---
title: "What Is Data Quality? Definition and Dimensions"
description: "What is data quality? A plain definition, the dimensions (accuracy, completeness, timeliness), how it differs from data governance and a quick self-check."
url: https://nexumautomations.com/glossary/data-quality
updated: 2026-11-10
publisher: Nexum Automations, an AI consulting firm
---

> Nexum Automations is an AI consulting firm that builds what it recommends. It finds where AI pays off in a company, builds it on the tools the team already uses and stays until it is live and working.

Glossary

# What is data quality? Explained

A plain definition of data quality, the dimensions used to measure it, how it differs from close terms and why it matters before you use AI.

Last reviewed November 10, 2026

Definition

Data quality

Data quality is how well a set of data fits the use it is meant for. Data that is good for sending invoices may be poor for forecasting. What is data quality in practice? It is measured on dimensions such as accuracy, completeness, consistency, timeliness, validity and uniqueness.

**In short:** How well data fits its intended use.

Our data

## What buyers see in this search

The first page of Google for "what is data quality", counted by type of result.

| | Results |
| ---------------------------------------------------- | ------------------------------- |
| Articles by companies that sell a product or service | 9 of 9 |
| Titles that answer this exact search | 9 of 9 |
| Google features on the page | An AI Overview, People also ask |

The 9 organic results on the first page of Google, United States, desktop, measured with DataForSEO on October 29, 2026\. Types only: no company or product is named.

Side by side

## Dimensions at a glance

Core six, with a seventh from Precisely and fitness for purpose from IBM.

| | Plain meaning | Customer or invoice example |
| ------------ | -------------------------------- | ---------------------------------------------- |
| Accuracy | Data matches reality | The phone number is the customer's real number |
| Completeness | No required value is missing | Every invoice has a due date |
| Consistency | Same fact, same value everywhere | Billing and CRM show the same address |
| Timeliness | Data is up to date | A moved customer's address is updated |
| Validity | Values follow the allowed format | A date field holds a real date |
| Uniqueness | No duplicate records | One customer is not entered twice |

## What is data quality, in plain words

1. Pick the use
Decide what the data must support.
2. Check the data
Test it on each dimension.
3. Compare
Is it good enough for that use?
4. Fix or accept
Repair the gaps that matter.

Data quality is a question of fit. The same customer list can be fine for a mailing and useless for a sales forecast.

IBM defines it as how well a dataset meets criteria such as accuracy, completeness, validity, consistency, uniqueness, timeliness and fitness for purpose.

Quality is never one score for a whole company. You judge each dataset against the job it has to do.

What is data quality in machine learning? The same idea applies: a model learns from its training data, so gaps, duplicates and stale records carry straight into its answers.

## The dimensions of data quality

* Accuracy
Values match the real world.
* Completeness
Required fields are filled in.
* Consistency
The same fact matches across systems.
* Timeliness
Data is current when you need it.
* Validity
Values follow the expected format.
* Uniqueness
Each thing appears only once.

The dimensions are the lenses used to measure quality. Sources do not agree on one list: Collibra describes six, Precisely lists seven, and IBM adds fitness for purpose on top.

This also answers the common questions about the six pillars and the seven components of data quality. They are the same idea counted differently.

The table below uses the six core dimensions, plus the one Precisely adds for seven.

## How data quality differs from close terms

Data quality

* Is the data fit for its use?
* Measured on dimensions
* The result you want

Close terms

* Governance: rules and owners
* Master data: one record per item
* Cleansing: the fixing itself
* Integrity: no unwanted change

Several terms sit near data quality, and they get mixed up. Each answers a different question.

[Data governance](https://nexumautomations.com/glossary/data-governance) is the set of rules and owners. It decides who is responsible for quality, so quality is one thing governance aims for.

[Master data management](https://nexumautomations.com/glossary/master-data-management) keeps one agreed record per customer or product. Data cleansing is the fixing: removing duplicates and correcting errors.

Data integrity means data has not been altered by accident or by an attacker. Quality asks whether the data is good enough to use. [Data lineage](https://nexumautomations.com/glossary/data-lineage) shows where the data came from.

## Why it matters before AI

Signs of good data

* One record per customer
* Required fields filled
* Records reviewed recently

Signs of weak data

* Duplicate customers or invoices
* Blank required fields
* Stale or outdated records

AI works on whatever data you give it. Weak data produces weak answers, only faster.

In a 2025 survey by the AI, Data & Analytics Network (PEX Report 2025/26, 200+ professionals), 52% said data quality and availability held back AI adoption more than anything else.

A quick self-check on your own data starts with three things. Our [data engineering](https://nexumautomations.com/services/data-engineering) work covers how pipelines keep data clean.

* Duplicates: the same customer or invoice twice
* Blanks: required fields left empty
* Stale records: old addresses, closed accounts

FAQ

## Data quality questions

* ### What are the six pillars of data quality?
Collibra lists six: accuracy, completeness, consistency, timeliness, validity and uniqueness. Other sources add a seventh, so the count depends on who you ask.
* ### What are the 7 components of data quality?
Precisely lists seven dimensions, and IBM adds fitness for purpose to its list. The core six are accuracy, completeness, consistency, timeliness, validity and uniqueness.
* ### What is another term for data quality?
Data health or data fitness are used loosely. Do not confuse it with data integrity, which only covers unwanted change.
* ### What are the principles of data quality management?
Managing quality means setting owners and rules, measuring dimensions, and fixing errors where the data is created. [Data governance](https://nexumautomations.com/glossary/data-governance) is where those rules live.
* ### What is data quality in machine learning?
Models learn from their training data, so duplicates, blanks and stale records carry into the results. Checking the dimensions before training works the same way as for reports.

---

Source: https://nexumautomations.com/glossary/data-quality

Free AI diagnosis (30-minute call): https://nexumautomations.com/#book - Services: https://nexumautomations.com/services - Guides: https://nexumautomations.com/guides
