What are the best problem-solving techniques for identifying data governance gaps?

In my experience, its rare to have a "pure" data governance problem. Instead, you have business problems where data governance is one of the capabilities that you can use to solve the problem. As examples: - you are unable to accurately track customer retention - your cyber security costs are too high because you hold redundant personal data - you have an inefficient business process that is causing poor customer experiences You need to define problems from a business perspective. That ensures you will later develop a holistic solution where data governance may play its part along with other teams and skillsets in your business. Avoid looking for pure data governance problems or defining problems solely through a data governance lens.

In my last role, I remember really hitting a wall with our data issues. It wasn't just that we had to catch and fix the errors; we needed to uncover the root causes. We had problems like data definitions that didn't line up and no data entry standards. Root cause analysis was my go-to technique. It cuts to the core of each issue. For instance, when faced with inconsistent data, instead of just correcting errors as they pop up, root cause analysis prompts us to ask: "Why are these inconsistencies recurring?" It often lead to enlightening discussions, unveiling not-so-obvious reasons like inadequate training or flawed entry processes. We turned from reacting to proactively preventing issues, ensuring data governance was not just a quick fix.

Organizations need to take an infinite mindset approach to managing data in the organization. In my experience the #1 challenge is ensuring that data issues are captured. Too often, data issues are hidden from view, dealt with within silos, with data heroes taking the stage applying band aids and not dealing with the root cause. Train your users to capture data issues like they would any other type of issue. Raising visibility to data issues in a consistent and holistic manner is a #Dataninja super tip. Make it part of the expectation. Treat Data Quality like Water quality and report any deviations.

To identify data governance gaps, it’s like playing detective. You need to look for clues that something isn't right. This could be inconsistent data, difficulties finding specific data or too many errors. Firstly, carry out a data audit - taking a thorough look at your data to spot any issues. It's like doing a health check-up for your data. Secondly, conduct surveys or interviews with your team . Ask about their data challenges – it’s like asking if they feel any ‘pain’ when working with data. Finally, create a data map (preparing for implementation phase), outlining the data journey from collection to usage. This can help you 'visualize' where problems exist, and solve them effectively. The journey is a road trip with several pit stops!

The biggest mistake in modern day companies is not utilizing the tools that you have. There are new technologies being created everyday that can help take away human error but time and time again I see people ignore the technology to do it themselves which creates more errors. Excel macros are a thing of the past. Learn new ways to streamline your analysis, use the tools you have to inform the data and stop trying to insert your manual process on modern technology.

Identifying data governance gaps is akin to detective work. You must examine the data landscape critically, tracing the lineage of data as if following clues—where it originates, how it moves, and who interacts with it. By using tools like lineage diagrams, you can visualize the data’s journey, pinpointing where the trail goes cold. Data catalogs act as dossiers, revealing what data exists and its attributes. Quality reports highlight the inconsistencies, errors, or omissions—much like inconsistencies in alibis. And stakeholder maps show who has a vested interest in the data, guiding you to interviews to understand the human elements behind the numbers. Thorough analysis not only uncovers the gaps but also the story behind them.

Root cause analysis is what typically done by Data Governance practitioners to identify the possible causes. All possible causes could be identified and possible ways to eliminate them happening again should be discussed before move on. The causes might be missing standards, unfollowed policies, gaps in the business processes or others.

In tackling data governance gaps, think of yourself as an architect designing a blueprint for data integrity. Gather a diverse team for a brainstorming session to leverage collective intelligence—your crew of experts in legal, IT, business, and data science. Look beyond your walls; benchmark against industry standards and competitors to identify best practices you can adapt. Weigh each potential solution with a cost-benefit analysis, ensuring the foundations you choose not only fill current gaps but also support the structure of future data endeavors. The goal is a robust data governance framework that's both resilient and flexible.

It is absolutely critical to engage stakeholders across the data ecosystem in solutioning discussions. Technical teams often come up with elegant tech solutions to filling data governance and data management gaps, but if those solutions are designed without the data users, stewards, and owners the table, successful implementation will be very limited.

Companies can also consider: 1) Continued maintenance of the solution- cost, resources, integration, time etc; 2) Priority - eg solutions that focuses on data encryption v data quality 3) Competitive Advantage - Will you stand out ? 4) Effectiveness - consider if it addresses the gaps, is it comprehensive,if it requires regular adjustments and updates.

After the causes are identified and possible solutions are discussed, the solutions should get implemented and monitored closely whether the solution prevents the issues happening and/or the implemented solution causes some other issues. A dashboard can be used to monitor the progress which will be helpful to understand by stakeholders.

A clear and unambiguous data ownership is the very first step towards having a robust data governance at any organization. Then, comes the understanding of overall data ecosystem, it's impact on the information needs of the organization and creating data policies to cater to the consumption needs of it.

Utilize a gap analysis framework to systematically compare your practices against best practices. You will find areas of improvement along this exercise

Assuming that the DG operating model is setup with the Roles, responsibilities, defined process flow for each of the data swim lanes namely Data Quality, Data integration, Creating new dataset, lineage etc. are place then the following steps will help.. 1. Once the issues is identified for example Quality is an issues, then do a value stream mapping across the DQ process that is defined. 2. This process will help identify the bottleneck in the process and the delay. 3. Once the root cause is identified for example the rules are not right, then work with the business to modify the DQ rules and run the job. 4. Do a continuous monitoring of the process using value stream mapping and optimise as required

1. What are the business values that need to be derived from data and at what points of data lifecycle? 2. Who are data owners, consumers and intermediary (in data value chain)? And do they know how value is created? Do they know how one data action creates domino effect in entire data value chain? 3. Is data understood by everyone? E.g. a product means a uniform entity across organization - case in point a "30 Year Fixed Interest Mortgage" is a product, not Mortgage (as a general category). 4. Do tools (any tools) give data governance (or just data) team to prove value to key stakeholders? Enable them to find gaps/issues and drawbacks in existing infrastructure?