Executive Insights: Data Governance – The starting point to becoming a data driven organisation!

The need for Data Governance

Data Governance is defined as a set of processes, policies, roles, and metrics to ensure efficient use of data and information across the organization to drive its business goals. The efficient and effective data management of an organization is driven by the data governance framework that is defined.

Data Governance is needed in an organization for efficient use of information and data. In Data and Analytics, the need for data governance is important to validate and make efficient use of the outcomes and consumed to its full potential. Outcomes and Target variables need to be validated and approved by decision-makers for a successful Data Governance program in Data and Analytics.

Challenges faced with implementing Data Governance initiatives

One of the most important challenges faced with implementing a good Data Governance strategy for Data and Analytics is identifying the key areas to track with regards to the long-term goals of the organization and a lack of support from leadership towards those areas. A good governance initiative is driven by the business and tied closely to the business goals of the organization. A lack of understanding regarding the need and the value of a good Governance strategy, lack of the skills and knowledge in relevant technologies coupled with lack of harmony across various business areas to arrive at an agreement are some additional challenges. A common mistake that most organizations make is to orient the Data Governance strategies towards the data instead of the business. This causes a lack of alignment of the initiatives with the business goals. Identifying the key areas of interest that influence the organizations long term goals along with gaining approval from the Business Leadership to focus/track these areas is the most efficient way to initiate a good governance program Governance policies and frameworks need to be developed in tandem with the use cases that they will cater to. Building an isolated governance framework with the use case definition may not be the right approach.

5 steps required to implement a good Data Governance Program


The first step is to define the metrics, key drivers, and policies for the governance program. How the outcome of a Data Science use case will impact the business outcome is an important step in the definition. Various options are evaluated, and outcomes are prioritized. Accountability and decision owners/rights are defined for each driver, metric, outcome attributes, etc. It is always a good practice to implement large programs such as the Data Governance programs in a phased manner.


After a well-defined framework for the program, the next step is efficient planning and deriving a good operating model. The first step in planning is an assessment of the current maturity level of the organization against the defined business drivers, rules, policies, etc. laid out as part of the framework. The next step in planning is defining the operating model on what to govern (the metrics, policies, rules), how to govern (monitoring through various approaches such as data lineage, security, etc.) and who will govern (the data stewards). There are different operating model strategies such as a centralized strategy where the decision making, authority, and responsibility lie within a central body and federated strategy where the authority lies within individual business units and the collaboration across various units are encouraged. In the real world most organizations follow a hybrid approach.

For a good Data Science focused governance initiative, it is suggested to follow the three-ringed model – Inner Ring (governs Master data/fewest attributes), Outer Ring (Data used by the fewest number of applications), Middle Ring (everything else in between). The challenge with these three-ringed approaches is that the middle ring becomes a bottleneck as all application data resides in this ring and alignment/agreement across all business areas may pose a challenge.

Once the operating model is defined, the next step is in defining the deployment approach and setting up the governance footprint. Data footprint is an important aspect of the governance programs as it provides valuable information about the data and its traceability.


Execution begins with a detailed analysis and assessment of the impact due to gaps identified in the organization’s maturity with the business goals. If there are overlaps and/or conflicts in definitions, they are identified too. The next step to is engage with the key stakeholders to assess the content and define/redefine processes that bridge the gap. The deployment plan is implemented along with setting up of the necessary workflows, compliance, and reporting the outcomes


A successfully executed program needs to be monitored for efficiency. Monitoring is done by assessing operating behavior and workflow processes against expectations and checking for anomalies, defining thresholds to drive behavioral change. Thresholds also help in evaluating the impact of the business policies against the actual outcomes and identify areas of improvement.


Optimization is a crucial aspect of the strategy as there is continuous learning and enhancements to enable an efficient program. This is done by tracking data-related issues and routing to appropriate resolution pipelines. Where needed, new processes are devised and the strategy is reassessed, controlled testing is done on a subset of the metrics before releasing across for enterprise-wide adoption

Data and Analytics governance programs lay more emphasis on some of the data elements such as the

  • —Data Architecture
  • —Data usage and accessibility
  • —Data quality and classification
  • —Master data
  • —Metadata
  • —Target reporting metrics/attributes
  • —Compliance metrics,
  • —Data retention


Teams behind a successful governance program

The Data Governance initiatives are driven by the CIO, supported by various teams across the organizations, that include:

  • —Application/Software Engineering Team that inform the strategies and provide guidance for enterprise application
  • —Data & Analytics Team that establishes the framework for Data and Analytics governance
  • —Program/Portfolio Management Team that defines the program plan for successful execution of the governance initiative
  • —Infrastructure and Operations Team that guides on the infra and the operational changes need for the initiative
  • —Security and Risk Management Team that works with Data and Analytics leaders to ensure that risk and security implications are understood and defined  appropriately
  • —Tech professionals who Architect and deploy Data Management strategies to improve data quality, increase Business Analytics and Intelligence

It is the primary job of the Data and Analytics leads and their teams to define an audit trail on the decisions made, actions taken, investments, expenditure, and compliance for the governance initiatives.

Training & Education

For successful implementation and sustenance of Data and Analytics governance programs, the people across the organization that are part of the initiative need to be trained on the value of the programs. Skills needed for the programs are identified by analyzing the roles and mapping against the skills needed for those roles. Preparing focused training modules and ensuring consumption of those modules can enable employees to get on board into the programs. Cross teams’ collaboration is also a valuable method to share knowledge and train.


The success of a Data and Analytics Governance program is supported by “People, Process and Technology”