Localization management software, localization tools and localization solutions are the same category under different names, and for a software team they have one job: keep the string catalog true across releases so that the release can be costed from the memory, the handoff generated from the record, the translations read back against their keys and the locale build scored before it ships. Localization technology is sold in three shapes, the enterprise platform, the agency's portal and the team's own tool, and the shape decides who owns the catalog. This page sets out what localization management is for a software product, what the software localization tools have to do at each step of a release, and what a team gives up with each shape.
Localization management is the release loop, kept as a record
Localization management is not a project plan, it is the loop every release runs: freeze the strings, cost the release, hand off, read back, test, ship, and start the next one from a catalog that is now larger and a memory that is now better. A localization tool that keeps the loop as a record, with each release's leverage sheet, handoffs and test results filed against it, makes the next release's cost and risk a reading rather than a guess. The free translation memory leverage sheet and localization testing checklist on this site work the two figures of the loop with no account.
The three shapes of localization technology, and who owns the catalog
An enterprise platform holds the catalog in the platform, with connectors, vendor management and machine translation around it, and the team owns it as long as the contract runs. An agency's portal holds the catalog with the agency, which is convenient until the agency changes. A team's own tool holds the catalog with the team, keyed, exportable, with the handoff generated from it and the memory in an open format. Localization solutions differ less in features than in that answer, and the answer decides what a team keeps when it leaves.
What software localization tools have to do at each step
At the freeze, import the changed keys from the formats the developers ship and count the words. At the costing, read the memory's exact and fuzzy match shares against the release. At the handoff, generate XLIFF per locale from the record. At the read-back, match the translator's file against the keys and validate placeholders and plurals. At the test, take the defect counts against the ship threshold. At the ship, export the catalog in the build's formats and file the release. A localisation software that does the middle four and leaves the first and last to a script is fine; one that cannot export is not.
Questions people ask about localization management software
Are localization management software and a translation management system the same thing?
Largely, yes: the TMS name comes from the translation industry and the localization management name from software teams, and both keep the catalog, the statuses and the memory. The difference is emphasis, vendor routing on one side and the release loop on the other.
Do we need localization tools if we use a translation agency?
The agency has its tools; the team needs the catalog. A team that keeps its own keyed catalog and memory can change agency without losing a string, and can read the leverage before accepting a quote, which is the figure the agency's tools will not show you.
What is the smallest localization tool that does the job?
A keyed catalog with a status per locale, an XLIFF handoff generated from it and read back, a memory in an open format, and an export. Everything else is a convenience the team can add when the locale count grows.