A context match is a translation memory match where the source segment is identical to a previously translated segment and its context also matches. In many translation tools, this is shown as a 101% match because it is treated as more reliable than a standard 100% exact match.
The extra “1%” is not a literal quality score. It is a convention used by some CAT tools and translation management systems to signal that the tool found not only the same source text, but also the same surrounding or structural context.
That distinction matters because identical words can mean different things in different places. A phrase such as “Home,” “Open,” “Save,” or “Current plan” may need a different translation depending on whether it appears in a menu, button, heading, sentence, or software string.
Context Match Definition
In translation memory workflows, source content is split into segments. A segment may be a sentence, heading, button label, table cell, or short software string.
When a translator opens a new file, the CAT tool compares each segment with entries already stored in the translation memory. If the source text is identical, the tool may return a 100% match. If the source text is identical and the context also matches, the tool may return a context match, often shown as 101%. Phrase explains a 101% match as an in-context match, while a 100% match means the source text matches but the new text’s context is different.
A context match is an exact translation memory match where the tool also recognizes that the segment appears in the same or equivalent context as the previously translated segment.
Why Context Matches Matter
Translation memory helps teams reuse previous translations. But reuse is only reliable when the old translation still fits the new context.
For example, the English word “Open” can appear in several ways. It might be a button label, where it acts as a command. It might describe a store’s status, where it means the business is currently operating. It might appear in a technical instruction, where the grammar may need to change depending on the target language. It might also appear as a menu item, where it has to follow established interface terminology.
A 100% match only says the source text is identical. A context match adds another signal: the same text has appeared in the same surrounding or structural situation before.
This is especially helpful for short UI strings, repeated documentation, product manuals, help-center articles, and version updates where many segments remain unchanged between releases.
Context Match vs 100% Match
A 100% match and a context match both begin with identical source text. The difference is whether the tool also recognizes the context.
A fuzzy match means the segment is similar but not identical. It usually needs review and editing.
A 100% match means the source text is identical, but the context may be different or unknown. The translation may be correct, but the translator should still check that it fits.
A context match, often shown as a 101% match, means both the source text and the context match. It is usually a higher-confidence match, but it should still follow the project’s quality rules.
A double-context match or 102% match is a stronger tool-specific category used in some systems when more than one type of context matches. memoQ describes double-context matches as 102% or XLT matches, where both ID-based context and text-flow context are identical.
A context match is stronger than a plain exact match, but it is not the same as an independent quality review.
How Translation Tools Determine Context
Different CAT tools and translation management systems use different context signals.
Previous and Next Segments
For running text, context often means the segment before and the segment after the current segment.
For example, a document may contain this sequence:
Go to Account Settings.
Click Save.
Your changes are now stored.
If the same three-part sequence appears in a later version of the file, “Click Save” may be returned as a context match because the tool recognizes both the segment and its neighboring text.
This type of context is common in documentation, manuals, policies, training content, and long-form text.
Segment IDs, Keys, or Resource Identifiers
For software localization, context may come from a string key or resource ID instead of neighboring sentences.
A UI string such as “Save” might have a key like:
account_settings.save_button
That key can tell the tool that “Save” belongs to a specific button in account settings. In structured formats, that context may be more useful than the segment before or after it.
This is why context matching can behave differently in software files, CMS exports, JSON files, resource files, and other structured content.
File Structure and Formatting
Some tools may also consider document structure, tags, formatting, or imported file behavior. This is one reason match percentages can change between tools or projects.
A segment may look identical to a human reader but be treated differently by software if tags, placeholders, punctuation, or formatting differ.
Why Is It Called a 101% Match?
The term 101% match is a software convention. It means the match is considered more specific than a 100% exact match because context has also been matched.
It does not mean the translation is “101% correct.” It does not mean the segment should always be accepted without review. It simply tells the user that the match carries an additional context signal.
