MVP in data governance
Petrus Keyter, Data Governance Consultant at PBT Group
In my previous article, I discussed some of the key principles of a data governance strategy. Now, the focus turns to the topic of minimum viable product (MVP) in data governance. The MVP has evolved significantly in recent years but for it to be effective, there must be top-down support. Here are some of the principles to consider when working on an MVP in data governance.
1. Data governance team
Any company in the world must have a proper data governance team in place. This team is responsible for driving this initiative and executing data governance policies. Typically, a team will consist of a Data Governance Manager or Chief Data Officer. There will also be Data Owners, Data Stewards, and Data Custodians.
2. Data governance framework, policies, and guidelines
There must also be a proper data governance framework in place that outlines and supports the goals and objectives of any data governance initiatives inside the company. With this in place, the data governance team can create effective policies. Furthermore, a Data Governance Committee can also ensure policies, processes and standards, and that the implementation and execution of data governance initiatives all align.
3. Build a data inventory
Data transparency and data quality are essential components of an MVP in data governance. This requires the company to build a data inventory or catalogue of what it already has in place. For instance, what data does the business have of its customers and on the products and services that it sells? This information must be available in a central place for employees to get answers to the questions they have.
A company must also catalogue all critical data assets. This includes definitions, business terms, classification, and business rules to ensure all employees speak about data in the same way. The data inventory should also map data flows to identify any integration points and dependencies.
4. Basic data quality measures
From there, a business must establish the baseline data quality metrics that cover the core data quality dimensions. These encompass elements such as accuracy, completeness, consistency, uniqueness, timeliness, and validity. It is here where the company needs to introduce basic controls and checks which will be supported by a well-defined data remediation process.
5. Data security measures
Of course, there must be data security measures implemented. Given the significant financial impact of fines for data security breaches as well as the associated reputational damage, businesses need to prioritise the safety of their data.
They therefore need to define sound access controls to all data in the business and implement basic security controls. These include encryption, secure access protocols, identifying which employees have frequent access, and conducting data usage audits.
6. Performance metrics and reporting
As the saying goes: “you cannot manage what you cannot measure”. With that in mind, the company must define KPIs to measure the effectiveness of its data governance programme. It must measure activities at a specific point in time and use the findings to realign data initiatives.
There must therefore be some form of metrics in place for companies to implement and report on their data governance activities. The resultant reports must bring value to all stakeholders in the business, from the executives and data stewards to any other employee requiring data governance.
7. Training and awareness programme
None of this will work if the business does not create educational material about data governance, including all core principles. It must educate all staff on data governance and their role in maintaining data integrity and security. Yes, the business can show metrics and reports to everybody, but employees must understand the detail and how the company got to it. Telling a data story through training and awareness becomes essential. Ultimately, people will not see the importance of these reports and data in their world if they do not understand its value.
If a business is to ensure a successful and sustainable data governance programme, the MVP must be supported by a well-defined implementation approach. The company should start with a small and iterative implementation approach and then go back to the metrics and value. Once these are measured, the MVP must evolve.
