What is Data Governance? And, why is it important?

What is data governance?

“Data governance” is a term that encompasses the array of tools and techniques that an organisation uses to manage and assure its data assets. This includes policies, strategy, processes, data quality and security standards, roles and responsibilities and issues management.

How is data governance different to data management?

Data management and data governance, as terms used across the industry, experience some overlap. However, generally, data governance refers to policy, accountability and decision-making, and data management refers to execution and operations. Data governance supplies the motives and reasoning; data management is the day-to-day activities driven by governance.

What makes data governance important?

Data enables business activities. It provides information for business intelligence, enables marketing and sales activities at scale, helps reduce waste and streamline operations and offers proof of performance.

But data is not usually valuable unless it can be used by someone. Because data governance is the decision-making infrastructure that makes data usable, it is essential for the ongoing enablement of any data-driven initiative undertaken by an organisation. Right now, it has received renewed attention from business leaders because robust data governance is critical to the successful deployment of AI tools.

More specifically, data governance is important because:

It improves trust and quality.

An organisation with strong data governance is one with a better knowledge of its own data resources. It ensures that the policies and processes relating to quality are formalised, clear and accessible to all relevant staff, which supports higher quality data. The organisation can trust the data it holds, and, where data quality problems may exist, it is aware of them and able to act to remedy or mitigate them before they cause harm.

It allows scaling up.

Automating processes is one of the only cost-effective ways to scale them up, and automation requires data that is high trust and demands minimal oversight. Good data governance helps businesses ensure they can scale up dynamically and flexibly.

It enables reliable businesses intelligence.

Business intelligence suffers when it’s derived from poorly managed, siloed and incomplete data. Strong data governance keeps it in a fit state to use for confident decision-making.

It supports compliance and risk mitigation.

There are many compliance obligations a business may have when it comes to how it stores and uses data. They range from banks’ complex anti-money laundering and counter-terrorism financing (AML/CTF) obligations to the SPAM Act’s more straightforward requirement that we action unsubscribes promptly. Whether organisations need to process and analyse, store or delete data, all of these activities benefit from well-supervised, clean, orderly data, which is among the key goals of data governance.

It preserves data security.

Distinct from an organisation’s compliance activities are its broader security priorities. Data governance ensures a clear and comprehensive hierarchy of roles, responsibilities and permissions, as well as specific security policies that apply to day-to-day data management. This supports consistency and reliability in data security practice across the business.

What are the key components of data governance?

Data governance, as we’ve seen, encompasses and affects a great many activities throughout a business. It’s the wide collection of elements that ensures everyone knows not just what data is being stored, but also where, how and why that data is stored, what it’s used for, who uses it, how to maintain it, who is responsible for it, and what processes to undertake if something goes wrong.

A data governance framework differs from organisation to organisation, but tends to include elements like:

• The organisation’s data strategy
• Documentation about systems, storage and architecture
• All data-related policies and procedures, including those concerning implementation and use of AI tools
• Roles, responsibilities, and issues management hierarchy — who is responsible for what, and by what method should issues be escalated and resolved?
• Metrics that inform reports on the performance of data
• Data quality, integrity and security standards
• Documentation regarding suppliers
• Risk and compliance requirements and procedures

Many of these elements overlap with one another, but, taken altogether, they represent key elements of a useful data governance framework.

Resourcing data governance

In ideal circumstances, data governance falls within the purview of a cross-departmental council or a dedicated and well-resourced team. Most businesses do not operate in ideal circumstances.

In a business context, resourcing anything without either an immediate pay-off or a minimum compliance requirement on the line can be a lot to ask.

The challenge of strong data governance is that it offers significant long-term benefit, but the immediate advantages can be murky to decision-makers. It rarely attracts the attention of a team — and, sometimes, it does not even attract the attention of a single full-time resource, instead relegated to a fragment of someone’s workload.

This can allow data governance priorities to fall by the wayside when more urgent projects are given primacy, resulting in a build up of outdated policies and procedures presiding over data that cannot support the technologies the business wishes to deploy.

Relatedly, it’s worthwhile to consider that the “who” and “how” of resourcing data governance (to whichever degree the business can do so) can significantly impact its effects on the business over time. There is an evergreen tension between the utility of data and the security of data. Concomitantly, there is also a difference in perspectives across a business when it comes to teams focused on using that data and those focused on protecting it. A compliance team and a sales team might, for example, have quite different priorities when it comes to the governance of customer data.

When resourcing data governance, the employment backgrounds and current priorities of staff members involved will tend to create trends in its deployment. This balance between risk and value can usually be resolved with reference to the goals in the organisation’s data strategy. But it can still cause challenges, particularly if teams are siloed or data governance is a minor project that sits isolated within one team or within the remit of one person. Under those circumstances, it must be anticipated for that team’s bias to prevail and make an impact over time.

Creating stronger data governance

Many businesses evolve using ad hoc and piecemeal data solutions, and in them, data governance is an afterthought. It’s common for leaders to realise much later on that the flow of data throughout the business is confused or risky, and that it’s insufficient to support the enablement of increasingly advantageous technologies.

Fortunately, stronger data governance isn’t an all-or-nothing proposition. Some changes might be resource intensive and challenging, and others might be high-impact easy wins. Fortunately, all improvements in data governance will produce better data outcomes — which is vital at the moment, when so many businesses are re-examining their data practices with a view to AI implementation.