Quality Assurance

Quality assurance (QA) is the process of reviewing customer interactions with Team Members to ensure consistent quality and to discover coaching opportunities. For example, while a five-star interaction might appear at first glance to be of high quality, an in-depth QA review might reveal that the Team Member gave the customer incorrect information, leading to more frustration for that customer later.

The Medallia Agent Connect QA experience focuses on real-time coaching that happens shortly after a customer interaction occurs. This approach is much more effective than waiting days, weeks, or even months to provide Team Members with valuable performance feedback. Team Members have access to their QA reviews immediately after they are completed, so they can review and make immediate performance improvements.

QA overview

QA reviewers can start a review for any interaction on an ad-hoc basis, as described in Reviewing interactions. To make QA as intuitive and actionable as possible, QA reviewers can also add annotations directly to interaction transcripts. These annotations are then visible to Team Members within the context of their interactions.

Starting a QA review

You can also create Assignments, which push a quota of random interactions to pools of reviewers automatically, on a schedule.

Reviewers base their reviews on the Scorecards you create. A scorecard provides the standards against which interactions are compared, and ensure that reviewers focus on the quality standards and coaching opportunities that are important to your company. For example, you might want all Team Members to begin and end all interactions with a greeting and a thank you. By adding these standards to a scorecard your reviewers must include those elements in each review.

Calibrations help remove subjectivity from reviews by measuring whether reviewers are reaching similar conclusions after reviewing the same interaction. This ensures that Team Members receive consistent coaching aligned with company goals.

Optionally, your company might use Acknowledgments and appeals.. An acknowledgment is a simple, one-click task for Team Members that lets reviewers know that the Team Member has read and understands the contents of a QA review. An appeal is a way for Team Members to ask reviewers to take another look at specific questions or answers on a review scorecard.


The Agent Connect QA solution includes reporting functionality that helps teams quickly understand the performance of a QA program over time, get granular insight about areas where the team excels and needs improvement, and observe trend details to better inform coaching. For information, see QA Dashboard.

Sensitive data

Team Leaders and Admins can configure how many days interaction transcripts remain available in QA reviews. This threshold helps protect customer data from being available indefinitely. To set the threshold, open the QA settings screen, and then modify the Expungement threshold field.

The Expungement threshold field on the QA screen

After a number of days that exceeds the threshold, transcript text is starred out in reviews, and call recordings are replaced with a blank file. Changes to the threshold setting also apply to any previously-completed reviews.

Enabling QA functionality

Admins must enable specific QA functionality on the Company Info screen for QA to be available for the company.

  • In the Products section, turn the QA field On.
  • To use email notifications for completed reviews, turn the QA review notifications field On.
  • To use acknowledgments and appeals, turn the QA Acknowledgements & Appeals field On.
  • To use acknowledgments, and not use appeals, turn the QA Acknowledgements only field On.
  • If you do not want Team Leaders to be notified about new calibrations, turn the Disable QA Calibration Notifications field On.

If acknowledgments and appeals have been turned on for your company by an Admin, Team Leaders can activate the functionality on the QA settings screen.

Acknowledgments and appeals settings on the QA screen