Translation Glossary Guide: Build and Manage Term Bases [2026]

A translation glossary (also called a termbase or terminology database) is a structured list of terms with their approved translations, definitions, and usage guidelines. Unlike dictionaries, glossaries focus on specialized terminology specific to your organization, industry, or product.

Here's a number that should get your attention: 70% of translation quality issues stem from terminology problems. This guide covers how to create, manage, and use glossaries to keep your translations consistent and accurate.

What is a Translation Glossary?

A translation glossary is a curated database of terms with their approved translations in one or more target languages. It's the single source of truth for how specific words and phrases should be translated.

That sounds simple — and it is. The hard part isn't understanding what a glossary does. It's building one that people actually use.

Glossary vs. Dictionary

Components of a Glossary Entry

A well-structured glossary entry includes:

• Source term — The term in the original language
• Target term — The approved translation
• Definition — What the term means in context
• Part of speech — Noun, verb, adjective, etc.
• Context/Example — How the term is used in a sentence
• Usage notes — Guidelines, restrictions, or preferences
• Domain — Subject area (legal, medical, technical, etc.)
• Status — Approved, pending, deprecated
• Do Not Translate flag — For brand names, product names

Standard Formats

The primary glossary exchange format is TBX (TermBase eXchange), an ISO standard (ISO 30042). Other common formats include:

• Excel/CSV — Simple, widely compatible
• XLIFF — XML-based, CAT tool compatible
• TBX — Industry standard for terminology exchange
• Proprietary formats — Tool-specific databases

Why You Need a Translation Glossary

The 70% Rule

Studies consistently show that about 70% of translation quality issues relate to terminology — inconsistent term translation, wrong technical terms, brand name variations, and unapproved terminology choices.

A glossary directly addresses every one of these.

Business Impact

Compliance Requirements

In regulated industries, terminology consistency isn't optional — it's mandated:

• Medical/Pharmaceutical — Drug names, dosages, warnings must be exact
• Legal — Contract terms have specific legal meanings
• Financial — Regulatory terms require precise translation
• Technical — Safety-critical terminology must be consistent

Get a drug name wrong and you don't just fail a quality check. You create a safety risk.

Types of Terms in a Glossary

Brand Terms

Terms representing your company identity:

Technical Terms

Industry-specific or product-specific vocabulary:

Do Not Translate (DNT) Terms

Terms that should stay in the source language: product names and trademarks, industry-standard abbreviations, technical identifiers (URLs, code), and proper nouns by policy.

Preferred vs. Forbidden Terms

This table is one of the most useful things you can put in a glossary. Translators appreciate knowing what not to write.

Industry Examples

Software/SaaS: UI elements (button, dropdown, modal), actions (submit, cancel, save), status messages (error, warning, success).

Medical: Anatomical terms, drug names and formulations, procedures and diagnoses.

Legal: Contract terminology, regulatory terms, jurisdiction-specific language.

E-commerce: Product attributes, checkout flow terms, shipping and payment terminology.

How to Create a Glossary: Step-by-Step

Step 1: Term Extraction

Identify terms that need standardization.

Manual extraction works well for smaller projects: review existing translations for inconsistencies, analyze customer support tickets for terminology confusion, consult subject matter experts, and review competitor terminology.

Automated extraction scales better: use CAT tool term extraction features, apply NLP tools for technical term identification, analyze frequency patterns in source content, or use AI-powered terminology extraction.

Start with manual extraction for your first 50 terms. You'll learn a lot about what matters to your organization.

Step 2: Term Definition and Translation

For each extracted term:

1. Write a clear definition — What does this term mean in your context?
2. Determine the approved translation — Research, consult experts, verify
3. Add usage examples — Show the term in context
4. Note any restrictions — When NOT to use this translation
5. Set the status — Approved, pending review, deprecated

Step 3: Expert Validation

Involve the right people:

• Subject Matter Experts — Technical accuracy
• In-country reviewers — Cultural appropriateness
• Legal/Compliance — Regulatory requirements
• Marketing — Brand voice alignment
• Product teams — UI consistency

Yes, that's a lot of stakeholders. But skipping validation creates glossaries that nobody trusts — and glossaries nobody trusts don't get used.

Step 4: Structure and Organization

Design your glossary for usability:

Company Glossary ├── Brand Terms │ ├── Product Names (DNT) │ ├── Slogans (translate) │ └── Company Values ├── Technical Terms │ ├── Software Features │ ├── API Terminology │ └── Industry Standards ├── UI Terms │ ├── Buttons and Actions │ ├── Navigation Elements │ └── Status Messages └── Domain-Specific ├── Legal ├── Medical └── Financial 

Step 5: CAT Tool Integration

Import your glossary into translation tools.

Integration options: direct import (TBX, CSV, Excel), API connection for real-time sync, cloud-based terminology management, and project-specific term extraction.

Key features to enable: term recognition highlighting, automatic QA checks, translation suggestions, and consistency warnings. These features make glossary compliance automatic rather than manual — and that makes all the difference.

