Create a data glossary

A data glossary is a simple list of your organisation’s key data terms and what they mean.

It's like a small dictionary. It helps everyone stay aligned across all areas of the business. 

AI tools learn patterns from data. If people use the same term in different ways, your data can quickly become confusing and inconsistent.

Common terms – like customer, order, supplier and transaction – can mean different things across your organisation. Without clear definitions, this can lead to:

• reports that don’t match
• key measures that are hard to compare
• decisions taking longer
• less trust in your data.

A data glossary helps people use key terms in the same way so AI tools and systems can use your data more consistently.

A data glossary also improves:

• data literacy – people understand the data they use
• data governance – terms are agreed, owned and kept up to date
• trust in insights – reports match and people make decisions using the same numbers
• AI readiness – AI tools work better with consistent, high-quality data. 

You don’t need a large project to get started. Start small and build it over time.

Step 1: Pick the most important terms

Choose key terms that your organisation uses often. Start with terms linked to money, customers, staff, stock or delivery. For example:

• customer
• order
• invoice
• refund
• supplier
• product
• staff member.

Step 2: Write a simple definition for each term

Keep each definition short and avoid technical language.

If a term is hard to explain, that may be a sign your organisation isn’t using it consistently yet.

Step 3: Add rules and examples

Add a brief note about what the term includes and doesn’t include. Examples help people apply the definition in real work.

Step 4: Assign an owner

Pick one person or team to keep definitions up to date. This doesn’t need to be a data expert – just someone who understands the business meaning.

Step 5: Put it somewhere easy to find

Use a shared document, spreadsheet or wiki page. Make sure people across your organisation can find the glossary and suggest updates. 

A basic data glossary can include the following information:

• Term – the official name used across the organisation.
• Definition – a simple, clear definition for non‑technical users.
• Location – where the term appears across the organisation.
• Alternative terms – similar words for the term that people can refer to.
• Scope – what the term does and doesn’t include.
• Examples – what the term means in practice.
• Related terms – connected terms in the glossary.
• Owner – who is accountable for the definition and keeping it up to date.
• Priority – high, medium or low based on where to focus your data quality effort first.

Keep it up to date

A glossary only works if people use it. These small habits help:

• Review your top terms every 3 to 6 months.
• Update definitions when systems or processes change.
• Add new terms when you add new products, services or software.
• Ask staff to flag terms that cause confusion.