Demystifying Domain-Specific Data Governance

Domain-specific data governance has been of focus lately in various industries. In this article, I simplify what it means and how it is done.

What is Data Governance? 

If you ask twenty people in a room what data governance is, you might get twenty different answers. None of them might be wrong, which is the beauty of it. The definitions can range from quite simple to complex, based on the individual’s knowledge, experience, education and how data governance takes shape within their own organization organically. While the basic data governance framework is industry-agnostic, it is also not a one size fits all and should be flexible based on the business need at hand. In simple terms, data governance is the art of aligning people, process, and technology to ensure data is fit for purpose, secure and easily accessible to authorized users.

What is a Data Domain?

Data domain refers to a logical grouping of items of interest to an organization. Data domains are also referred to as data areas. Some examples include Employee, Product, Client, Customer, Consumer, Vendor, and Facility.

What is Domain-specific Data Governance? 

Domain-specific data governance means exactly that i.e., governing data for one or more data domains within an organization. Here, the prioritized data domains are in full focus and are tailored to in terms of governance, rather than the generic enterprise-wide data governance. These data domains are governed consistently across an organization’s business lines, functions, and products.

What are the Advantages of Domain-specific Data Governance?

Domain-specific data governance helps an organization to focus on the most critical areas of interest or problem areas right away, and keeps it customized for the domain(s). This helps accomplish target business outcomes quickly, and in turn promotes higher acceptance from the business. But this should not replace enterprise data governance as domain-specific data governance would still need to be within the realm of high-level enterprise-wide data governance policies, standards, procedures, and guidelines.

How Do You Establish Domain-specific Data Governance from Scratch?

Here are some guidelines for Domain Data Governance Leads and Leaders that are new to the space, to establish domain-specific data governance ground up. Try to keep it simple, and sustainable.

