101% matching

To understand 101% matching, it helps to start with 100%. A 100% match means the current segment is word-for-word identical to a segment already stored in the translation memory. The existing translation can be reused directly without any editing.

A 101% match goes one step further. Not only is the current segment identical to the TM entry, but either the segment immediately before it or the segment immediately after it also matches. This surrounding context, the neighboring segments, confirms that the segment is appearing in the same position and meaning as it did when it was originally translated. A 102% match extends this further: both the preceding and following segments match, providing the maximum level of contextual confirmation.

These matches are sometimes referred to as in-context exact (ICE) matches or context matches, depending on the CAT tool.

🔍 Why context matters beyond 100% #️⃣

A 100% match at the text level is not always enough to guarantee the stored translation is appropriate. Consider a word like “Close”, identical text, but meaning “to close a window” in one context and “close proximity” in another. If only the segment itself matches without any context, a CAT tool cannot distinguish between the two uses. A 101% or 102% match narrows the ambiguity significantly by confirming the surrounding content also matches, making it far more likely the stored translation applies correctly.

This is particularly relevant in software localization, where short UI strings appear in many different contexts and identical text can carry different meanings depending on what surrounds it.

💰 Key points about 101% matching #️⃣

• 101% and 102% matches carry higher confidence than standard 100% matches because context is verified, not just the segment text itself.
• Many LSPs and project managers charge less, or nothing, for ICE matches, since the additional context confirmation means human review is rarely needed. This makes them one of the most valuable levers for reducing translation costs on repeat or update projects.
• Not all CAT tools implement context matching in the same way. Some use the segments immediately before and after; others use document-level context or metadata. Understanding how your specific tool calculates these matches helps set realistic expectations for TM leverage.
• For billing purposes, 101% and 102% matches are typically grouped with 100% matches under “no-charge” or “full-leverage” tiers, or priced at a minimal review rate.
• High percentages of 101%+ matches on a project are a strong indicator that the TM is well-maintained and that content is being updated rather than rewritten from scratch, a healthy sign for any mature localization workflow.

🔄 101% vs. 100% vs. fuzzy match at a glance #️⃣

🛠️ How Localazy supports context matching #️⃣

Localazy provides “Context Match” certainty through Duplicity Linking and key-based context. Instead of guessing based on neighboring text, Localazy uses the developer key (e.g., settings.btn_save) to anchor a string’s meaning. Using the feature, you can link identical strings across your project so they sync automatically. Once a translation is verified for a specific key, it is “locked in” and reused everywhere that key appears. This gives you the same reliability as a 101% match, ensuring that once a string is right in its specific context, it stays right without needing a second look.

Learn more about managing duplicities and Translation Memory in our docs.