Implementation of Data Governance Concepts in A Digital Transformation Organization

Data-driven decision-making approaches have been popular nowadays to formulate new products, discover new features, or even measure the success of digital products. Organizations are aggressively collecting data to be further utilized as the fuel of their business and operational decisions. Amazon, IBM, and Oracle are a few tech companies implementing a data-driven decision-making approach (Mishra, 2021). At the stage when the data is massively collected, the data should be treated as an organization’s assets.

Despite its infinite potential benefits, a huge amount of data holds a number of potential significant risks too. For instance, an access management problem, where people within an organization need clarification about what data they can access and how to access it. This article discusses how we manage data assets by starting a Data Governance chapter from scratch.

The Concept of Data Governance

Data Governance is frequently confused with data management and master data management. The relationship between data governance and data management is similar to an auditor and financial management team in a financial process. An auditor controls financial processes, but it does not actually execute financial management — just like data governance which ensures data is appropriately managed without directly executing data management. In the simplest terms, data governance establishes policies and procedures around data, while data management enacts those policies and procedures to compile and use that data for decision-making. (Susan Earley, 2017)

Figure 1. The relation between Data Governance and Data Management

Data Governance Framework in Govtech

Data Governance aims to ensure that data is managed properly, according to policies and best practices (Ladley, 2012). The policies and procedures around data will be developed within a data governance program. Getting peers to onboard and adopt formal data governance programs is challenging for an organization, as it will involve cross-functional teams and engaging organizational change management. One of the most common misconceptions is the roles and responsibilities related to data are assigned to the Data Team (Data Engineer, Data Analytics, and so on);. At the same time, those should be a collective responsibility across functions.

At GovTech, we started our Data Governance journey when GovTech was still in the very early stage. Our journey began with observing the condition and various challenges in the organization at that moment. Some functional representatives (including the data engineer, data analytics team, policy & transformation team, and product manager) were involved in an in-depth interview to understand the data-related roles and responsibilities, data collection and processing, and data utilization at that moment. Second, we formulated a data governance charter based on the interviews and literature review results on the best practices of data governance (DAMA-DMBOK Data Management Body of Knowledge book). Two months after the kick-off, we came up with a data governance charter to define the goals, scopes, and data maturity assessment method (see here for more elaboration).

Building Data Governance Fundamentals in Govtech

We identified three fundamentals that were necessarily built at the initial stage of implementation, as follows

1. Data Governance Office

Data Governance Office (DGO) is a virtual office that executes the data governance policy, standard, and process within the data domain based on the initiatives and roadmaps developed by the Data Governance team. It is embedded in the existing role available in the organization. Three actors in the DGO are

• Data Domain Owner — a senior leader with a solid business understanding of the product & data. We embed this role to a person who has ultimate accountability for a product
• Data Stewards — a business professional, most often already the go-to subject matter expert on any product inquiries. They work with stakeholders to define and control data. We appointed the subordinates of the Data Domain Owner for this role
• Technical Custodian — technical experts who manage the technical environment where data resides. Various functions in the engineering team cooperate to pick up this role

Figure 2. Data Governance Office Structure

2. Data Maturity Assessment

Data maturity assessment can be seen as a complement to the previous preliminary observation and is needed as input to develop data governance programs. We customized the Data Maturity Assessment framework introduced by Stanford to fit our organization's condition. The assessment consists of seven components (i.e., awareness, formalization, metadata, stewardship, data quality, master data, and data management), and each component consists of three dimensions (i.e., people, policy, and capability). We elaborated each of those components and dimensions on the supplementary materials. The levels of maturity fall into the following stages (in order, lower to higher): the initial, managed, defined, quantitatively managed, and optimizing. We shared the questionnaire used in our organization to give the readers more concrete ideas on the assessment.

An example of the assessment’s result is depicted in Figure 3, where it tells us the following insights

Figure 3. An Example of Data Maturity Assessment Results and Score

• The team has been aware of the importance of the data. However, the data governance efforts are unsystematic and insufficient across functions, with no standardized process, tools, and roles in place. The governance is done on an ad-hoc basis.
• Talking about metadata, both business metadata and technical metadata, data are not appropriately documented while sitting in silos.
• The data quality process and effort are inconsistent across functions. Hence, the remediation was also conducted on a reactive basis.
• No clear and formal management of sensitive data within the Organization. Low awareness of risks, people are convenient to share and use the data, a proper access right management is needed.

The results of the data maturity assessment should become the input for possible upcoming programs at the Data Forum. We plan to conduct the Data Maturity Assessment on a bi-annual basis to quantify the impact of Data Governance programs.

3. Data Forum

The Data Governance Team empowers the Data Governance Office by forming a Data Forum. At the forum, the team gives technical support, operationalization, and assistance, as well as running DG Initiatives. Examples of regular topics during this forum are (but are not limited to)

• Reviewing the support from the Data Governance team to the DGO.
• Aligning future possible support needed from the Data Governance team to actors at the DGO.
• Aligning the overall data governance objectives and ensuring the data governance execution is coherent with the strategic plan in Data Governance roadmap
• Synchronizing the progress of the deployment of data governance initiatives in a business unit

The Operationalization of Data Governance in GovTech

Once the three fundamentals are in place, the next challenge is actually to operationalize the initiatives. A Data Governance Specialist formulates the initiatives by making the standards and upscaling the capabilities, while DGO actors will execute them. To better understand the actual technical implementation, here are two examples of operationalization of Data Governance.

We elaborated on each of those activities in the supplementary material.

Closing

The implementation of Data Governance in an organization requires significant effort from cross-functional teams, and the process involves change management. Hence, support from upper management is highly recommended before socializing and operating the data governance framework in an organization. It is worth noting that the data governance implementation might differ for different organizations. There are some factors that might affect the framework and the programs of data governance in an organization, such as the data maturity stage, the organization structure, the business model, and also the product roadmap.

References

• Susan Earley, et al. (2017). DAMA-DMBOK Data Management Body of Knowledge, Second Edition. Basking Ridge, NJ 07920 USA: Technics Publications.
• Utsav Mishra. Best Data-driven Companies in the US. Analytic Steps, 6th December 2021, https://analyticssteps.com/blogs/best-data-driven-companies-us, accessed on 15th October 2022.
• George Firican. Stanford data governance maturity model, https://www.lightsondata.com/data-governance-maturity-models-stanford/, accessed on 16th October 2022.
• A Guide to Data Driven Decision Making: What It Is, Its Importance, & How to Implement It, https://www.tableau.com/learn/articles/data-driven-decision-making, accessed on 17th October 2022.
• The Scaling Data Framework: Data Informed to Data Driven to Data Led. https://www.reforge.com/blog/scaling-data#three-stages-scaling-data accessed on 8th November 2022