Before we explore the importance of data stewardship, let’s take a step back and consider the term 'Steward'. A steward is someone who looks after something on behalf of another. A steward cares for that which is entrusted to them.

So, a Data Steward is someone responsible for the care of the Data in the organisation—they are the caretakers of the Data Assets.

Everyone in the organisation should strive to produce quality data. However, the Data Stewards bear the responsibility for ensuring that this occurs and for addressing the situation when it does not.

Why is the role so important?

Data Stewards play a crucial role in the hands-on management and governance of data within the organisation. They act as custodians to ensure that the data is accurate, consistent, accessible, and secure throughout its lifecycle.

Data Stewards serve as the eyes and ears, the feet on the ground, and the heartbeat of a data-driven organisation.

Where to find Data Stewards

The organisation is filled with potential Data Stewards. All you need to do is identify them, nurture them, inspire them, and, most importantly, educate them.

Data Stewards are found in all areas of the organisation, including Business and IT. They are usually chosen because they are Subject Matter Experts (SMEs) and have expert knowledge of their Data Domain. Then, they need to be enabled to perform the role to the best of their ability.

What do Data Stewards do?

Data Stewards must implement mechanisms to ensure adherence to standards and processes. They play a role in defining Governance Controls to ensure compliance with ethical and regulatory requirements. 

They define the quality expectations of the data – the rules, targets, and thresholds – and they monitor and report on improvements to data quality. Data stewards must manage what occurs in the event of data issues, posing questions such as: How should we handle this? How should we escalate this? How should we resolve this? How can we prevent this from recurring?

Here is a breakdown of what Data Stewards typically do:

• Data Quality Management
  • Monitor and improve data accuracy, completeness, consistency, and timeliness.
  • Identify and resolve data quality issues (e.g., duplicates, missing values).
• Data Governance Implementation
  • Enforce data standards, policies, and business rules defined by the data governance team.
  • Ensure compliance with regulatory requirements (e.g., GDPR, POPIA, HIPAA).
• Metadata Management
  • Maintain data dictionaries and business glossaries.
  • Document data lineage (where data comes from, how it’s transformed, and where it goes).
• Data Lifecycle Oversight
  • Track data from creation to archival or deletion.
  • Ensure proper data retention and disposal practices.
• Projects and Data Initiatives
  • Participate on projects and working groups to provide insights into the Data and ensure all aspects considered, e.g. Data Lineage, Data Issues, Data Governance requirements etc.
• Data Privacy and Security
  • Understand Data Privacy requirements
  • Ensure Privacy and security are applied in all environments e.g. Operational, Archive, Disaster recovery, Data Warehouse
  • Oversee implementation and testing of Data Breach response
• Collaboration with Stakeholders
  • Act as a bridge between business units, data owners, IT teams, and data governance bodies.
  • Translate business needs into data requirements and vice versa.
• Issue Escalation and Resolution
  • Identify data-related risks and escalate unresolved issues to governance committees.

In the past, things were simpler. All you had were the operational Data Environments. Times have changed; now there are Data Warehouses, Big Data, Data Lakes, Data Lakehouses, and Data Meshes. You also have Reference and Master Data Management hubs, Data Integration and Data Migration, along with new architecture requirements.

The Data Steward must contend with numerous demands related to data and respond to questions they may not have even considered. They are managing more data than we ever imagined. Furthermore, they now face the challenge of how to care for the data that AI and digital transformation require. 

The role of the Data Steward is evolving to encompass collaboration, enabling, and facilitation as some traditional tasks are managed by AI.

Of course, Data Stewards can also use AI to assist them with some of their tasks!

What Data Stewards need to know?

Here are just some of the skills required by Data Stewards:

• Strong understanding of business processes and how data supports them;
• Knowledge of Data Privacy laws and compliance standards (e.g. PoPIA, GDPR);
• Attention to detail and problem-solving ability;
• Excellent communication and documentation skills;
• Understanding of Data Quality requirements;
• Familiarity with data management tools (e.g., Collibra, Informatica, Alation).

