String Changes Report

The String Changes Report allows you to view and download the history of changes that were made to a translated string. Common use cases for this report include:

  • Auditing: To track any changes made to the content, and the number of changes.
  • Insights: To learn what was changed from the initial translation to the final published translation.

Access the String Changes Report

  1. Log in to Smartling and select Reports > String Changes Report.
  2. Filter for the information that you need, then click Run Report.
    • Depending on your user role, you will see different filters.


Report Filters

There are a number of Report Filters on the left-hand panel to help refine your report. If you choose to run a report without any filters, the report will pull all data for all Jobs in all projects, over the last 7 days.

You can choose to refine your report by specifying the Project(s), Job(s), Translation Language(s), and User(s) on the account. The Content Filter will allow you to specify if the report should include All Content, Content With Changes, or Content Without Changes.

By default, the report will only include published strings. You can include unpublished strings in the report by selecting the "Show unpublished strings" checkbox. 

There is no column in the CSV download that separates unpublished and published strings.

Data can be retrieved for the past 90 days, and a maximum of 500 items will be shown within the dashboard. However, when you download the content to a CSV, the full result set (anything more than 500) will be available in the file. 

Data is refreshed only once every 24 hrs, so depending on when you run the report, recent string changes might not be included.

For Agency Account Owners, Translation Resource Managers, Translators, Editors, and Reviewers 

All the above applies to your user role except the ability to filter by User(s) on the account. You can view and download content for which you participated. For example, if you're a Translator, you can view and download the translations of the workflow(s) for which you took part.

Report Headers

The CSV download of the String Changes Report will contain the following headers to note;

  • Edit Distance: The number of character changes in the content, between each step.
  • Edit Word Distance (Smartling Dashboard only): The number of word changes in the content, between each step.
  • Edit Distance to Publish: The number of character changes from the current step to the final, published translation.
  • Fuzzy Score: If a fuzzy score is available, it will be displayed. If the content is SmartMatched, no Fuzzy Score will be available. The Fuzzy Score in the downloaded CSV file is represented by a number between 0 and 1000. 
    • 1000 represents a 100% match
    • 995 represents a 99.5% match
    • etc.
  • Outgoing Translation: This group of columns contain translation for all plural forms of a string. In the case where no plurals are contained and only single translation is expected for a string, it will be available in outgoingTranslationNonPlural field.
  • Workflow Action Type: Shows the action that caused a translation to move to the next step in the workflow. Possible values include:
    • TRANSLATION_SUBMITTED: Translation submitted from the translation step.
    • EDIT: Translation submitted from an edit step.
    • REVIEW: Translation submitted from a review step.
    • MOVE: An Account Owner or Project Manager submitted content from a "hold" step.
    • REJECT: A translation that was rejected and sent back to the previous step in the workflow.
    • SMARTMATCH: A translation was SmartMatched from the translation memory (TM) and moved to the next step in the workflow.
    • IMPORT: A translation was imported manually from a file.
    • CUSTOM_MOVE: An Account Owner or Project Manager manually moved content between workflow steps.
    • POST_MACHINE_REVISION: A translation was edited after it was machine translated.
    • AUTO_MOVE_FUZZY_MATCH: A translation skipped a step due to a configuration that allows a translation to meet a fuzzy match threshold to skip workflow steps.

Word counts are space-based for all languages, with the exception of Chinese and Japanese (character-based). It is likely that Edit Distance and Edit Word Distance for Chinese and Japanese will always be equal.

Was this article helpful?