Fuzzy Match
A translation memory match that is similar but not identical to the source text.
Definition
A fuzzy match occurs when a new source segment is similar but not identical to a previously translated segment in the translation memory. Fuzzy matches are shown as percentages (e.g., 85% match). Translators review fuzzy matches and make small adjustments instead of translating from scratch. This significantly speeds up translation of repetitive content.
Examples
- →New: 'Save all changes' → 85% match with existing 'Save changes'
- →New: 'Delete selected items' → 75% match with 'Delete items'
- →New: 'Welcome back, John!' → 90% match with 'Welcome back!'
Frequently Asked Questions
What fuzzy match percentage should I trust?
90-99%: Minor differences, usually safe with quick review. 75-89%: Moderate differences, needs careful review. Below 75%: Significant differences, may not save much time. Most TMS platforms let you set thresholds for auto-suggesting matches.
Do fuzzy matches reduce translation costs?
Yes. Translation pricing often uses tiered rates: 100% match (lowest cost), 85-99% fuzzy (moderate discount), 75-84% fuzzy (small discount), no match (full rate). Cumulative savings can be 20-40% for repetitive content.