• How many duplicate definitions exist across the company for a specific concept? Each duplicate definition represents a duplication of research, writing, and editing tasks, and possibly duplicate localization costs in each target language.

• How many synonym pairs exist for a specific product? It is difficult to assign an exact cost to synonyms, but at the very least they will entail higher localization costs (in the form of higher word counts and time spent researching and untangling the synonym pairs in each supported language).

• How many terminology-related “bugs" are filed during a product cycle and on average how much time does it take to resolve them? Many companies already track this metric, and being able to show a substantial reduction here may be a good foundation in making a case for managing source terminology.

• In how many languages is the software published? The more languages a company needs to support, the more cost-effective it will be to invest in source-language terminology management.

• How much does it cost to localize each word? This cost will vary by localization vendor, language, and volume.

• How many fuzzy matches would become 100% matches if the source terminology were managed? During the localization process, 100% matches cost less.

• Are there any terminology issues that have ended up costing extra money because they had to be fixed at an expensive point late in the development process (after the product shipped, for example)? In addition to expense, mistakes arising from unmanaged terminology can result in legal action or simply embarrassment. Often, one or two such errors suffice to drive home the value of source terminology management.

Answers to these questions can provide the foundation for ROI measures. Each of the answers above reflects potential expenditures that a software company wouldn't have to make if it managed source-language terminology.

A company may need to start small and focus on just one domain or product that matters to upper-level management. After operating for 25 years without actively managing source-language terminology, for example, Microsoft began with just one group. We have since moved on to support other projects that are important to upper-level management, such as security, developing more robust processes, communication paths, and tools. Without upper-level management support, we might never have made headway toward a single set of terms and definitions.

Every company that has successfully implemented terminology management can testify to the fact that it requires a lot of dedication and determination. Setting up a database and convincing people what it costs not to manage terminology is only part of the battle. One must also change people's attitudes toward the importance of consistent, transparent terminology in software.

责任编辑：admin