Business Glossaries: Agreeing What "Customer" Actually Means

Opening Scene

Two branches of the same family, reunited after decades apart, discover they’ve been using the surname “Reyes-Martin” to mean two different combined households, and untangling family history turns out to require agreeing, once and for all, exactly who counts as a “Reyes-Martin” going forward, in writing, so the next reunion doesn’t repeat the confusion.

In Plain English

A business glossary is a governed set of agreed-upon definitions for the business terms an organization uses — “customer,” “active user,” “revenue,” “churn” — independent of any specific table or system. It differs from a data dictionary in altitude: a dictionary defines what a specific field in a specific table means technically; a glossary defines what a business concept means at the organizational level, and then links that concept to every technical field across the catalog that implements it.

The Old Way

Before business glossaries were treated as governed, cross-team artifacts:

Marketing counted a “customer” as anyone who signed up, while Finance counted only paying accounts, and both teams presented conflicting numbers in the same meeting.
Term definitions lived informally in individual teams’ heads or local documents, invisible and unreconciled across the organization.
Disputes over whose number was “right” consumed meeting after meeting, because there was no governed, agreed source of truth to point to.

A business glossary exists specifically to force that reconciliation once, in writing, instead of relitigating it every quarter.

What’s Changing (and Why AI Is the Reason)

Glossaries are moving from static documents maintained by a single governance team into living, catalog-integrated artifacts that link directly to the technical fields implementing each term.
This linkage builds directly on the field-level work covered earlier in this series’ data dictionary article, connecting business-level meaning to technical-level implementation.
AI systems generating business-facing answers from data need glossary-grounded definitions to avoid quietly picking whichever technical field happens to be named closest to the business term asked about, which is a common and hard-to-detect source of AI-generated reporting errors.

The Metaphor, Fully Extended

Agreeing on a Family Name	Business Glossary Concept
Two branches using “Reyes-Martin” to mean different households	Two teams using “customer” to mean different populations
Reconciling the name’s meaning once, in writing, for the whole family	Reconciling a term’s meaning once, in writing, for the whole org
A glossary entry every branch can point back to	A glossary entry every team can point back to
Avoiding the same confusion at the next family reunion	Avoiding the same dispute in the next reporting cycle

For Beginners: What to Actually Do

When two reports disagree on a metric, check whether they’re actually using different definitions of the same business term before assuming one is simply wrong.
Look up business terms in the glossary, not just field names in the dictionary, when trying to understand a report.
If you notice a business term being used inconsistently across teams, raise it as a glossary gap rather than quietly picking a side.

For Practitioners and Leaders: The Deeper Layer

Establish clear stewardship for glossary terms, since a term without an accountable owner will inevitably drift back into inconsistent usage.
Link every glossary term explicitly to the technical fields that implement it across the catalog, so the connection between meaning and implementation stays visible.
Treat glossary disputes as a signal worth escalating to governance leadership, since unresolved term conflicts tend to produce recurring, expensive reporting disagreements.

Quick Recap

A business glossary defines organization-level business terms, distinct from the technical field-level definitions in a data dictionary.
Without a governed glossary, teams routinely define terms like “customer” or “revenue” inconsistently and argue over conflicting numbers.
Modern glossaries link directly to the technical fields implementing each business term.
AI systems need glossary-grounded definitions to avoid silently guessing which technical field a business term refers to.

Where This Fits in the Series

Article 9 covered choosing the platform to house an organization’s catalog. This article covers agreeing on the shared vocabulary that platform needs to enforce consistently. Article 11 shifts from defining known terms to discovering data nobody knew existed — finding distant relatives no one in the family had heard of yet.