How to choose a terminology management tool in six steps

Start with discovery, not demos

Before you look at any tool, document your current state: existing terminology resources, workflows, the technology ecosystem and the people involved. Then define the desired future state. Two questions shape almost every requirement:

• Who will use it? A handful of terminologists, or hundreds of writers, engineers and reviewers who only need quick lookup?
• Where must terms appear? In your TMS and CAT tools, authoring tools, machine translation, CMS, or an AI assistant?

Requirements checklist

• Central, web-based access from anywhere, as a single source of truth.
• Effortless lookup with little or no training for basic users.
• Customizable entry structure and search filters that match your metadata model.
• Workflows and access rights for term requests, review and approval.
• Tool independence: import and export in standard formats such as TBX, so data can move into any CAT tool.
• Integrations with your TMS, CAT, authoring and content systems, via connectors or an API.
• Scalability for more languages, users and business units.
• Total cost of ownership: licences, technology, support and maintenance, not just the price per seat.

The tool landscape

Terminology functionality comes in three broad flavors:

• Dedicated terminology management systems such as Kalcium Quickterm and TermWeb, built for enterprise-wide workflows and broad user access.
• Termbases inside translation tools, such as Trados MultiTerm, memoQ and QTerm, Phrase or Wordbee, convenient when translation is the main use case.
• Authoring and content-governance tools such as Acrolinx and Congree, which enforce terminology at the source while writers create content.

Many organizations combine them, for example a central termbase that feeds both the translation tools and an authoring assistant. As an independent consultant, I have worked with all of these and recommend whatever fits the stack and the budget.

A six-phase roadmap

1. ROI calculation: validate the business case (see terminology management ROI).
2. Discovery and recommendations: compare the current and desired state, define technical requirements, shortlist tools.
3. Proof of concept: import a sample of real multilingual terms, gather stakeholder feedback and document gaps in an evaluation report.
4. Implementation: centralize, clean up and import term resources, set up workflows, train users.
5. Integration: connect the termbase to TMS, CAT, MT and authoring tools and automate data exchange.
6. Management: maintain the termbase, add terms for new launches and keep governance running.

Common mistakes

• Choosing on features alone and underestimating the clean-up and migration effort.
• Designing for terminologists and forgetting the casual users who make up most of the audience.
• Skipping the proof of concept, then discovering integration gaps after the contract is signed.
• Going live without owners, so the termbase slowly goes stale.