Translation Quality

Quality Confidence Score

The Quality Confidence Score (QCS) is a fundamentally new approach to Quality Management that enables organizations to recognize and address quality issues for all translations.

QCS is a prediction of translation quality outcomes in Smartling. It is a 1-100 score that indicates the likelihood that a translation is of “professional quality” if it were to be manually evaluated by a human linguist.

Viewing Quality Confidence Scores

A Quality Confidence Score (QCS), is calculated after a Job that contains 200 or more words is completed in Smartling. To see QCS for completed jobs, log in to Smartling and follow these steps: 

  1. Select the Jobs menu from the top navigation.
  2. Select All Jobs from the left menu.
  3. Filter Status for Completed jobs.
  4. Click the Job name to enter the Job details and click the Quality tabScreenshot_2021-04-01_at_14.39.51.png

Interpreting Quality Confidence Scores

The Quality tab in Smartling provides a summary of the 7 key drivers that influenced the calculation of QCS for the job. These are visualized in a radar chart similar to the one below. Green indicates that a driver had a positive impact on the Quality Confidence Score for the job, and red indicates that the driver had a negative impact on quality confidence for the job.


Clicking on one of the score drivers generates an Insight. For example, Resolving Translation Issues can improve Quality Confidence for a job by up to 8%.

For this job, 10 issues were created, and 2 issues were resolved. This is 22% lower than the average Issues Resolution Rate in Smartling.

As a starting point for understanding the Quality Confidence for a job, consider looking at the strongest positive driver (green) and strongest negative driver (red). In the above chart, Translation Memory Leverage had the strongest negative impact on the score, and Visual Context had the strongest positive impact on the score. Every Insight should lead to an action that you and your team can take to improve quality in Smartling.

Was this article helpful?