Across industries, data governance is still treated as a heavy, compliance-driven exercise, slow, bureaucratic, and disconnected from real business value. Yet in my experience working with financial institutions, regulators, and mining organisations, the real problem is not a lack of frameworks. It is the persistence of flawed assumptions about how governance should work in practice.

To make this real, let’s move away from theory and into everyday operational contexts.

In a commercial bank, data governance often becomes visible when something goes wrong, such as incorrect customer balances, inconsistent loan exposure reports, or delayed regulatory submissions. The typical response is to form a committee, introduce more controls, or invest in a data governance tool. But the root cause is usually much simpler: a lack of clear data ownership.

Take customer data as an example, if no one in Retail Banking “owns” the definition of an active customer, Finance may define it one way for reporting, Risk another way for exposure calculations, and Digital another way for app usage metrics. The result is constant reconciliation and mistrust of data. Good governance in this context is not about a new tool; it is about assigning a Data Owner in the business, agreeing on a standard definition, and ensuring that every system aligns to it. Suddenly, reporting becomes faster, disputes reduce, and decision-making improves.

Now consider a Central Bank or regulator overseeing multiple financial institutions. Here, the instinct is often to enforce governance through strict compliance requirements, such as templates, reporting standards, and penalties for noncompliance. While necessary, this approach alone does not guarantee quality data.

For example, a regulator may require banks to submit capital adequacy reports. If each bank interprets data definitions slightly differently, the regulator receives data that is technically compliant but analytically inconsistent. This creates risk at a systemic level. A more effective governance approach is for the regulator to define critical data elements clearly, what constitutes Tier 1 capital, and how exposures are classified, and to enforce standardisation at the definition level, not just the reporting format. In this case, governance improves not just compliance, but the regulator’s ability to make sound policy decisions.

In a mining company, the challenges look different but follow the same pattern. Data is generated across geology, operations, maintenance, and finance, often in silos. A common issue is the disconnect between production data and financial performance.

Imagine a situation where the operations team reports high production volumes, but finance reports lower-than-expected revenue. The immediate assumption is a performance issue, but often the real problem is data inconsistency. Ore grades, recovery rates, and stockpile measurements may be defined differently across systems. Without clear ownership and standard definitions, the organisation spends more time reconciling data than improving performance.

Effective governance in this context means assigning ownership of key data elements such as ore grade and recovery rates, standardising definitions across systems, and embedding these standards into daily operational processes. When done correctly, production and financial data align, enabling more accurate forecasting, better decision-making, and improved investor confidence.

One of the most common misconceptions across all these environments is that governance must be centralised. In practice, centralising execution often creates bottlenecks. In a bank, if every data issue must go through a central Data Office, business teams quickly become frustrated. In a mining operation, central control can slow down time-sensitive decisions on site.

A more effective model is to centralise principles but decentralise execution. Define standards centrally, but allow business units, Retail Banking, Risk, Operations, and Geology to own and manage their data within those standards. This keeps governance close to where data is created and used, making it far more practical and sustainable.

Another area where organisations struggle is governing everything at once. A bank may attempt to catalogue all its data assets. A regulator may try to standardise every reporting element. A mining company may aim to clean all historical data. These efforts often stall under their own weight.

A more pragmatic approach is to focus on what matters most. In a bank, start with critical data used for regulatory reporting and financial statements. In a regulator, prioritise data elements that drive systemic risk assessment. In mining, focus on production and cost drivers. By targeting high-value, high-risk data first, organisations can deliver quick wins and build momentum.

There is also a persistent belief that governance slows things down. In reality, poorly designed governance slows things down. Well-designed governance accelerates execution.

Consider a credit approval process in a bank. If data definitions are unclear and customer information is inconsistent, approvals are delayed due to repeated checks and corrections. With clear governance, standard definitions, validated data, and clear ownership, the same process becomes faster and more reliable.

In mining, maintenance decisions rely on equipment data. If this data is inconsistent or unreliable, decisions are delayed or incorrect, leading to downtime. With strong governance, data becomes trusted, enabling faster and more effective decisions on the ground.

Ultimately, the biggest shift organisations need to make is recognising that data governance is not an IT responsibility. In every example above, the value of data is defined by the business. IT enables, but it does not own the meaning of data. When governance is treated as an IT initiative, it remains disconnected from real operational needs.

Good governance lives in the business. It shows up in how a banker defines a customer, how a regulator interprets risk, and how a mining engineer measures production. It is embedded in daily decisions, not confined to policies or committees.

For organisations undergoing digital transformation, this distinction is critical. You cannot digitise or automate effectively if your data is inconsistent, poorly defined, or mistrusted. Governance is not a compliance exercise; it is a performance enabler.

The organisations that succeed are those that simplify their approach. They start with ownership, focus on what matters, embed governance into workflows, and evolve continuously. They move away from theoretical frameworks and towards practical, business-driven execution.

In the end, data governance is not about control. It is about clarity, accountability, and enabling better decisions, whether in a bank approving loans, a regulator safeguarding the financial system, or a mining company optimising production.

That is where real value lies.