With data underpinning everything the organisation needs to do, the role of the Data Stewards is more crucial than ever before. What efforts are being made to identify and support the Data Stewards in your organisation, and to equip them to address any challenges that future innovations may present?

In uncertain times, organisations often delay projects due to limited funding.

But you don’t have to come to a complete standstill.

Instead, you can focus on maximising what you already have by doing more with your current team and existing data to achieve new improvements in your business!

One critical area to address is Data Debt.

What is Data Debt?

• Data Debt is a technical debt that specifically applies to data created when individuals and teams fail to manage their data effectively
• It arises when shortcuts or compromises are made to quickly access data, leading to poor data quality, incomplete information, and inconsistencies.
• It results from not correctly cataloguing, or categorising data, leaving it undocumented, unused, or inconsistent.
• Data Debt results from poor data quality such as missing data, duplicate data, incomplete data, inaccurate data etc.

Why is it a problem?

• Data Debt can lead to unreliable data and manual data management, increasing costs and decreasing efficiency.
• Poor data quality can erode trust in analytics and compromise security.
• It can hinder productivity, data-driven decision-making, and slow down innovation.

Data Debt is the hidden cost of neglecting data quality and management. It can significantly affect an organisation's ability to operate efficiently, innovate, and make informed decisions.

Even with limited budgets, you can do something about this.

When you train individuals who generate data to manage it accurately, the entire organisation will benefit from less rework, enhanced productivity, and higher quality data.

Education is particularly crucial when budgets are tight.

At a fraction of the cost of the significant transformation initiatives that have been postponed, you can focus on improving knowledge and skills and still find opportunities to improve the effectiveness and efficiency of existing business processes and the supporting IT systems.

Train your people, and your data improves; improve your data, and your business improves.

P.S. If Artificial Intelligence is in your vision, Data Debt can also impact initiatives like generative AI, as the quality of data used to train these models is crucial.

What you do with data depends on the job you do and how you do it.

What job do you do?

• Is it your job to respond to customer queries or complaints?
• Is it your job to prepare reports?
• Is it your job to increase sales?
• Is it your job to put out fires?
• Is it your job to  <<fill in the blank>>?

What does every job in your organisation have in common?

Have a guess. It's easy - it's data!!

Data is central to performing our jobs, yet no one teaches us enough about the data we use and produce daily, or why its quality is important.

What do you need to know about the data to do your job better? 

You need to know:

• What the data means and how it is used
• Who produces this data and who uses it
• That the data is correct and of good quality
• The data is up to date and trustworthy.

If you knew this stuff:

• You would do your job better and make less errors
• You would you be able to be more productive
• You would face less challenges and enjoy your job more
• You would be more valuable to the organisation.

So, we go back to the question, what are you doing to the data?

Let's improve our data skills, data literacy, and data quality awareness, and grow your data culture.

I remember when Data was …

Way back in 1968 when I started my career as a junior programmer – our Data was easy! We had text and numbers. That was it! No images, no videos, no audio clips. Just text and numbers.

All those years ago, we communicated with our Data in a very different way. Punch Cards, Paper Tape, enormous Magnetic Tapes and huge Disk Drives. Can you believe it was even more exciting? We could read the Data in its internal format. We all spoke hexadecimal. (You may have to google what this is). Hexadecimal was just one of many languages we had to learn to process and understand our Data.

In those early days of computing, we had 3 priorities – Requirements, Data, and our Programs.

I remember when Data was easy. I remember when Data was cherished, I remember when we were all responsible and accountable for the Data.

What's changed in the last almost 60 years? Well, I have got a lot older and we have lost or forgotten some of the key principles of managing our Data. Those key principles were; understanding the meaning of our Data, understanding the importance of our Data, identifying all the rules around the Data and validating against those rules before we stored the Data.

I remember when Data was understood consistently. I remember when Data Rules were defined. I remember when we made no compromises.

Have I got you thinking? Do you think we can recover some of the good stuff we used to do and combine it with the smart stuff we are now able to do? Let’s all remember the right stuff we should be doing.