Bound, and Be Antecedent Driven

The final theme that emerged from the analysis complements, rather than aligns

with, the conceptual frameworks discussed earlier (Alhassan et al., 2019). First, an

established data strategy is crucial. While a fully developed data lifecycle is not required

for a data governance (DG) program, DG must function within the context of existing

data infrastructure and processes. A clear data strategy is essential for guiding any DG

roadmap and implementation. Second, DG implementation should be targeted and use

case-bound. Begin with a focused, specific use case driven by a business need or

challenge, such as improving product sales. The use case should directly address this

need.

All participants agreed on the importance of establishing a clear data strategy as

the foundation for a data governance program. However, their strategies varied. Some

advocated for starting small, while others emphasized embedding data governance from

the outset to enhance compliance, privacy, security, and risk mitigation. Additional

approaches included making data governance antecedent-driven, adopting a “governance

by design” model, and following a “don’t boil the ocean” strategy.

Participant 1 emphasizes an agile, iterative approach, embedding DG into the data

lifecycle from the outset to ensure compliance, enhance privacy and security, and

mitigate risks, particularly in regulated industries. Participant 2 echoes the need for a

clear data strategy but stresses that DG should be driven by specific business needs or

challenges, rather than being IT-led, to empower better data management and

decisionmaking. Participant 3 aligns with Participant 1, advocating for embedding DG

from the ground up in data management processes to ensure a comprehensive

understanding of data sources and usage. Participant 4 highlights that many companies

treat DG as an afterthought, which is problematic without executive and business buy-in.

They stress that securing this support should be a priority. Participant 5 also supports a

“governance by design” approach, where governance controls are integrated from the

start, with a focus on one use case at a time. They emphasize the importance of

budgeting for DG, having the right tools, and engaging stakeholders early. Participant 6

and Participant 7 advocate for starting with small, specific use cases rather than large,

enterprise-wide programs, which can be costly and prone to failure. They recommend a

comprehensive data strategy as the first step. Participant 8 suggests beginning with data

quality, lineage, and master data management, aligning with FAIR principles, and

iteratively implementing DG. Participant 9 underscores that strategy is fundamental,

guiding the creation of tactical steps for implementation, including prioritizing based on

business alignment, defining roles and responsibilities, and measuring success with KPIs.

Participant 10 advises starting small to build momentum and gain executive support,

using a proof of concept to demonstrate the value of DG. Finally, Participant 11 stresses

that DG should be integrated from the start as an ongoing process, with buy-in from data

producers who are responsible for governance implementation.

Table 15

Frequency of Eighth Major Theme

Source of data collection n

Interviews 9

Documents 0

Note. Theme 8: Successful DG programs should have a clear strategy, be targeted, use

case bound, and be antecedent driven; n = frequency.

Applications to Professional Practice

Data Governance (DG) is increasingly crucial, particularly in the context of

advanced analytics, AI, security, privacy, and regulatory compliance. As a strategic

enterprise asset, data requires a cross-functional DG framework that outlines decision

rights and responsibilities for managing it effectively (Abraham et al., 2019). This study

offers valuable insights for IT managers, identifying strategies to overcome the key

challenges in DG implementation. By gathering knowledge and experiences from

seasoned DG professionals across various industries, the study highlights key drivers and

practical considerations essential for successful DG initiatives.

The study’s findings are significant because they align with the data governance

decision domains outlined in Khatri and Brown’s (2010) Unified Framework for Data

Governance and the dimensions of the Data Governance Conceptual Framework by

Abraham, Schneider, and Vom Brocke (2019). These frameworks provide guiding

principles for understanding the implementation of DG, including its challenges and

successes, through six dimensions of analysis. Methodological triangulation with several

industry reports on DG confirmed the study’s findings.

Effective strategies such as data discoverability and data cataloging with lineage

are essential for IT managers to meet regulatory compliance requirements, particularly in

regulated industries. A major challenge for DG managers is often justifying the value of

DG to executive leadership to secure their buy-in. This study suggests strategies to

address this challenge, such as starting small, focusing DG programs on specific use

cases or data domains, building pilot projects and proof of concept programs, recording

performance KPIs, and demonstrating value. Additionally, the study highlights the

importance of training and socializing data skills to empower those in DG roles to

implement and drive programs successfully. Socialization and incentivization, supported

by an executive champion, can help incorporate DG into the organization’s culture and

processes.

Some participants in the study recommended a shift in mindset towards data

engineering with DG embedded from the outset. This approach is akin to the early 2000s

when software was often designed without security in mind, leading to frequent

malicious attacks. Companies like Microsoft led the industry in changing this approach

by designing software with security at its core. The study suggests a similar shift for DG,

advocating for its integration from the beginning rather than as an afterthought in data

management. These strategies enhance the critical success factors discussed in the

study’s literature (Alhassan et al., 2019), providing IT managers with additional

knowledge to implement successful DG programs.

Implications for Social Change

Technological innovation should prioritize societal well-being, driving positive social

change through enhanced data quality, security, privacy, and the democratization of data

governance. Effective data governance improves the quality, trustworthiness, and exchange

of data (Hikmawati et al., 2021), enabling companies to harness advanced analytics and AI to

extract value from Big Data. This study identified data quality, security, privacy, and

regulatory compliance as key drivers of data governance.

By focusing on these drivers, IT managers can contribute to better data quality across

various sectors, including healthcare, finance, and retail. Stronger data governance helps

prevent data breaches that could harm institutions and, ultimately, consumers. Additionally,

high-quality data is crucial for the AI-driven innovations transforming industries and

enhancing people’s lives. Improved data governance also ensures compliance with

regulations, helping companies avoid costly fines that could be passed on to consumers.

Recommendations for Action

This study confirmed that data is a strategic organizational asset that must be

governed with the same priority and rigor as any other company asset. Data quality,

security, and privacy are critical for both companies and users to enable strategic

business decisions and drive innovation. Abraham et al. (2019) emphasized the

importance of data governance in ensuring that data assets are properly managed to

maximize value and minimize risks.

Company executives must recognize data as a strategic asset within the data

management ecosystem and allocate sufficient resources to enhance its quality, security,

privacy, and compliance. Crucially, they must move beyond treating data governance as

an afterthought and instead fund robust programs that address the challenges and needs

associated with these key drivers. IT managers should focus on implementing data

discoverability, data cataloging, and the standardization of data terms across the

organization, while also democratizing data skills.

To overcome the key challenge of securing executive buy-in for DG, IT managers

should adopt the strategies unearthed in this study. In addition to following best practices

outlined in critical success factors (CSF), IT managers can demonstrate the value of data

governance through quick wins, agile methodologies, proof-of-concept initiatives, and

pilot programs, all supported by clear KPIs. Finally, technical and data skills are

essential for the success of data governance. IT managers must seek funding and

resources to properly train those in DG roles, ensuring they can effectively execute data

governance programs.

As promised in the interview protocol and during the interviews, I will share the

results of this study with the participants, many of whom expressed eagerness to learn

about the findings. My hope is that this cumulative knowledge will assist them in their

DG implementation efforts and benefit other professionals as well. I also plan to explore

ways to disseminate the study’s findings to a broader audience, including publishing a

summary in the Data Governance Professionals Organization (DGPO) journal and other

relevant publications. Additionally, I intend to reach out to Yebeness and Zorilla (2022)

to discuss strategies for Big Data governance.

Recommendations for Further Study

This study highlighted several areas that warrant further exploration to deepen the

understanding and enrich the body of knowledge on Data Governance (DG). One key

area is DG for Big Data. Yebeness and Zorilla (2022) laid a solid foundation with a

framework and tentative reference architecture for understanding the motivations and

challenges of DG in the context of Big Data. Future research could build on their work

by focusing on the specific challenges IT professionals face in implementing DG for Big

Data and Cloud Data. Another important area for further study is the implementation of

data discoverability and data catalogs as enablers of effective data governance.

Interviews with IT managers and professionals could provide valuable primary data and

insights across industries, shedding light on the needs and challenges associated with

these aspects of DG.

Reflections

My primary motivation for choosing this research topic was to gain a deep

understanding of the theoretical foundations of Data Governance (DG), enabling me to

better grasp its challenges, opportunities, and importance in data management. As a

professional data engineer and architect, I have developed solutions for clients where DG

was claimed to exist, supported by a dedicated team. However, my assessment of data

quality and processes often revealed significant gaps. This research allowed me to

identify common challenges companies face, such as siloed organizations, DG efforts

driven by individual business teams, and a lack of executive support, budget, and buyin

—all factors that hinder the success of DG programs.

Through interviews with experienced DG managers across industries and an

extensive review of the literature, I have come to fully appreciate the critical role of DG

in managing data as a vital enterprise asset. This newfound knowledge will enable me, as

a data engineering professional, to design data platform solutions with DG integrated

from the ground up.

Summary and Study Conclusions

The primary motivation for this study was to explore strategies employed by IT

managers to effectively implement data governance (DG) programs. The findings

confirmed that data quality, security and privacy, and regulatory compliance are the main

drivers behind DG initiatives. Additionally, the study revealed a growing trend:

companies are increasingly treating data as a strategic asset, a shift from previous

practices.

Two key challenges to DG programs were identified: securing executive support

and justifying the value of DG initiatives to both leadership and mid-management. The

data collected through participant interviews provided actionable strategies for IT

managers. Chief among these is winning executive buy-in by demonstrating the value of

DG programs through pilot projects, proof of concepts, and agile methodologies.

Another successful approach involves designing DG programs with a clear, use-case-

driven strategy. Methodological triangulation was employed to validate the themes and

findings, comparing them against industry DG reports from various companies and

organizations.

This study offers valuable insights for IT managers, highlighting strategies to

overcome the main challenges in DG implementation, particularly in securing executive

support, obtaining funding, and demonstrating ROI through quick-win strategies. The

study’s findings also have broader implications for social change by emphasizing the

role of effective data governance in enhancing data quality, security, and privacy,

ultimately protecting end users from identity theft and misuse of personal data.

Information Technology Managers’ Strategies for Implementing Data Governance

Students also viewed