Step 6: Team Training

Make sure translators know how to use the glossary: provide access documentation, explain term lookup procedures, clarify when deviation is acceptable (and the approval process for it), set expectations for adherence, and create feedback channels for term suggestions.

That last point matters. Translators working with the content daily often spot terminology gaps before anyone else does.

Terminology Management Best Practices (2026)

The CCU Principle

Every glossary term should be:

• Correct — Accurate translation for the domain
• Consistent — Same term used throughout
• Unambiguous — Clear meaning, no confusion

If a term fails any of these three tests, it needs work.

Governance Structure

Term Approval Workflow

┌─────────────────┐ │ Term Suggestion │ └────────┬────────┘ │ ▼ ┌─────────────────┐ │ Initial Review │ │ (Terminology Mgr)│ └────────┬────────┘ │ ▼ ┌─────────────────┐ │ Expert Review │ │ (SME + In-country)│ └────────┬────────┘ │ ▼ ┌─────────────────┐ │ Approval │ │ (Sign-off) │ └────────┬────────┘ │ ▼ ┌─────────────────┐ │ Glossary Update │ │ (All systems) │ └─────────────────┘ 

Regular Audit Cadence

AI-Powered Terminology (2026)

Modern tools offer automatic term extraction from source documents, translation suggestions based on context, consistency checking across projects, similar term detection to prevent duplicates, and usage analytics to identify underused terms.

The extraction part is genuinely useful — AI can scan 10,000 pages and surface candidate terms in minutes. But human review is still required for approval. AI finds the needles; humans decide which ones to keep.

Terminology Management Tools

Dedicated Terminology Tools

CAT Tools with Built-in Terminology

Most CAT tools include terminology features: SDL Trados has MultiTerm integration, memoQ has a built-in term base, Phrase has a terminology management module, Smartcat offers cloud glossary, and MateCat supports glossaries too.

KTTC Quick Glossary

KTTC offers Quick Glossary extraction for quality assessment — it extracts key terms from source/target pairs, identifies terminology inconsistencies, validates translations against existing glossaries, and integrates with Translation Memory workflows.

Glossary Integration in Translation Workflow

Pre-Translation Setup

Before translation begins:

1. Load relevant glossaries into the CAT tool
2. Brief translators on terminology requirements
3. Highlight critical terms that must match exactly
4. Provide context for ambiguous terms

During Translation

The glossary actively supports translators through term highlighting (recognized terms are marked in the source), auto-suggestions (approved translations appear), warnings (alerts fire when glossary terms aren't used), and one-click insert (apply the approved term instantly).

Quality Assurance

Post-translation QA checks whether all glossary terms were used correctly, whether the same term was translated consistently throughout, whether there are any new terms that should be added, and whether any intentional departures are documented.

Feedback Loop

After project completion: collect translator feedback on term usability, identify new terms for glossary addition, flag problematic terms for review, update the glossary with approved changes, and sync across all systems and projects.

Key Takeaways

• Glossaries store individual terms with approved translations, not full sentences
• 70% of quality issues stem from terminology problems
• CCU principle: Terms must be Correct, Consistent, Unambiguous
• Regular audits (every 6-12 months) maintain glossary quality
• CAT tool integration turns terminology rules into automatic checks

FAQ

What is the difference between a glossary and Translation Memory?

A glossary stores individual terms (1-3 words typically) with their approved translations. Translation Memory stores complete segments (sentences or phrases) from previous translations. Glossaries keep terminology consistent; TM enables reuse of full translated segments. Learn more in our TM vs. Glossary comparison.

How many terms should a glossary have?

Quality matters more than quantity. Start with 50-100 critical terms — brand names, key product terms, industry-specific vocabulary. Grow based on actual translation needs. A well-maintained glossary of 500 terms is more valuable than an unmaintained database of 5,000.

Who should approve glossary terms?

Term approval typically involves a terminology manager (linguistic accuracy), subject matter experts (technical correctness), in-country reviewers (cultural appropriateness), and legal/compliance (for regulated terms). Establish a clear approval workflow with defined roles.

How often should I update my glossary?

Monthly: Review and add new term suggestions. Quarterly: Audit existing terms for accuracy and usage. Annually: Full review including deprecation of obsolete terms. And update immediately when products change, new features launch, or terminology errors surface.

Can AI automatically extract glossary terms?

Yes, modern NLP and AI tools can identify potential terms from source content based on frequency, patterns, and context. But human review remains essential for approval. AI is great at suggesting candidates — humans make sure they're right.

Build Your Glossary Now

A well-managed translation glossary is your best defense against terminology inconsistency — and terminology inconsistency is responsible for 70% of quality issues. That's not a small number.

Start with your most critical terms, set up a clear governance process, and plug your glossary into translation workflows. The effort pays off across every project.

For how glossaries work alongside Translation Memory, read our TM guide and our TM vs. Glossary comparison.

Ready to validate your translations against terminology standards? Try KTTC for AI-powered quality assessment with Quick Glossary extraction and terminology consistency checking.