- Coaching and Development
- Measures and Results
Who is eligible to receive performance exception notifications?
Any Performance Management user who has an out-of-range daily performance result for a calculated leaderboard, regardless of role.
How often are performance exception notifications sent?
Once per day, as soon as the system calculates leaderboards for the prior day (typically in the early morning hours of the timezone in which the account is installed).
Are performance exception notifications sent for daily, weekly, and monthly results?
Notifications are only sent for daily results. If a performer’s result would be out-of-range based on a weekly or monthly aggregation, an additional notification is not triggered.
Are performance exception notifications to be triggered on every single queue?
Notifications will be triggered for any and every out-of-range result on a unique leaderboard. Since leaderboards are calculated separately by queue, there is the possibility of having out-of-range results on multiple queues, including the queue choice “Non-queue.” Note: Notifications are not triggered for the aggregate queue combination of “All.”
Are results shown in the performance exception notification?
No. The measure and queue are shown, and the date may be assumed to be the prior day. Recipients may view the result values by clicking on the notification to drill through to the leaderboard, and view the results in context.
How is a result determined to be a performance exception?
Individual results that depart 1.5 standard deviations, or more from the mean result (Average) for each unique leaderboard calculation, will trigger a notification.
Why is an excellent result triggering a performance exception? Shouldn’t this be for poor results only?
Individual results that depart 1.5 standard deviations or more from the mean (Average) trigger a notification, regardless of whether they are considered “good” or “bad.” The purpose of exception notifications is to bring awareness to unusual performance results.
Will multiple performance exceptions be shown in a single notification?
No. There is a separate notification for each performance exception triggered. A digest-style notification is not possible today.
As a management’s member, can I see the performance exception notifications that performers have received?
Currently, only the performer may see the notification. However, management may view the leaderboard in its entirety to interpret results, including top and bottom performers. Results that have triggered a performance exception will be shown in red and a bell icon will be displayed.
Is there a way to deactivate performance exception notifications?
Presently, the only way to prevent performers from receiving exceptions is to remove them as users from the “Performance Management” application.
How often are the leaderboards generated?
Leaderboards are generated daily in the early morning hours of the timezone, in which the account is installed (data to generate the leaderboards is gathered hourly, however.)
How far back does leaderboard data go?
Leaderboard data will be calculated 31 days into the past at the time of installation; it will be retained inside the Performance Management application for a minimum of one year.
Where does leaderboard data come from?
Leaderboard data originates from data written to the Talkdesk data platform by the individual applications that generate the source data. For example, “Average Handle Time” is generated by Talkdesk’s CCaaS system, whereas “Average CSAT” is generated by the Feedback application.
What time zone is being used for leaderboards?
Leaderboards are calculated and displayed in the time zone in which Performance Management is installed. Neither user’s team time zone, computer settings nor browser settings change this. The scope of data included on a leaderboard and the generation time of the leaderboard may not match expectations for users in different time zones.
Why do leaderboards skip ranking numbers when there are ties?
Performance Management uses standard competition ranking, often referred to as “1224” ranking. This widely-used style of ranking allows lower-ranked performers to understand their placement without being affected by the ties of performers ahead of them.
Can I export leaderboards?
Not at this time. Results data will be made available on the Talkdesk data platform as a near-term roadmap item.
Where can I see a trend view of the leaderboard results?
Trend views are not available today, but this popular data visualization is on the roadmap. For a quick seven-day trend, the coach option can be chosen to view these results; saving the coaching is not required.
Is there a way to anonymize or hide names on leaderboards?
Yes. There is a specific permission that will anonymize all results below average (other than the user’s own results, which they can always see). To adjust this, select the Admin icon > Click on the Roles and Permissions tab > Choose the role > Click on the “Permissions” tab > Press the Edit button > Scroll down to the “Performance Management” section > Select the Anonymized option, under “Results”, “Leaderboards”.
Note: The default “Agent” role and all custom roles are given “Anonymized” access only to safeguard data. An administrator may increase the visibility of results for any custom roles to “Full view,” if desired.
What do the color circles on the leaderboards signify?
The color circles have no specific meaning in the leaderboards.
When does a “week” start?
A week in Performance Management starts at 00:00:00 Sunday morning and concludes at 23:59:59 Saturday night. It is not customizable at this time.
Why can’t I select an arbitrary date range of my choice?
A separate leaderboard is calculated for every Measure and Queue combination for single dates, current week, last week, current month, and last month. Calculating every possible combination of user-selected dates, in addition to this, would involve excessive computations, resulting in a poor user experience.
What is a queue? Is this a ring group?
Queue is a generic term that includes Talkdesk ring groups today and will include e-mail, SMS, and text groupings in the future.
Why are there so many queues in the dropdown?
The system remembers every queue for which a result has been created. There is not a way to remove unwanted choices; however, a search is available.
Why is there a “Non-queue” choice in the queues' dropdown?
Non-queue is available for Quality Management evaluations that are not associated with any particular inbound voice ring group; for instance, quality evaluating a face-to-face interaction, or back-office casework.
Why can’t I multi-select a group of queues?