  1. Secure Senior Leadership Support
    1. Understand the triggers/need to establish data governance for the domain(s). Assess the current state of governance, required maturity and what it will take to get there. Leverage any standard data governance maturity models out there or create your own.
    2. Make a compelling business case with supporting data from above analysis. 
    3. Secure senior management support for the program at various levels. While you are taking a bottom-up approach to establishing data governance, you need top-down support from senior leadership to get business and other stakeholders’ engagement, interest, and support.
  2. Conduct Business Discovery
    1. Start by setting up meetings with key stakeholder teams for the domain(s). Understand the gaps in terms of people, process, and technology, and what it takes to address the gaps. 
    2. Gather as much business knowledge as possible for each domain. Understand the big picture for the domain(s), what makes your domain(s) unique, and where you need to tailor governance activities for each. This will help with formulating your strategy on how to establish governance in the domain(s) space and in developing a roadmap.
  3. Gather/Create Domain Inventory
    1. Identify the systems of record for the domain(s). Identify/define authoritative sources that can be used across the organization for the domain(s).
    2. Create inventories of master data, reference data, upstream and downstream systems, business owners, etc. Gather/ask to create or update data dictionaries, data models and system integration diagrams. 
    3. Work with key stakeholder teams to identify Critical Data Elements (CDEs) for the domain. Understand rationale for criticality, business processes that those elements touch for the team(s) and any applicable regulations. Regulatory CDEs get higher priority in terms of governance as regulatory requirements are non-negotiable.
    4. Create Business Glossary, starting with prioritized CDEs. Have working group meetings to gather business definitions, data quality rules, metrics, targets, and thresholds. This helps realize a significant advantage of domain-specific data governance i.e., to establish common and consistent set of definitions, data quality rules and measurement across businesses/ functions within the organization, providing a holistic view of the domain(s).
    5. Gather process maps for key business processes creating or modifying key domain attributes. This will help with resolving data issues, identifying roles and responsibilities, establishing standards, etc.
    6. Leverage tools available in the market for building enterprise and domain data catalogues, data lineage, data quality management, domain dashboards, etc. Having integration between the systems can help with faster user adaptability.
  4. Formally Identify Data Roles and Responsibilities
    1. Identify key data roles that should be in place such as data owners, sponsors, data stewards, consumers, data custodians, etc.
    2. While it is debatable if a certain role like data owner is required or not, it is important to understand what makes sense for your organization, for your domain(s), and pursue accordingly. Having a domain data owner helps with having an overall vision, strategy, and guidance for the domain data across the organization. You could have a committee of data owners a level down, identified by a specific criterion like functional area, working in alignment with the domain data owner. 
    3. Identifying domain-specific data stewards helps with bringing them onto a common platform, especially when they are located in different businesses. 
    4. Outlining the responsibilities, going over the same with the people identified for the key data roles, and getting mutual understanding and acceptance is key to forming a structure that will provide transparency, trust, work, and sustain.
    5. As Bob Seiner (the TDAN.com publisher) mentioned in his book on Non-Invasive Data Governance, it is most likely that people are already assuming the data roles and doing the responsibilities in an informal manner. What data governance does is make this formal and put a structure around it so there is clarity on who is responsible for what from a data perspective and communication is streamlined. This should be done in a non-invasive manner as how this is done often determines the success of the program. Explain to the members assuming data roles how they are already fulfilling the responsibilities, the value they are bringing to the organization, examples of where it will benefit them, the other cross-functional stakeholders, and the organization, if this is formally recognized.
  5. Establish Domain Data Governance Committee
    1. Break the Barriers
      1. Understand the organizational and data silos that exist. Break the silos by establishing a governance committee with key data roles across the organization for each domain. Domain committees should be dotted line to the enterprise data governance council. Establish a regular cadence.
    2. Speak the Same Language
      1. Get approval on the business definitions and make them available in a centralized repository accessible to everyone within the organization. This helps with establishing a common understanding of the business terms and in speaking the same language. 
    3. Establish Standards
      1. Establish policies, standards, procedures, and guidelines to manage master data, reference data, data quality, and define required controls at each stage of the life cycle for the domains. This will help with ensuring good data quality from get go, provide clarity on required data cleanups for existing data, and to prioritize those projects. Note: While it is great to have domain specific data standards, ensure you have well established enterprise-wide standards and that the domain data standards fall within the realm of those standards.
      2. Develop implementation plans to execute the standards effectively.
    4. Support Business to be able to Govern the Data
      1. Data governance team enables governance by bringing people, process, and technology together, but the actual governance of data is mostly done by the business. Educate business stakeholders on data governance and provide required support for them to be able to do this effectively.
      2. Ensure the committee is business focused, with enough support from IT and other supporting/ enabling teams.
    5. Have a Solid Roadmap
      1. Collectively formulate the roadmap for the domain governance and regularly report on the progress.
      2. Have both soft and smart metrics to show the success of the data governance program.
  6. Influence Culture
    1. Look for opportunities to influence the culture to be data-driven and for data to be treated as an enterprise asset. Provide support to the user community to build/upskill their domain data literacy as required, according to the data role they are playing. This will help them with understanding the data, the value it brings to the organization, and using it to make things better within the organization.
    2. Provide easy access to data to authorized business users using systems of reference and provide training on data reporting, visualization, and analytics tools. When they are able to slice and dice the data as required, they can use it effectively for operational reporting, decision making, and to identify data issues easily and early.
    3. Effectively running committee(s), early successes with the program, and defining standards can help with influencing a culture where a proactive, rather than reactive, approach is taken for data.
  7. Data Governance Reviews
    1. Establish a process where critical development items/ database changes are reviewed and approved by the data domain committee(s), while ensuring this does not become a bottleneck for the required changes.
    2. he objective is to establish a recurring process where different perspectives are considered from data side before implementing the changes. 
    3. Before providing approval to proceed:
      1. Ensure that the standards established are being adhered to, with the proposed implementation.
      2. Identify any gaps and obtain remediation plans from the requesting teams.
  8. Data Quality Monitoring and Dashboarding
    1. Here the focus is on monitoring the data quality for each domain, across business lines, products, functions, etc., and providing a holistic view of data quality for each domain using dashboards.
    2. Utilize the data standards defined and the data quality rules, metrics, targets, and thresholds collected for the prioritized CDEs, to know what exactly makes the data fit for purpose for each domain. Confirm your understanding within the domain committee(s). Remember that this is point in time understanding. Future standards, requirements and what makes the data fit for purpose can always change so be prepared to change as per the need. Agility in terms of adapting to changes will help to take the program further.
    3. Establish a process for measuring & monitoring data quality and create simple dashboards that show data quality clearly. Baseline the data quality as of a certain date and show improvement over a period of time.
    4. Review the dashboards with the committee members to ensure all required metrics are captured, prioritized, targets and threshold limits are specified, monitored and data quality improvement plans are put in place.
    5. Along with required data cleanup projects, leverage any existing and recurring audit processes to ensure that the data is accurate on a go-forward basis.
    6. Sometimes, bad data quality could be purely process related. In these cases, having certain controls in place should help.  
  9. Establish Cross-Functional Partnerships
    1. Establish partnership with Information Risk & Security, Enterprise Architecture, Business Process, and other shared services to assure these other areas understand the data domains you have prioritized for governance, how their frameworks interact with the domains and to ensure there is no redundancy or conflict with each other’s standards.

In summary, domain-specific data governance offers significant advantages to an organization, helping with targeted, tailored, and quick focus on areas of interest in terms of governance, providing a holistic view of domain(s) across the businesses, functions, products, etc. in terms of business definitions, data standards, data quality rules & metrics, alignment between domain data stewardship processes, alignment between shared services for the domain(s) and more.  

I hope this is helpful for you in getting started on your journey. I would love to know how you established data governance within your organization and if you would like to share any tips and tricks.

Share this post

Lata Garlapati

Lata Garlapati

Lata Garlapati has worked in several industries including Financial Services, Healthcare, Retail, Manufacturing, and Travel, and has spent about 9 years in data governance leadership roles. Lata is highly skilled in establishing enterprise data governance and domain-specific data governance from scratch. Lata is a versatile professional with grit and passion for data governance. Lata strongly believes data is one of the most valuable assets of an organization. Lata holds Post Graduate Diploma in Planning and Management, with specialization in Finance and Marketing from IIPM, India and Bachelor of Technology (Mechanical), with specialization in Industrial and Production from KL University (formerly Koneru Lakshmaiah College of Engineering), India. Lata lives in New Jersey with her daughter, who studies at Rutgers Business School. Lata is an avid gardener, and she loves travelling, playing carroms, and doing social service. Lata currently works for Quest Diagnostics. The views expressed in this article are Lata's own and not those of Quest Diagnostics.

scroll to top