Everyone has heard of it, but is everyone doing it? Data Governance is the act of governing your data or managing it with guidance. No matter who you are or what you do, data should be important to you! Let’s see if you can answer a few simple questions: Where is all of your data? What are people doing with your data? What type of data quality do you have? How long do you keep your data? If you can’t answer those questions, then these are good motivations to dive in!
We all know how important data is to us personally and to our company. In my experience, there is no ‘right way’ to implement data governance, you just have to start. Not to be clique here, but you need to start with three things; People, Process, and Technology. Build a Data Council that is staffed by business personnel who are the authority for the definition of their data. Build Data Governance Processes that are easy to follow and don’t slow the business down. Procure an Enterprise class Data Governance tool, which minimally will be your source of truth for business terms and data assets.
"The success of data governance is typically as simple as getting people to do the right thing with their data"
The success of data governance is typically as simple as getting people to do the right thing with their data. Much easier said than actually done! Data needs to have distinct ownership by the business. For all business processes run on data, the business needs to know what they need and when. Once the business realizes that they need to partner with technology to define their critical data, the collaboration becomes addictive. Coveted characteristics in data governance personal are honesty and authenticity.
Oftentimes business partners need to understand the benefits of data governance; the ‘what’s in it for me’? Here are a few successful ways to grab their attention:
- Democratize your data – make it easy to find
- Decrease your risk profile – the data will be labeled, it shall be traceable, and properly retained according to privacy and security guidelines. This will reduce potential data misuse, once you have implemented data usage guidelines.
- Reduce your cost – if you procure data, you can enforce data standards upon receipt, which will eliminate rework and technical debt. Data Governance helps people understand their data, which reduces confusion. Many times, it takes a lot of time to find the right data – to understand what it ‘means’ – and feel confident that it is what you are looking for.
- Finally, Trust–by operationalizing data governance you will increase trust in your data. You will enable certified data to run reports on.
Ideally, business processes will be defined first (note: I’ve never seen this precede data governance anywhere). Therefore, it is on the governance team to develop processes that are repeatable and non-disruptive. A few that have been successful are:
- Information Technology engagement
- Data certification process
- Develop guidelines that will drive engagement
- Roadmaps and scorecards
The best way to operationalize (deeply ingrain data governance into everyone’s daily life) is to embed your work into already successful processes like your IT project portfolio. Likely your IT projects are run using a software development lifecycle process (waterfall, agile, Kanban, etc). The data governance team needs to create light weight tasks that are adopted into the IT project development lifecycle. For example, in the planning phase of a new IT project, there should be a task related to listing the data that will be needed to support the new project.
Once the data governance council has collaborated and agreed upon the standards for their critical data, it’s time to certify ‘sources’ to those data standards. Once you start to actively work with IT application owners, they will see the value in using data standards that the business has agreed on. It is recommended that you start certifying the data at the source of truth (or data hubs) and work towards the source systems, and then the consuming systems. Over time, there will be apparent increase in data quality because data will be passing from system to system using the same standards! Once all sources are certified, we will benefit from significantly improved data quality in our data flows and reporting.
Author and implement guidelines that embed data governance into processes. For example, implement a guideline in collaboration with a procurement organization related to purchasing data. If data governance is engaged in external data procurement, the data standards can be implemented for the procured data. Contract language related to data quality can be also influenced. Another successful example is to create a guideline that will enforce engagement with data governance any time someone wants to exchange data. This will help the data governance team to keep their finger on the pulse of data movement.
As a data governance organization (formal or virtual) you need to develop roadmaps and scorecards. Roadmaps will demonstrate what work will be accomplished and when it can be expected. Scorecards help to drive accountability and show successful execution! Publishing roadmaps and scorecards drives relevance and increases your ability to professionally communicate the objectives.
In order to support long term growth for the data governance organization, technology that minimally supports a central Business Data Glossary, Data Lineage, and Data Quality dashboards will be essential. Having a central location for the business glossary will create a ‘source of truth’ for the data definitions and data standards. Data lineage becomes important when people start to understand the relationships between their reports and data quality. Data quality will become a hot topic once you are able to display it.
In Summary, data governance is superbly important regardless what line of business you are in. Start with the people – build a team of like-minded people who agree with your mission. Roll out some processes–typically documented processes are the ones that work, and they give the illusion of having your act together. Finally, adopt a technical solution that will support all the good work that the data governance team would execute. This will be key to an organization’s success.