We've launched a new support center at help.smartling.com. This site will continue to be maintained into January 2017. Please update your bookmarks.

Important changes to Translation Memory Leveraging

Smartling is changing the way we leverage the existing translations in your Translation Memory. These changes will bring the way Smartling translations are reported and billed in line with industry standards and make Fuzzy Estimates and Word Count reports more accurately reflect the work performed by translators.

Previously, translations were leveraged at the level of whole strings, rather than segments. From July 1st, translations will now be leveraged at the segment level.

For example, take a string with three segments, and a total of 47 words. The whole string does not match any existing strings in the Translation memory, but two of its segments match exactly to existing segments in the Translation Memory.



Segment-Level Match

Segment 1



Segment 2



Segment 3



Here’s how this string would be counted in Fuzzy Estimates and Word Count reports, before and after July 1st.

Fuzzy Match Bracket

Words before July 1st (comparing whole strings only)

Words after July 1st (comparing segments)

0 to 85%

47 words

13 words

85% to 95%

0 word

0 words

95% to 100%

0 words

0 words

100% to 100%

0 words

34 words


One additional effect of these changes is that unsegmented content in the Translation Memory is no longer leveraged in Fuzzy Match Estimates and Word Count Reports. You may have unsegmented content your Translation Memory if it includes:

  • External translations imported into Smartling via our File API
  • Content translated in Smartling prior to the adoption of segmentation in 2012
  • Content translated using the the  Offline CAT feature.
Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request
Powered by Zendesk