1. Data Governance Context Diagram
1. Data management and data governance
Before formally talking about data governance, first distinguish the difference between data management and data governance. The overall drive for data management is to ensure that organizations can derive value from their data , more at the level of overall strategy ; data governance focuses on how decisions are made about data, and how people and processes behave with respect to data, more at the level of detail execution level .
2. Overall introduction to data governance
Data governance is the core content of DMBOK, not only in the central position among the 11 data management functions, but also in each individual data management function, there is a part of data governance. This emphasizes that data governance is not a separate process, but must be integrated into the system design and development process, and implemented throughout the entire process of system construction, so that data governance can be implemented better .
Achieving good data governance requires building a data-centric organization: Organizations must change the way they translate strategy into action. Data is no longer viewed as an adjunct to processes and business products. The goal of business processing is to obtain high-quality data . Effective data management has become a high priority as organizations strive to gain insights and make decisions through analytics.
2.1. General content of data governance
The specific content of data governance is generally related to the actual business needs of the organization, but basically includes the following content:
-
- Strategy . Define, communicate and drive execution of data strategy and data governance strategy.
- System (Policy) . Set up policies related to data, metadata management, access, usage, security and quality.
- Standards and Quality . Set and enforce data quality and data architecture standards.
- Oversight . Provide observation, audit and corrective actions (commonly referred to as stewardship) in key areas of quality, policy and data management.
- Compliance . Ensure organizations can meet data-related regulatory compliance requirements.
- Issue Management . Identify, define, escalate, and address issues in areas such as: data security, data access, data quality, compliance, data ownership, policies, standards, terminology, or data governance procedures.
2.2. General principles of data governance
Every organization needs to have its own principles, but those looking to get more value from their data are likely to share the following:
-
- Data should be managed as an enterprise asset.
- Best practices in data management should be encouraged throughout the organization.
- An enterprise data strategy must align with business strategy.
- Data management processes should be continuously improved.
2.3. Main activities of data governance
In most cases, data governance activities will focus on the following parts:
1) Create and manage core metadata . It includes the definition and management of business terms, valid data values, and other key metadata. Usually, the management specialist is responsible for sorting out the business glossary, which becomes the data-related business term recording system.
2) Document rules and standards . It includes the definition and documentation of business rules, data standards, and data quality rules. Meeting expectations for high-quality data is often based on business process specifications for creating and consuming data. To ensure consensus within the organization, data governance professionals help define the rules and ensure their consistent application.
3) Manage data quality issues . Data stewardship professionals are typically involved in identifying, resolving data-related issues, or facilitating the process of resolution.
2. Business drivers
1. Reduce risk
1) General risk management . Gain insight into the financial or reputational impact of risk data, including responses to legal and regulatory issues.
2) Data security . Protect data assets through control activities, including availability, availability, integrity, continuity, auditability, and data security.
3) Privacy . Control private, confidential, personally identifiable information (PII) and more with policy and compliance monitoring.
2. Improve the process
1) Regulatory compliance . The ability to respond effectively and consistently to regulatory requirements.
2) Improve data quality . The ability to drive business performance through truthful and trusted data.
3) Metadata management . Establish a business glossary for defining and locating data in the organization; ensure that the vast amount of metadata in the organization is managed and applied.
4) Project development efficiency . Improve in the system lifecycle (SDLC) to address data management across the organization.
5) Supplier management . Contracts governing data processing, including cloud storage, external data procurement, data product sales, and outsourced data operation and maintenance.
3. Goals and principles
1. Goal
(1) Sustainable development (Sustainable)
Governance procedures must be attractive. It is not a project as the end, but an ongoing process. It needs to be the responsibility of the entire organization. Sustainable data governance relies on the support of business leaders, sponsors and owners.
(2) Embedded (Embedded)
Data governance is not an additional management process. Data governance activities require a fusion of software development methodologies, data analytics applications, master data management, and risk management.
(3) Measured
Data governance done well has positive financial impact, but demonstrating that impact requires understanding the starting process and planning for measurable improvements.
2. Principles
(1) Leadership and Strategy
Successful data governance starts with vision and committed leadership. The data strategy guides data management activities and is driven by the enterprise business strategy.
(2) Business-driven
Data governance is a business management initiative, so IT decisions related to data must be managed just as business activities related to data are managed.
(3) Shared Responsibility
In all knowledge areas of data management, business data management professionals and data management professionals share responsibility.
(4) Multi-layered
Data governance activities occur at the enterprise level and at the grassroots level, but often at levels in between.
(5) Framework-based
Since governance activities need to be coordinated across organizational functions, an operational framework must be established for data governance projects to define their respective responsibilities and work content.
(6) Principle-based
Guiding principles are the foundation of data governance activities, especially data governance policies. Sometimes principles can be reverse engineered from specific policies. However, it is best to work on the articulation of core principles and best practices as part of the strategy.
