We continue elaborating on the four most often used cliches about data management demonstrated in Figure 1.
Figure 1: Logical relationships between the four most often used data-related cliches.
In the first two articles of the series, we discussed the implementation aspects of the statement “data is a company’s asset.” If data is an asset, then it should generate value for the company and its stakeholders. In the second article, we elaborated on five steps to analyze how a company can get more value from different types to sustain the company’s competitive advantage. To get value from data, a company should become “data-driven.”
In this article, we will:
- Discuss different viewpoints on a “data-driven” organization
- Assist a company in choosing its definition of being “data-driven.”
- Show the 7-step method to become “data-driven” following the chosen definition
Various viewpoints on a “data-driven” organization
The first time I dealt with the term “data-driven” organization was five years ago. At the time, I concluded that no common view existed. For this article, I again made some Google searches. I still got the same feeling: there is no agreed common view and approach to the meaning of this statement. The term “data-driven” has been put into a different context and therefore has different meanings. The following facts support this conclusion:
- The kick-off words in the definitions differ.
The kick-off word is the first noun or noun phrase in the definition. The term “data-driven” is defined as a “business state,” “the recognition,” “a situation,” “an approach,” a business condition,” and so on. These kick-off words put this term in different contexts. A state or situation describes an “as-is” status. An approach focuses on how to achieve a “to-be” status. A business condition defines the circumstances or factors needed to achieve something.
- Definitions have different focuses. Some definitions answer the question “WHAT” does it mean to be “data-driven?” Others focus on the question, “How to become data-driven?”
According to standard rules in designing business terms and definitions, a definition should answer the question “WHAT” is a thing. When you know WHAT you need to achieve, you can find multiple ways to reach this status. Many definitions are unable to define the “data-driven” status clearly. Instead, they explain how to do some things.
Below, there are examples of the definitions that focus on the “WHAT” question:
- Data is organized and examined with the goal of better serving one organization’s customers and consumers
- “makes strategic decisions based on data analysis and interpretation.”
The following definitions focus on the “HOW” question:
- “[…] Data must be managed efficiently and with professional discipline[…]”
- “Integrated data analysis” is at “the core of its business processes.”
- Building a data lifecycle that includes data sourcing, sharing, analytics
- Implement a data management
- “Data powers digital transformation“
- “building tools, abilities, and, most crucially, a culture that acts on data“
- in all its forms to achieve digital transformation, compete on analytics, or become AI-first.”
Figure 2 summarizes the abovementioned definitions and views on the “data-driven” term.
Figure 2: The summary of definitions and views on the “data-driven” term.
So, if a company decides to become data-driven, it should first design a clear view of the applicable meaning of this term. And only after that the company should specify how to achieve the state of being “data-driven.”
Further, I will share the Data Crossroads’ view on the “data-driven” concept in this article.
“WHAT” does it mean to be a data-driven organization?
“Data-driven” is a feature of a business model that enables a company to use data as the most significant factor in decision-making at all organizational levels.
So, the term “data-driven” reflects the ability of a business to include data management in its business model. Then, data management organizes the data lifecycle so that the data becomes the most significant factor in decision-making.
The current interpretation of the “data-driven” concept does not demonstrate the role of human beings in business decision-making. By now, the participation of a human being remains unavoidable. However, the part of data changes: the business decision by human beings should be based on reliable data instead of “gut” feeling.
“HOW” to establish a data-driven organization?
Figure 2 demonstrates different answers to this question. However, all answers can be combined into one: establish an operational data management capability by implementing a DM framework. The reasons for that are the following:
- Data analytics is part of a data lifecycle.
So, operational data analytics alone is insufficient to create a “data-driven” company. Data analytics is only one step in a data lifecycle. - A data lifecycle model depends on the company’s business model. Different data chains realize the data lifecycle model.
A company should optimize different data chains to adjust the data lifecycle to the updated business model. - A data management capability enables data chains.
A data management capability includes several other capabilities needed to design, implement or optimize data chains. To perform it, data management should become a business function. - A data management framework defines the data management function.
A data management framework is a collection of interrelated components that shapes data management into a business function.
Figure 3 shows the visual representation of all reasons discussed above.
Figure 3: The relationships between various advice to build the “data-driven” organization.
An organization can become “data-driven” by implementing the data management framework to establish a data management function.
The “Orange” data management framework offers a ready-to-use approach to establishing an operational data management function. Figure 4 demonstrates the concept of the data management framework:
Figure 4: The definition of the data management framework.
A data management framework is a collection of interrelated components that shapes data management into a business function.
A data management capability is an ability of an organization to safeguard data assets and deliver business value from them.
A data management function is implementing the data management capability in the organizational structure.
The DM framework has several components and serves various goals, as shown in Figure 4. Models and methods are the key components of the data management framework.
A method is “a procedure, technique, or way of doing something, especially in accordance with a definite plan.”
The key goals of the data management framework are to design, implement, and measure the maturity and performance of the data management capability.
The “Orange” data management framework includes a model of a data management capability and a 7-step method to implement it. Figure 5 represents this method,
Figure 5: The 7-step method.
Phase 1: Scoping the data management framework
Phase 1 is crucial for the success of a data management initiative. During this phase, a company identifies its business needs, sets long-term goals, and assesses feasible resources. It leads to the definition of the scope of the DM capability and framework. The feasible scope guarantees the achievement of goals within a defined period.
Phase 2: Performing a preliminary maturity assessment
Phase 2 of the “Data management star 2.0” is essential for long- and medium-term planning. During this phase, a company identifies its “as-is” situation and assesses the desired “to-be” status. The results of this assessment form the basis for developing a data (management) strategy and/or roadmap.
Phase 3: Designing a data (management) strategy and/or roadmap
The data management strategy is a long-term future state document that demonstrates the intent of a company to manage and use data in accordance with its business strategy. Information technology should enable these intentions. Data comes first; technology follows.
The “Orange” data management framework (the “Orange” DMF) recommends designing a data strategy after performing the preliminary maturity assessment and before the designing of key data management capabilities.
Phase 4: Designing data management capabilities
Every company can use or adjust the existing data management framework or create its framework. Any option is viable as long as the framework fits the purpose and resources of the company. The “Orange” DMF includes the most common data management capabilities:
- Governance of the data management framework
- Data modeling
- Information systems architecture
- Data chain management
- Data quality
Phase 5: Implementing the data management framework
Phase 5 is the most challenging step in the success of a data management initiative. A company should bring theoretical developments from Phase 1 up to Phase 4 into practice. Phase 5 is also the most time and resource-consuming. The key rule is that once started, the development of the data management capability will never stop. Challenges in the business environment will always generate new demands for data management.
Phase 6: Performing the detailed maturity assessment
A company can proceed with Phase 6 of the “Data management star 2.0” when it has already made some progress with implementing the data management framework. The key goal is to assess progress and improve and optimize the framework.
Phase 7: Designing and implementing the performance management
Data management development is an ongoing process. Changes in the internal and external business environment will always demand modifications and further developments in data management capability. Data management is resource-consuming. Therefore, a company should continuously monitor the effectiveness of a data management capability.
I hope I have achieved two goals of this article: provide a definition of a “data-driven” organization and describe how the company can achieve it in practice.
In the final article of this series, we discuss digital transformation as one of the change processes to transform a business.
For more insights, visit the Data Crossroads Academy site: //academy.datacrossroads.nl.
