Data Quality and the Finance Function

Data quality is the foundation on which everything the finance function does is built, and yet it is one of the most underappreciated determinants of how well the function performs. Every number the finance function produces, every report, every analysis, every decision it informs, rests on the quality of the underlying data — and where that data is poor, everything built on it is compromised, however sophisticated the systems and however skilled the people. Conversely, a finance function built on good data can produce reliable information efficiently and can trust the numbers it works with. Data quality is therefore not a peripheral technical matter but a fundamental determinant of the finance function’s effectiveness, and a finance leader who understands and attends to it builds a stronger function than one who takes the data for granted.

This guide is written for finance leaders and professionals who want to understand and improve the data quality their function depends on. It covers why data quality matters so much to finance, what good and poor data quality look like, the causes of poor data quality, how to improve and maintain data quality, and the role of data quality in the modern, increasingly automated and AI-enabled finance function. It is a practical orientation to a foundational issue that affects everything the finance function does. The aim is an understanding of why data quality matters, where it comes from, and how to build and maintain the good data quality that an effective finance function depends on.

Why Data Quality Matters So Much to Finance

Data quality matters to finance because finance is fundamentally about producing reliable information from data, and the reliability of that information cannot exceed the quality of the data it is built on. Poor data quality undermines everything: the reporting is unreliable because the data behind it is wrong; the analysis is flawed because it works from poor data; the decisions informed by the numbers are misguided because the numbers are wrong; and the effort consumed correcting and working around poor data is effort not spent on value-adding work. A finance function struggling with poor data quality is fighting a losing battle, because the foundation it builds on is unsound, and no amount of skill or sophistication above that foundation can fully compensate for it.

The cost of poor data quality is often hidden but real. It shows up in the time the finance function spends correcting errors, reconciling discrepancies, and working around the unreliable data — time that could go to analysis and partnering. It shows up in the errors that flow through to the reporting and the decisions, with their consequences. It shows up in the lack of trust in the numbers, which undermines the finance function’s credibility and the business’s confidence in its information. And it shows up as a fundamental constraint on the function’s ability to do more sophisticated work, including the automation and AI that depend critically on good data. Understanding that data quality is foundational — that everything finance does rests on it and that poor quality undermines everything — is the basis for giving it the attention it deserves, which many finance functions do not until the problems become acute.

What Good and Poor Data Quality Look Like

Understanding data quality requires understanding what good and poor quality look like in practice. Good data quality means data that is accurate — correctly representing what it is supposed to; complete — with the necessary data present rather than missing; consistent — the same thing represented the same way across the data rather than in conflicting forms; timely — available when needed and up to date; and valid — conforming to the rules and formats it should. Data with these qualities can be relied upon, supports accurate reporting and analysis, and provides a sound foundation for the finance function’s work. These dimensions — accuracy, completeness, consistency, timeliness, validity — together define good data quality.

Poor data quality is the absence of these qualities: data that is inaccurate, incomplete, inconsistent, out of date, or invalid, in any combination. It shows up as errors in the data, as duplicate or conflicting records, as missing information, as inconsistent representations of the same thing, as data that does not conform to the rules it should. These problems undermine the reliability of anything built on the data, and they accumulate where data quality is not actively maintained. The finance leader who understands what good and poor data quality look like can assess the quality of the function’s data — recognising the problems where they exist — and can define what good quality requires. This understanding is the basis for both diagnosing data quality problems and working toward good quality, because you cannot improve what you cannot define and recognise. Knowing what good data quality looks like is the foundation for achieving it.

The Causes of Poor Data Quality

Poor data quality has identifiable causes, and understanding them helps the finance leader address it at the source rather than perpetually correcting the symptoms. A common cause is poor data entry — errors, inconsistencies and omissions introduced when data is entered, whether through human error, inadequate validation, or unclear conventions — which puts poor data into the system at the source. Another is the lack of controls and validation — systems and processes that do not check the data as it enters and flows through, allowing errors to enter and propagate undetected. A third is the proliferation of data across multiple systems and sources that are not properly reconciled or integrated, leading to inconsistency and conflict between different versions of the data.

Further causes include the accumulation of errors over time where data quality is not maintained, the carrying forward of poor data through migrations and system changes, the lack of clear ownership and accountability for data quality, and the absence of the processes and discipline that would keep the data clean. These causes share a common theme: poor data quality results from the lack of the controls, processes, discipline and ownership that would prevent it and maintain good quality. This is actually encouraging, because it means data quality is largely within the finance function’s control — the causes are addressable through better controls, processes, discipline and ownership. The finance leader who understands the causes of poor data quality can address them at the source, building the controls and discipline that prevent poor data rather than perpetually correcting it. Addressing the causes, rather than the symptoms, is the path to genuinely improving data quality.

