The most common trigger for a data audit is a specific incident: a report that produced a number nobody believed, a model that gave contradictory outputs on different days, a board question that nobody could answer with confidence. These incidents are useful. They create the organisational will to look carefully at something that is easy to defer. But they are rarely the root cause. They are symptoms.
What a Data Audit Actually Examines ¶
A data audit at Quantum Lattice covers three things: the provenance of your data (where it comes from and how it is collected), the consistency of your data (whether the same concept is defined the same way across systems and teams), and the fitness of your data for the decisions you are trying to make. The third category is the one most organisations have not thought about explicitly.
The Interview Phase ¶
We spend a significant portion of every data audit interviewing the people who maintain and use the data, not just the people who commissioned the audit. The person who enters data into a system every day knows things about its limitations that are not written down anywhere. These conversations are often the most valuable part of the engagement. We have found inconsistencies in this phase that no amount of technical review would have surfaced.
Common Findings ¶
The most common finding is definitional inconsistency: the same field means different things in different parts of the organisation, and nobody has noticed because the reports that use it are read by different people. The second most common finding is a gap between the data that exists and the data that would be needed to answer the organisation's most important questions. Both are fixable, but fixing them requires knowing they exist.
The Roadmap Output ¶
The audit concludes with a prioritised roadmap: a written document that identifies the issues found, ranks them by their impact on decision quality, and recommends specific remediation steps. We distinguish between quick wins (things that can be fixed in a week with no new infrastructure) and structural changes (things that require investment and planning). Most organisations find that the quick wins alone justify the cost of the audit.
A data audit is a good starting point if you suspect your data is not serving your decisions as well as it should, but you are not sure where the problem lies. We can scope one for your situation in a short initial conversation.