Terms worth understanding before — and during — an audit
Data auditing involves a specific vocabulary. This glossary explains key terms in plain language, with attention to how they apply in practice rather than in theory.
D
Data Audit
A structured examination of an organisation's data collection practices, data quality, and the relationship between collected data and actual decisions. A data audit typically produces a catalogue of existing data sources, an assessment of how each is used, and recommendations for closing gaps between collection and use. Unlike a technical data quality audit, a broader data audit also examines the organisational and behavioural dimensions of how data reaches decisions.
Context: Core term in Buzomu's workData Governance
The framework of policies, roles, and processes that determine how data is collected, stored, accessed, and used within an organisation. Governance covers questions of ownership (who is responsible for a given dataset), access (who can view or modify it), documentation (what metadata exists), and quality standards (what constitutes acceptable data). Weak governance is a common contributor to the collection-use gap, because data collected without clear ownership often lacks the documentation needed for others to find and use it.
Context: Governance review is part of every full auditDormant Data
Data that is actively collected and stored but has no current downstream use — no reports draw from it, no decisions consult it, and no processes depend on it. Dormant data is distinct from archived data, which was once used and is preserved for historical reference. Dormant data represents collection cost without corresponding value. Identifying dormant data is one of the consistent findings of a data audit, and the discovery often prompts either a decision to start using the data or a decision to stop collecting it.
Context: Identified in the inventory phase of an auditDecision Quality
The degree to which a decision is supported by relevant, accurate, and timely information. Decision quality is distinct from decision outcome — a well-supported decision can still produce a poor outcome due to factors outside the decision-maker's control, and a poorly supported decision can produce a good outcome by chance. Improving decision quality means improving the information available to decision-makers, not guaranteeing results. Data audits focus on identifying where decision quality could be improved by making better use of existing data.
Context: Central concept in audit recommendationsE
Event Tracking
The practice of recording specific user actions or system events as discrete data points. In web analytics, event tracking captures interactions such as button clicks, form submissions, video plays, and scroll depth. In operational systems, event tracking records process milestones, errors, and state changes. Event tracking data is frequently underused because the events are defined and captured by technical teams but the resulting data is not systematically connected to the business questions that would benefit from it.
Context: Common area of dormant data discoveryG
Gap Analysis
In the context of data auditing, gap analysis refers to the process of identifying the distance between the data that exists and the data that decisions actually use. A gap can take several forms: data exists but isn't reaching decision-makers; data exists but isn't trusted; data is consulted but is of insufficient quality to support the decision reliably; or a decision is being made without data that would be feasible to collect. Gap analysis is the analytical core of a data audit.
Context: Central methodology in audit analysis phaseI
Integration Gap
A break or discontinuity in the flow of data between systems, where information that exists in one system does not reliably reach another. Integration gaps are common in organisations that have accumulated multiple platforms over time without a coherent data architecture. They result in data being collected in one system and decisions being made in another, with no reliable mechanism connecting them. Integration gaps are often discovered during the inventory phase of an audit and are a frequent source of preventable information loss.
Context: Identified in integration gap analysisInformation Hierarchy
The implicit or explicit prioritisation of certain data sources over others within an organisation. Every organisation has an information hierarchy, though it is rarely documented. Some sources are trusted and regularly consulted; others are available but routinely ignored. Understanding the information hierarchy helps explain why certain data remains dormant even when it would be relevant to decisions — it may be lower in the hierarchy than the sources decision-makers habitually consult.
Context: Explored during discovery interviewsM
Metadata
Data that describes other data. In the context of a data audit, metadata includes documentation of what a dataset contains, how it was collected, when it was last updated, who owns it, and what it has been used for. Metadata is essential for making data findable and usable by people who were not involved in its collection. Poor metadata is one of the most common reasons that data remains dormant — it exists but cannot be found or understood by those who might benefit from it.
Context: Assessed during data inventoryO
Operational Data
Data generated as a byproduct of operational processes rather than collected for analytical purposes. Operational data includes transaction records, system logs, process timestamps, error records, and throughput metrics. This type of data is often abundant but underused analytically, because it is generated and stored by systems designed for operational purposes rather than for reporting. Surfacing operational data for decision support is a common opportunity identified in data audits.
Context: Common source of untapped dataP
Pipeline
In data contexts, a pipeline is the sequence of steps through which data travels from its point of collection to its point of use. A pipeline might include collection, cleaning, transformation, storage, and delivery to a reporting tool or decision-maker. Pipelines can fail or degrade at any point, and a data audit often examines pipelines to identify where data is being lost, distorted, or delayed in transit. A broken pipeline can mean that data is being collected but never reaches the people or systems that would use it.
Context: Examined during integration gap analysisR
Reporting Habit
The recurring pattern by which an organisation consults data — which reports are reviewed regularly, by whom, and in what context. Reporting habits are often established early in an organisation's history and change slowly, even when the available data has expanded significantly. A data audit frequently finds that reporting habits have not kept pace with data collection, meaning that new data sources are being ignored not because they're irrelevant but because they're outside the established reporting routine.
Context: Explored during discovery interviewsS
Signal-to-Noise Ratio
In data analysis, the ratio of meaningful information (signal) to irrelevant or misleading variation (noise). A low signal-to-noise ratio means that useful information is difficult to extract from a dataset because it is obscured by irrelevant variation. Data audits consider signal-to-noise in the context of reporting: dashboards that display many metrics often have a low signal-to-noise ratio for any given decision, making it harder for decision-makers to identify what's relevant. Simplifying and focusing reporting is sometimes a more effective improvement than adding new data.
Context: Relevant to reporting and dashboard reviewSource of Truth
A designated data source that is treated as the authoritative reference for a particular type of information within an organisation. When multiple systems contain overlapping data, conflicts arise unless one is designated as the source of truth. Organisations without clear sources of truth often experience decision-making friction, where different teams cite different numbers for the same metric because they are drawing from different systems. Identifying and clarifying sources of truth is a common recommendation in data audits.
Context: Governance review findingQuestions about how these concepts apply to your organisation?
We're happy to discuss your specific context and what an audit might surface.