Improving and Maintaining Data Quality

Improving and maintaining data quality is a matter of building the controls, processes, discipline and ownership that good data quality requires. The controls and validation that check data as it enters and flows through the systems — catching errors at the source, enforcing the rules and formats, preventing the poor data from entering — are foundational, because preventing poor data is far more efficient than correcting it. The processes that maintain data quality — the reconciliations that catch discrepancies, the routines that keep the data clean, the discipline of correcting errors promptly — keep the quality from degrading over time. And the ownership and accountability for data quality — clear responsibility for the data being right — ensures that data quality is actively maintained rather than neglected.

Improving data quality where it is poor often requires a deliberate effort — cleaning the existing data, resolving the errors, duplicates and inconsistencies that have accumulated — followed by the ongoing discipline to maintain the improved quality. The deliberate clean-up addresses the accumulated problems; the ongoing discipline prevents them recurring. The finance leader who builds the controls, processes, discipline and ownership, and who undertakes the deliberate improvement where needed, builds and maintains the good data quality the function depends on; one who neglects these allows the data quality to degrade and the function to struggle on a poor foundation. Improving and maintaining data quality is largely a matter of the controls and discipline that prevent and correct the problems, sustained over time, and it is one of the more valuable, if unglamorous, investments a finance leader can make in the function’s effectiveness, because it strengthens the foundation on which everything else is built.

Data Quality in the Modern Finance Function

Data quality has become even more critical in the modern finance function, because the automation and AI that increasingly characterise finance depend critically on good data. Automation applies consistent processing to the data, which means it processes good data efficiently but also propagates poor data efficiently — an automated process built on poor data produces poor outputs at scale and speed. AI is even more dependent on data quality, because AI systems learn from and operate on data, and poor data produces poor AI outputs, sometimes in ways that are hard to detect. The drive toward automation and AI in finance therefore raises the stakes on data quality, because these technologies amplify both the value of good data and the damage of poor data.

This means that data quality is not just a foundation for the traditional finance function but a prerequisite for the modern, automated, AI-enabled one. A finance function seeking to automate its processes or apply AI must first have the data quality these technologies require, because building automation and AI on poor data produces unreliable results at scale. The finance leader who recognises this — that good data quality is the prerequisite for the automation and AI that the modern finance function increasingly relies on — gives data quality the priority it deserves, because it is the foundation not just for today’s work but for the function’s ability to modernise. Data quality, long underappreciated, is becoming recognised as a critical enabler of the modern finance function, and the finance leader who builds and maintains it is laying the foundation for both reliable information today and the automation and AI of tomorrow. This connects to the broader use of AI in finance covered in our hub on AI in finance, where the dependence of AI on good data is a recurring theme. Attending to data quality is therefore one of the more important, if less visible, things a finance leader can do to build an effective and future-ready finance function.

Building a Culture That Values Data Quality

Sustaining good data quality over time depends not only on controls and processes but on a culture that values it, and the finance leader should work to build that culture. Data quality is maintained by the people who enter, handle and use the data, and where those people understand its importance and take responsibility for getting the data right, the quality is maintained; where they treat data quality as someone else’s concern or as unimportant, it degrades despite the controls. Building a culture that values data quality — in which the people understand why it matters and take ownership of the data they handle — is what sustains good quality beyond what controls alone can achieve.

Building this culture means helping the people understand the importance of data quality and the consequences of poor data, establishing clear ownership and accountability so that responsibility for the data is genuine, and reinforcing the discipline that keeps the data clean. It also means the finance leader treating data quality as a priority rather than an afterthought, because the function takes its cue from its leadership, and a leader who visibly values data quality fosters a function that does too. The finance leader who builds a culture that values data quality — understanding, ownership, discipline, prioritisation — sustains the good quality that controls and processes establish; one who relies on controls alone, without the culture, finds the quality harder to maintain. The cultural dimension — the people understanding and owning data quality — is what makes good data quality durable, and building it is part of the finance leader’s role in establishing the foundation on which the function depends. Good data quality is ultimately sustained by people who care about it, and fostering that care is one of the more valuable things a finance leader can do for the function’s long-term effectiveness.