In production workflows, a project manager may choose to pre-translate, lock, discount, or route context matches differently from 100% matches. Those decisions depend on project risk, client requirements, translation memory quality, and quality assurance policy.
Context Match Example
Suppose a translation memory contains this sequence:
Account settings
Save changes
Your profile has been updated.
If a new file contains the same sequence, the segment “Save changes” may appear as a context match.
Now compare it with a different sequence:
Document editor
Save changes
Continue editing your file.
The source segment is still “Save changes.” A tool may still find a 100% match because the words are identical. But because the surrounding context is different, it may not return a context match.
The old translation might still be correct. The changed context simply tells the translator that the segment deserves attention.
Context Match, Perfect Match, and Double-Context Match
Localization tools do not always use the same terminology. Three terms are commonly confused.
Context Match
A context match usually means the source segment matches a translation memory entry and the relevant context also matches. In many tools, this is shown as 101%.
Perfect Match
In some translation environments, “Perfect Match” refers to a match created by comparing updated source files with previously translated bilingual files, rather than relying only on a translation memory. RWS describes PerfectMatch as a form of context match that compares updated source files with existing bilingual files instead of a translation memory.
This can be useful for updated versions of previously translated files. However, because “Perfect Match” is tool-specific terminology, it should not be used as a generic synonym for “approved translation.”
Double-Context Match
A double-context match usually means that two forms of context match. For example, a tool may check both the surrounding text and an ID-based context such as a string key.
This is most relevant when content has both running-text context and structured identifiers, such as software strings, CMS content, or files with stable segment IDs.
What to Do When You See a Context Match
A context match is usually a strong reuse signal. How it should be handled depends on the project.
For low-risk content, a team may decide that context matches can be pre-translated and reviewed lightly. For high-risk content, such as regulated, legal, medical, financial, or safety-related material, context matches should still be reviewed by a qualified human reviewer.
Before accepting a context match, check whether the translation memory source was reviewed and approved. Confirm the file type, because running text and structured software strings may use different context logic. Review variables, tags, placeholders, numbers, and formatting. Check whether terminology has changed since the translation memory entry was created. Follow client instructions, especially if the client requires review of all matches.
A 101% match does not mean the same thing in every tool. Context matching is useful, but it is still shaped by software rules, project settings, and translation memory quality.
When Context Matches Are Most Useful
Context matches are most valuable when the translation memory is clean, reviewed, and relevant to the current project.
They are especially useful for updated product documentation, repeated legal or policy templates, software UI strings with stable keys, help-center articles with recurring instructions, versioned content where only some segments changed, and large localization programs with consistent terminology.
They are less reliable when the translation memory contains unreviewed legacy translations, the product or terminology has changed, the content is highly regulated, the source text is short and ambiguous, the file has complex tags or variables, or the tool lacks enough context data to compare.
Can Context Matches Reduce Translation Cost?
Sometimes, but not automatically.
Many localization workflows use match categories to estimate effort. A file with many context matches may require less new translation work than a file with mostly new or fuzzy segments. However, pricing policies vary by language service provider, client, tool, risk level, and review requirements.
A context match should not be treated as “free work” by default. It still depends on whether the translation memory is trustworthy and whether the project requires review.
For buyers, the better question is not simply, “Do we get a discount for 101% matches?” A more useful question is:
Which match categories are reviewed, which are locked or pre-translated, and what quality checks apply to each one?
That question creates a clearer conversation than relying on match percentages alone.
Conclusion
A context match is an exact translation memory match with matching context. It is often shown as a 101% match because it gives translators and project managers more confidence than a standard 100% match.
The concept is simple, but the implementation is tool-specific. Some systems use neighboring segments, some use string keys or IDs, and some use multiple forms of context. A context match is best understood as a strong reuse signal, not a guarantee.
Used well, context matches help localization teams work more consistently and efficiently. Used carelessly, they can repeat outdated or unsuitable translations. The safest approach is to combine match data with clean translation memories, clear review rules, and professional judgment.
