This is Part 2 of the article “A Data Strategy: Theory vs. Practice.” In Part 2, we will continue the analysis of data strategy examples and discuss the following:

  • A data (management) strategy content: recommended vs. presented in strategies mentioned above (sections 2 and 3)
  • Recommendations for developing a (meta)data (management) strategy

A data (management) strategy content: recommended vs. presented in the real strategies (Continuation)

“What” section

This section forces a company to make a serious decision that will impact the success of the data management initiative. This decision is the balance between “we want” and “we can.” In other words, it goes about the feasibility of the strategy. When you start writing a data (management) strategy, you should be honest with yourself and your company about the goal of writing the strategy. Do you write it pro forma for demonstration purposes because “all others do it,” or do you really need it with the key goal of its implementation? You can stop reading this article if your goal is the first one. If your goal is strategy implementation, I encourage you to dive into the topic of data management principles, framework, and core data management capabilities.

To define the feasible data management strategy, you should make decisions about the following topics:

Topic 4: Data management principles


A data management principle is a rule that regulates the way data management is implemented.

Different leading industry guidelines have pretty different approaches to defining data management principles. The DAMA-DMBOK approach in the second edition deviates from the approach in the first. In the first edition, the data management principles were more generic. The second edition defines principles per Knowledge Area. I like the approach taken by The Open Group in the TOGAF® Standard. In my practice, I apply the method that connects business drivers with data management principles and analyzes the consequences of applying principles.

Figure 5 demonstrates this approach.

Figure 5: The approach to formulating data management principles.

Figure 5: The approach to formulating data management principles.

Data management principles must be formulated with a key focus on the feasibility of their implementation. The analysis of consequences must include potential benefits and challenges and required actions.


I found the formulated principles in four of the five referenced strategies.

Let us review these principles that I grouped into several categories related to:

Data governance:

  • “Data are an asset” (Strategy 4)
  • “Data must have clearly defined accountabilities” (Strategy 4)
  • “Data must follow rules and regulations” (Strategy 4)
  • “Data should be managed consistently” (Strategy 4)


  • Principles of professional ethics (Strategy 2)
  • Ethical use (Strategy 4)
  • Ethical governance (Strategies 3 and 5), including:
  • “Uphold Ethics” (Strategy 5)
  • “Exercise Responsibility” (Strategy 5)
  • “Promote Transparency” (Strategy 5)


  • Conscious decisions (Strategies 3 and 5), including:
  • “Ensure Relevance” (Strategy 5)
  • “Harness Existing Data” (Strategy 5)
  • “Anticipate Future Uses” (Strategy 5)
  • “Demonstrate Responsiveness” (Strategy 5)


  • Data-informed culture (Strategy 4)
  • Learning culture (Strategies 3 and 5), including
  • “Invest in Learning” (Strategy 5)
  • “Develop Data Leaders” (Strategy 5)
  • “Practice Accountability” (Strategy 5)

Data management-related

  • Data-centric principles for IT architecture (Strategy 2)
  • Governance and effective management (Strategy 4)

As we can see, these strategies have a lot of similar principles.

Topic 5: The data management framework


I wrote multiple articles on the differences between viewpoints on data management structures of leading data management guides, DAMA-DMBOK2 and DCAM. Years ago, this unalignment brought me to the idea of developing the “O.R.A.N.G.E.” data management framework to solve the issues I found in the leading guidelines. This framework is a set of methods and models to establish an operational data management function.

One of this framework’s foundational models is the data management capability model, presented in Figure 6. The data management capability consists of several core sub-capabilities. These sub-capabilities play different roles in delivering business value from data management.

Figure 6: The “O.R.A.N.G.E.” model of a data management capability.

Figure 6: The “O.R.A.N.G.E.” model of a data management capability.

Data lifecycle management is the core capability that delivers business value for an organization’s stakeholders. Business architecture and data governance are strategic capabilities that define the direction of data management development. Data governance is a special capability. The title “data governance” incorrectly reflects this capability’s real role. This capability governs data management, not data. Its core task is to establish data management as a business function applying a data management framework and then control data management function operational efficiency and effectiveness. Data governance does it by controlling the establishment of the data management organizational structure, processes, policies, and tools and ensuring resources for all data management sub-capabilities.


Several strategies (1, 3, 4) refer to themselves as a framework. It looks like the organizations did not use any industry framework and tended to develop their frameworks to meet their goals.

Topic 6: The scope of the data management capability


The “O.R.A.N.G.E.” data management model demonstrates the core data management sub-capabilities. The most important thing is that all data management sub-capabilities are interrelated. Organizations that started implementing data quality as their first initiative have a high probability of failing. In order to properly manage data quality, an organization must have data governance and data and application architectures, including data modeling and metadata management, including data lineage, etc. This is something that many data management professionals do not realize. This situation is partly due to the DAMA-DMBOK2 approach. This guideline creates the impression that all Knowledge Areas can be implemented independently. However, I have to give it my due that on page 38, the authors stated the following: “None of the pieces of existing DAMA data management framework describe the relationship between the Different Knowledge Areas.”

The rule of scoping the data management capabilities is simple: the business drivers identified in the strategy section “WHY?” will define the set of required sub-capabilities. The level of development of these sub-capabilities will depend on an organization’s resources.


I will use the “O.R.A.N.G.E.” data management model to analyze the core data management capabilities mentioned in the strategies. I did not find direct references to the data management capabilities in the strategies. To solve the challenge, I have translated the statements (goals, objectives) mentioned in the strategies into related capabilities and maps with the “O.R.A.N.G.E.” model.