4. Activities
| level one activity |
secondary activity |
Supplementary details |
| Plan your organization's data governance |
Executive Readiness Assessment |
Including data management maturity, change capability, etc. |
| Discovery Aligns with Business |
/ |
|
| Develop Organizational Touchpoints |
|
|
| Develop a data governance strategy |
Develop a data governance operational framework |
See VII. 2 Operational Framework for details |
| Formulate goals, principles and systems |
Goals and principles are detailed in three Systems can be considered: confirm organizational data, designate business owners, assign data management specialists, provide standardized reports/scorecards, control access rights, and regularly review certification data |
|
| Drive data management projects |
Build organizational buy-in and reduce friction by demonstrating that data management improves efficiency and reduces risk. |
|
| Participate in change management |
Appropriate initiators and managers are required: Planning - Training - Influence System Development (PMO) - Institutional Implementation - Communication |
|
| Participate in Issue Management |
Problem handling and reporting mechanism |
|
| Assess regulatory requirements |
Focus on relevant regulatory requirements and institutionalize compliance |
|
| Implement Data Governance |
Initiate data standards and procedures |
Formulate standards and enforce standard measurement of data |
| Make a Glossary of Business Terms |
/ |
|
| Coordinating Architecture Team Collaboration |
/ |
|
| Initiate data asset valuation |
Giving monetary value to data is a very challenging task. Common ones include cost sharing, virtual settlement, etc., which need to be developed and agreed upon according to organizational conditions |
|
| Embedded Data Governance |
Embedded Data Governance |
Governance activities need to be embedded in a series of processes related to data as asset management |
5. Implementation Guidelines
1. Organization and culture
The overall organization and culture emphasizes high-level support and considers organizational and personal factors during execution, slightly
2. Communication and adjustment
With the continuous development of governance, the following key content needs to be continuously updated and synchronized with relevant parties in a timely manner:
1) Business strategy/data governance strategy blueprint (Business/DG Strategy Map) . These blueprints link data governance activities to business needs. Regularly measuring and communicating how data governance helps the business is critical for data governance to gain ongoing support.
2) Data Governance Roadmap (DG Road Map) . The data governance roadmap should adapt to changes in business circumstances or priorities.
3) Ongoing Business Case for DG . The business case for data governance must be adjusted periodically to reflect the changing priorities and financial situation of the organization.
4) Data governance indicators (DG Metrics) . As the data governance procedures mature, the relevant indicators of data governance should also gradually grow and change accordingly
6. Metrics
To address the resistance and challenges of long learning curves, it is imperative for data governance projects to measure progress and success through metrics that demonstrate how data governance participants add business value and achieve their goals . Examples of data governance metrics include:
(1) Value
1) Contribution to business goals.
2) Risk reduction.
3) Improvement of operational efficiency.
(2) Effectiveness
1) The realization of the goal.
2) Expand the relevant tools that data management professionals are using.
3) Effectiveness of communication.
4) Effectiveness of training.
5) Speed of adoption of change.
(3) Sustainability
1) How systems and processes are performing (i.e. are they working properly).
2) Compliance with standards and procedures (ie, whether employees follow instructions and change behavior when necessary).
7. Key concepts/tools/methods
1. Data Governance Organization
Data governance can be understood in terms of political governance: it includes legislative functions (defining policies, standards, and enterprise architecture), judicial functions (issue management and escalation), and executive functions (protection and services, stewardship responsibilities).
The following diagram depicts a typical Data Governance organizational model:
Among them, the responsibilities of some key roles are as follows:
For the specific responsibilities of more key roles (such as CDO, data management specialist), see DMBOK for details.
2. Operating framework
The following two figures introduce the typical operational framework in the process of data governance, including system formulation and distribution, daily work of data governance, problem handling and promotion, etc.
3. Different governance operation models
There are three main organizational models of data governance: centralized governance, distributed governance, and federated governance. Which organizational model to choose needs to be confirmed according to the actual needs of the organization and business.
4. The value of data governance
The value assessment of data governance mainly includes two parts: the value created and the cost saved, which is often referred to as cost reduction and efficiency increase. The cost is not only the direct resource cost, but also the potential risk cost and so on.
1) Replacement Cost (Replacement Cost) . The cost of data replacement or recovery in the event of a catastrophic data corruption event or data outage, including transactional, domain, directory, document, and metrics information within the organization.
2) Market Value . The value of a business asset when a business is merged or acquired.
3) Discover business opportunities (Identified Opportunities) . Discover the revenue value of business opportunities from data (business intelligence) by trading data or selling data.
4) Selling Data . Some organizations package the insights gained from the data into packages for product or sales.
5) Risk Cost (Risk Cost) . It is based on an estimate of potential fines, remediation costs and litigation costs. Legal or regulatory risks include:
①Lack of required data.
②There is data that should not be retained (for example, unexpected data discovered during a legal audit; data that needs to be purged but has not been purged).
③In addition to the above costs, including damage to customers, company finances and reputation caused by incorrect data.
④The decrease in risk or risk cost is actually the overflow after offsetting the operational intervention costs such as upgrading and verifying data.
In order to describe the concept of information asset value, generally accepted accounting principles can be transformed into generally accepted information principles to assist asset value judgments:
