QA Assignments
Updated over a week ago

QA Assignments

With Agent Connect’s QA assignment feature, you can assign a customized QA completion target for your QA analysts or team leaders, and easily keep track of progress against your targets. Agent Connect will automatically distribute reviews to reviewers and select interactions for them to review.

This feature is only available to clients sending Agent Connect survey requests in an integrated fashion. For more information on getting integrated, contact your Client Success Manager.

Note: Only Admins are able to create assignments

Creating a New Assignment

To create and configure a new QA Assignment:

  1. Click the QA tab in the main navbar of Agent Connect

  2. Click Assignments in the dropdown menu

  3. Click on the New Assignment button

You will now be taken through a series of steps to configure your new assignment.

Step 1: Name your assignment

This name will help you identify the assignment from the Assignments index page. You might choose to categorize your assignments by team name, by interaction types, or some other meaningful name for your team.

Step 2: Specify a cadence

A cadence specifies the number of days in the assignment. You can select a cadence of either ​Weekly​, ​Monthly​, or ​Custom​. ​Weekly cadences begin on Monday and end on Sunday. Monthly cadences begin on the first day of the month and end on the last day of the month.​​

​If you selected a custom cadence, in the ​Start Date​ field select the first day of the cadence, and in the ​Number of days​ field select the duration of the cadence.

An assignment cadence can ​begin​ in the past, but the entirety of the cadence cannot be in the past. For example, you can configure a weekly cadence in the middle of the week, even though that cadence begins on a Monday. You cannot configure a custom cadence that ends before the current day.

Step 3: Select a default scorecard

When reviewers work on the assignment, this scorecard is selected for them. Reviewers can change the scorecard used when starting a review.

Step 4: Add participants

Next, you'll select which teams or team members should be reviewed as part of this assignment.

This allows you to create assignments to review any team on one particular team, or any team member regardless of team leader.

*You can choose to review a user who is a Team Leader as a participant should you want to review them as well.

Note: A team member is not eligible for an Assignment until their profile is approved and "Active."

Step 5: Select reviewers

You can choose whether your team members should be evaluated by their own team leader if you are assigning by team.

If Allocate to team leader is selected, all reviews for each team member will be assigned to their direct team leader automatically. Selecting this option will disable the ability to indicate the reviewers.

mceclip0.png

Conversely, if your workflow requires that specific individuals complete reviews for your team members, you will need to indicate which users should be considered as reviewers. Agent Connect will calculate the number of reviews to be assigned to each reviewer. All reviews will be evenly distributed across all designated reviewers.

mceclip1.png

While adding reviewers, you also have the option to set reviews as "blind". Selecting this option will hide the team member name and photo from the reviewer while they are completing the review.

mceclip2.png

Step 6: Set filters

Determine whether you want to scope any interactions types out of the assignment using the exclude tags filter. This option may be useful for your team if certain interaction types are generally not worth reviewing e.g. if a simple macro is used. This option can help keep the pool of interactions available for review more meaningful.

Now, you can use any field in your CRM/source system to build eligibility criteria for QA Assignments to ensure you are pulling the right interactions in your automated queues of work for reviewers for each assignment.

On the Integrations page in Agent Connect, clients using Agent Connect QA will see their Custom Properties and will be able to tell us the Field Name they’d like us to use for this Custom Property, tell us which data type we should use for the Custom Property, and lastly, whether they should be visible on QA Reviews or available to use as a filter for QA.

Step 7: Set the review target

Indicate how many reviews should be completed as part of this assignment by entering the expected number of reviews per team member. You can allocate reviews across specific criteria like interaction channel, tag, or CSAT score. If you do not have any requirements for the type of interaction expected to be reviewed, simply enter the number of reviews using Any interaction.

For example, if you expect four reviews total for each team member to be completed, and would like at least one review to be completed for a phone interaction, and at least one review to be completed for an email interaction, but the other 2 can be for any interaction, your assignment quantity selection would look like the example below.

mceclip0.png

In the right-hand side column, you'll see a breakdown regarding the number of total reviews your assignment will require, as well as a breakdown showing how many reviews each reviewer will need to complete.

mceclip4.png


Please note: Review distributions across analysts may result in a decimal number. If this is the case, then each reviewer will be assigned a total of the closest rounded down whole number. For example, if Reviews per reviewer is equal to 6.5, this means that each reviewer will be assigned to review at 6 interactions.

Step 8: Publish assignment

Once published, Agent Connect will automatically begin to distribute interactions to reviewers based on the parameters set in your assignment. Interactions that fit your parameters are selected randomly in order to eliminate bias in the review process.

However, interactions will always be pulled from a rolling 7-day look-back period. This ensures that reviewers are never reviewing interactions that occurred longer than 7 days ago. For this reason, there may be a different number of interactions eligible for review as the assignment progresses.

Once configured, the assignment will run automatically per your cadence, and the assignment progress for each reviewer will reset until the assignment is archived. If edits are made to the target criteria, adjusted assignments will be reflected at the start of the next assignment period.

Scheduling and Frequency

Weekly & Monthly Assignment Cadences

Assignments start at 12:00:00 UTC at the start of your cadence, and end at 23:59:59 UTC on the last day of your cadence. For example, a weekly assignment cadence will start on Monday, at 12:00:00 UTC and end on Sunday, 23:59:59 UTC. A monthly assignment cadence will start on the 1st of the month, at 12:00:00 UTC and end on the last day of the month, 23:59:59 UTC.

Custom Assignment Cadences

Custom assignment cadences start on the date and time that a user configures/sets the assignment cadence up for and repeats based on the number of days configured. Please note that you can set up your custom cadence to start at a certain time in your local timezone. Agent Connect displays help text to convert that local timezone into UTC so globally distributed teams can understand at what time a new assignment cadence begins.

Allocating Reviews

The narrower the criteria for assigning reviews, the more likely it is that Agent Connect will not have any reviews to distribute to your reviewers.

If your reviewers do not have enough eligible interactions you might consider the following:

  1. Expand your assignment criteria.

  2. Wait! Because we distribute reviews from a rolling period of time, it is possible that more eligible interactions will occur and can be distributed to your reviewers during upcoming days.

Note that even if you do not have narrow criteria, it is still possible that Agent Connect does not have any interactions available to assign. Check back later, as your team members handle more support interactions, more work will become available for your reviewers. Or, you may want to consider leveraging our QA Assignments Out of Interactions Preference setting.

Reviewing Interactions

If you are a reviewer and you'd like to begin reviewing, click on the QA tab in the main navigation bar at the top of the screen, then select Assignments from the dropdown menu.

You will see a list of all assignments for which you are a selected reviewer. You'll also see details about the Assignment (cadence, due dates), the Status (New or In Progress), the Progress of the Assignment, a list of Team Members to be reviewed and a list of Reviewers assigned to the Assignment. From here, you can take additional actions, such as copy the Assignment to create a new Assignment based on the existing criteria, archive the Assignment, click to see more details about the Assignment or review the Assignment.


To begin your work, select Review. Agent Connect will automatically select an interaction that meets your assignment criteria for you to review. We will continue to select interactions for you that match your assignment criteria until you've met your target for the assignment, or until eligible interactions run out. If there are no eligible interactions, check back later or utilize the QA Assignment Out of Interactions Preference setting to serve up additional interactions that may not match the original criteria.

When a QA Reviewer starts reviewing an Assignment, they can click on Skip Interaction if the interaction isn’t eligible for review at that point in time. The QA Reviewer may select one of our preset skip reasons or choose Other. When Other is selected, the QA Reviewer is prompted to type the reason. Please note that the Other open text form will populate other reasons that other QA Reviewers might have inputted for that QA Assignment. This is to reduce the amount of variations of similar Other skip reasons. Once the QA Reviewer confirms the skip by clicking Skip Interaction, that reason will be saved for reporting purposes. For more information on skip reason reporting, please see Skip Reason Reporting.

Once Skip Interaction has been selected, Agent Connect will automatically assign a different interaction for review, and will mark the skipped interaction as not eligible for the assignment.


Reviewers can track their progress from their assignments list, and whenever they have completed their work for the assignment, we will celebrate!

Tracking Assignment Progress

You will be able to track overall progress for each assignment in the Overview section, where you will see a count of completed reviews, remaining reviews needed to hit the overall target for the assignment, and a progress bar. Click Details in order to see a more in depth view of the Assignment.

In Assignment Details, you can get high level statistics about the Assignment, including understanding the current progress for the Assignment, when the Assignment started/how long it's been running for, how many Team Members are to be QAed as part of the Assignment, how many Team Leaders have been assigned to review the Assignment and the goal for number of reviews to be completed. Additionally, you can easily identify Team Members who are Champions (high achievers compared to other Team Members within that Assignment) or Team Members who may need additional coaching.

Skip Reason Reporting

Within Assignment Details, Agent Connect provides in-depth reporting on the reasons for why an interaction has been skipped. Because QA Reviewers are required to provide a reason when they now skip interactions in Assignments, Agent Connect displays these metrics in the QA Assignment Details page.

A QA Manager can evaluate the trends for why interactions are being skipped, review the transcripts that were originally skipped by clicking into View Interactions upon hovering the reasons and request QA Reviewers to review if it was an interaction that shouldn't have been skipped.

Please note that in order for a QA Reviewer to review the interaction again, they must do so outside of the QA Assignment by clicking on the New Review button from the QA top navigation dropdown menu or on the Completed reviews tab (they will not be able to review the skipped interaction directly from the Skipped Interactions page).

Archiving Assignments

You can archive an assignment by clicking on the Archive button while viewing an assignment on the Assignments list. Archiving will disable the assignment for all of the selected reviewers. Past reviews completed as part of an archived assignment will remain active and will be accessible as normal.

How Adding or Removing Agent Connect Users Will Affect Your QA Assignments

As you add new users or deactivate existing users on the Manage Team page, Agent Connect will automatically update existing QA assignments to reflect the current state of your team's hierarchy.

Removing reviewers

If you have an assignment where you have specified reviewers, and you deactivate one (or more) of those reviewers on the Manage Team page, their share of the remaining reviews within the assignment will be evenly distributed across the other designated reviewers.

Removing team members

If you deactivate a team member that is to be reviewed as part of an assignment, the review target will be automatically updated to reflect the status of the hierarchy. For example, if your assignment requires reviewing interactions for 10 team members across 2 teams, where the target is to complete 2 reviews per team member (a total of 20 reviews), and one team member is on either team is deactivated, the new target for the assignment will be 18 total reviews.

Adding team members

If you add a team member to a team that is covered by an existing assignment, that team member will be automatically included in the assignment when progress resets at the start of the following assignment period, and the target number of total reviews for the assignment will automatically be increased.

Peer Review Assignments

At the moment, our assignments feature is not designed for peer-to-peer reviews. You are able to create assignments for members of teams on the team level, and the ability to perform a review is limited to team lead and higher. Please also note that team leads are not able to be reviewed.

QA Assignment Out of Interactions Preference

Note: this feature is currently in beta and is not Generally Available at this point. Please reach out to your Agent Connect representative to get further information.

What is it?

The QA Assignments Out of Interactions Preference gives Medallia Agent Connect customers the option to choose what happens when a QA Assignment runs out of interactions.

Why does it matter?

There are many reasons why a QA Assignment could run out of interactions (eg: a Team Member who must be reviewed as part of the Assignment lost their voice and couldn’t take phone calls for a week; another Team Member was on vacation for a week and had no interactions in that timeframe). When a QA Assignment runs out of interactions today, QA Reviewers are unable to complete the Assignment 100% because Agent Connect is unable to serve up interactions that meet the original criteria.

Providing some flexibility for the QA Assignment where QA Reviewers can be served up the next best interaction (even if it doesn’t match the original QA Assignment criteria completely) can help reduce frustration for QA Reviewers and allows teams the opportunity to get closer to completing a QA Assignment.

How does it work?

Turning on Out of Interaction Preference

A Company Admin must first navigate to Settings > QA and scroll down to Assignments. Under Review Similar, flip the toggle to ON in order to surface the Out of Interaction Preference setting.

Setting Up a QA Assignment with Out of Interaction Preference

When a new QA Assignment is set up, the Allow Partial Match checkbox should be selected. This will ensure that when the QA Assignment runs out of interactions that meet the original criteria, it will automatically pull in similar interactions, with the criteria relaxed.

Out of Interaction Preference in Action

When a QA Assignment has run out of eligible interactions based on the original criteria, a QA Reviewer will see the following message:

When a user clicks on the Review Similar button, they will be served an interaction that best matches the original criteria, while relaxing certain other criteria.

Alternatively, a QA Reviewer can return to QA Assignments and wait for an interaction that matches the original criteria before continuing QA.

How Does Agent Connect Serve Similar Interactions?

Agent Connect uses 3 steps to prioritize the next best interactions to serve:

  1. Use guidance from the “Exclude interactions from more than x days” criteria

    1. If “Exclude interactions for more than 7 days” was set, Agent Connect will pull interactions to serve from the 7 days prior to that. This setting is capped at 6 months (ie: we will not pull interactions older than 6 months) as any interactions older than 6 months may be considered out of date and irrelevant for QA.

  2. Within the batch of interactions from the first step, interactions will be sorted by the highest matching criteria/conditions from the original QA Assignment. For example:

An original QA Assignment criteria was:

  1. Star rating: 1-3 stars = 5

  2. Integrated channels = 3 phone

  3. Number of interactions with Tag = 3 shipping

Interaction X fulfilled the following criteria:

  1. Star rating: 1 star

  2. Integrated channels = phone

Interaction Y fulfilled the following criteria:

  1. Star rating: 2 stars

In this scenario, Interaction X would be served for QA before Interaction Y is served because Interaction X matches on 2 criteria from the original QA Assignment whereas Interaction Y only matches on 1 criteria.

3. If interactions have the same number of criteria matching, they will be ordered by the interaction creation date. For example:

An original QA Assignment criteria was:

  1. Star rating: 1-3 stars = 5

  2. Integrated channels = 3 phone

  3. Number of interactions with Tag = 3 shipping

Interaction X fulfilled the following criteria:

  1. Star rating: 1 star

  2. Integrated channels = phone

  3. Interaction creation date = 2022-10-02 01:05:00 UTC

Interaction Y fulfilled the following criteria:

  1. Star rating: 2 stars

  2. Integrated channels = phone

  3. Interaction creation date = 2022-10-01 15:25:00 UTC

In this scenario, Interaction Y would be served before Interaction X because they both match on 2 criteria but Interaction Y was created before Interaction X.

Did this answer your question?