The strategies assume the development of the following capabilities

Data governance

  • “Data Governance” (Strategies 1,3,4,5)
  • “Ethical Use” (Strategy 4)
  • “Data-Driven Culture” (Strategy 5)
  • “Data Process & Policy” (Strategy 5)

Data literacy

  • “Data and analytics workforce” (Strategy 1)
  • “Implement a data workforce plan that addresses the need to support data-driven decisions (Strategy 3)
  • “Data-Informed Culture (Strategy 4)
  • “Data Workforce” (Strategy 5)

Enterprise architecture (data-, application-, and technology)

  • “Open Data” (Strategy 1)
  • “Deliver Data Interoperability” (Strategy 2)
  • “Information collection strategies ((Strategy 3)
  • “Improve the data sharing process” (Strategy 3)
  • “Publish the Department’s Open Data Plan” (Strategy 3)
  • “Expand the Comprehensive Data Inventory” (Strategy 3)
  • “Enhance transparency through expanded access to agency administrative data” (Strategy 3)
  • “Strengthen the agency’s data release and disclosure review process” (Strategy 3)
  • “Create the State’s first Enterprise Data Inventory” (Strategy 4)
  • “Launch the State’s secured Geographic Data Sharing Hub” (Strategy 4)
  • “Conduct a survey of automated-decision systems in use” (Strategy 4)
  • “Improve data products and data principles” (Strategy 5)
  • “Build an Enterprise Data Architecture” (Strategy 5)
  • “Data Architecture” (Strategy 5)

Data quality

  • “Improve data quality with a focus on fitness for purpose” (Strategy 3)

Data lifecycle management

  • “Perform End-to-End Data Management” (Strategy 2)
  • “Strengthen the evidence-building pipeline in mission-critical domains (Strategy 3)
  • “Ensure the agency’s grant management system supports strategic data use”
  • “Effective Management” (Strategy 4)
  • “Data Tooling” (Strategy 5)
  • “Data Management” (Strategy 5)

Metadata management

  • “Greater use of common data standards” (Strategy 3)
  • “Establish a cohesive data skills program” (Strategy 3)

Data analytics

  • “Data and Analytics Workforce” (Strategy 1)
  • “Common Data & Analytics Toolsets” (Strategy 1)
  • “Analytics at Speed and Scale (Strategy 2)
  • “Build capacity to use data visualization and storytelling” (Strategy 3)
  • “Create value through the agency data analytics platform” (Strategy 3)

We can see that the strategies consider developing common data management capabilities with a key focus on data governance, literacy, analytics, and enterprise architecture.

“How?” section

This section must demonstrate how an organization will implement a strategy scope defined in the “What” section within a defined period. Several topics assist in doing this:

Topic 7: Long-term goals, objectives, and a roadmap


The roadmap defines the set of actions, stakeholders, and resources needed to achieve the goals and objectives specified in the strategy.

The most important point is that the roadmap must confirm the strategy’s feasibility. The roadmap must take into account dependencies between various sub-capabilities we discussed above.


All strategies were concluded with clearly identified goals and objectives—the examples of the goals and objectives I listed while discussing Topic 6.

Some of them also included roadmaps (Strategies 4 and 5).

Topic 8: Measures for success


Usually, we measure success in data management by setting up performance management and using key performance indicators as a measuring tool. The “O.R.A.N.G.E.” data management framework recommends the set of KPIs per organizational level. A data strategy belongs to the strategic organizational level. An organization can use three types of KPIs to measure the strategy’s success: leading, financial, and outcome.

The leading KPIs measure long-term trends and predict successful future outcomes of the data management strategy.

The financial KPIs identify the expected monetary return on investments.

The output KPIs measure whether the strategy meets the goals and objectives for the long term.


I could not find any direct reference to methods of measuring success.

Topic 9: Risk assessment and mitigation


Factors of external and internal environments can cause multiple risks for successful strategy implementation. The strategy must foresee these risks, develop mitigation actions, and turn the mitigation of the risks into success factors.


I could not find any analysis of potential risks and mitigation actions in any of the strategies.


Below is the summary of what we discussed in this article:

  1. Several industry authorities provide guidelines for developing a data (management) strategy. We considered and compared two examples: the DAMA-DMBOK2 and Gartner models.
  2. The proposed contents of the strategy are similar and encourage answering three key questions:
    • Why does an organization need a data (management) strategy?
    • What does “data management” mean for the organization, and what is the required scope?
    • How can the organization implement the identified scope of data management?
  1. Governmental institutions share their data strategies with the public. It is not easy to find data strategies published by commercial organizations.
  2. The content of the strategies analyzed in this article generally corresponds to the content recommended by industry authorities.
  3. The strategies focus on developing core data management capabilities like data governance, analytics, literacy, quality, enterprise architecture, and data management lifecycle. Unfortunately, the focus on metadata management is limited.

Recommendations for developing a (meta)data (management) strategy

If an organization decides to develop a (meta)data (management) strategy, it should do the following:

  1. Identify the focus of the strategy “data” vs. “data management”

A data strategy mainly focuses on the role of data in the business, while a data management strategy demonstrates the viewpoint on improving data handling practices.

  1. Define the key business reasons for establishing a solid data management strategy

Business reasons are the leading reason for investing in data management. The strategy must identify the key stakeholders, external and internal. It is worth noting that data management delivers different values to different stakeholder groups.

  1. Identify the scope of the required data management capability that fits the business drivers

Data management is a foundational capability. No organization can successfully operate its business without this capability. However, the scope of developing this capability must be linked to the most significant business drivers. The organization must define its core data management sub-capabilities to implement the strategy.

  1. Develop feasible objectives and goals and translate them into a long-term roadmap

While developing the strategy, the organization must consider the available resources. The strategy must be feasible. Setting up S.M.A.R.T. goals and objectives helps develop realistic and realizable long-term plans.

The strategy must also include risk analysis and mitigation actions. It also must demonstrate how it will measure progress an