A separate leaderboard is calculated for every Measure and Queue combination for single dates, current week, last week, current month, and last month. Calculating every possible combination of user-selected queues in addition to this would involve excessive computations, resulting in a poor user experience.
Note: Talkdesk expects to offer this ability in a future version of Performance Management.
Why don’t the team names always show on the Leaderboard?
The team names will only show if a “Team” filter is applied. All teams may be chosen purposely to make the team names appear next to the agent names on the Leaderboard.
Is there a limit to the number of teams I can select?
There is a 20 team limit when choosing teams in the filter, however the default view of Leaderboards (with no filter applied) can include results spanning more than 20 teams.
When I apply a team filter, why isn’t the leaderboard renumbered?
When a “Team” filter is applied, the agents not on those teams are removed from the leaderboard. Those remaining retain their original numbered rank for overall reference. The overall goal and average remain as well.
The result count will update and breaks will appear as a visual indicator where results have been removed.
Why is there a “No Team” choice in the Teams filter dropdown?
“No Team” is used to locate results for agents who have not been assigned to any team in Talkdesk. For these agents, a “blank” rather than a “Team Name” will be next to their name on the Leaderboard when team filtering is applied.
What happens to the rankings if an agent leaves?
The performer’s name is replaced with “Deleted User” and the result remains to retain the integrity of the leaderboard including results’ count, rankings, and averages.
How is the average calculated?
The performer results displayed across all pages of the individual leaderboard are averaged. The average is not weighted, meaning it is calculated irrespective of the volume of each result; pagination of results affects the average in no way.
Why do double dashes “--” appear in the “Goal” area?
There is no calculated goal for the leaderboard being viewed. This can be for several reasons; however, the most likely reason is due to inadequate historical results (fewer than a seven-day minimum) needed to establish an automated goal.
Where did the goal come from? Can I change it?
The goal is calculated by the system. Performance Management analyzes up to 31 days of prior results and establishes the goal at the 67th percentile of those results for a challenging, but achievable goal. The goal can be changed today by the support desk accessing an end-point. In the future, the “Configurations” menu will allow an administrator and/or supervisor to adjust the goals, either manually or by recalculating.
Coaching and Development
I don’t see the coaching features. Why are they missing?
Coaching features are controlled by the application-specific Roles and Permissions settings in the Admin area of Talkdesk Classic. Look for the application “Performance Management”. Make sure the permission level desired for “Coaching” is chosen there, in increasing order of privilege: view, update, create, and delete. Note that any user-created custom roles will receive permissions as if they are an Agent, and the administrator will need to expand the permissions on these custom roles if desired by the business users.
Is there a character limit for the notes section? Can I copy and paste a URL?
There is a 10,000-character limit. A URL may be copied and pasted, but it will not be a clickable hotlink.
Is there a way to enter confidential notes on a coaching session?
Today, there is a general notes field that is not specially protected. In the future, notes fields will be considered with varying levels of confidentiality.
Can I search, sort, or filter the development list?
There are no dedicated functions for searching, sorting, or filtering the development list. The list is displayed in reverse chronological order and may be searched with your browser’s search function (most likely Ctrl-F or Command-F). Searching, sorting, and filtering will be offered soon!
Is the date the result date or the date of the coaching session?
The date at the bottom of the coaching panel is the date of the coaching session itself. The result being addressed on the coaching date may be any date from an existing leaderboard. For now, a best practice is to type the result date in the notes field.
Can I create a coaching session that is not associated with any measure? What about coaching associated with multiple measures?
Both of these are capabilities being targeted for future development.
Why is there a column titled “Resource”? Why isn’t it “Coach”?
The development page is designed to hold sessions like mentoring, learning, self-study, and quizzes in the future. So this column has a more general header in anticipation of these future types of development.
I accidentally deleted a session. Is there a way to get it back?
No. There is a confirmation pop-up in the software displayed when “Delete” is pressed. This pop-up clearly states this action cannot be undone.
Where can I view an audit trail of who has created, updated, or deleted a coaching session?
Audit logging is part of the system, and detailed activity may be discovered by contacting the Talkdesk support desk.
Can agents see the coaching sessions?
Yes. Agents may view their own personal coaching sessions, but not those of others. Furthermore, the coaching session views have been optimized for agents using the small screen view inside the Desktop version of Talkdesk. Users with the “Anonymized” leaderboard view permission are considered agents.
Why can’t all results be applauded?
Applauding every result is not appropriate to ensure fairness. Average or above results only may be applauded, and an agent may not applaud their own personal results. Only daily results may be applauded, not weekly or monthly result roll-ups.
What does the applause do? Is it reportable?
Applause causes confetti to drop on the leaderboard where the agent has received applause. The applause count is not reportable today. In fact, 10 or more applause simply show as “9+”.
How is the applause system different for Agents vs. Supervisors?
Users with the “Anonymized” leaderboard view permission are considered agents. Those with a “Full” view are considered supervisors. Supervisor applause counts “5x” because being recognized by management carries extra weight, agent applause counts “1x”.
How is the confetti triggered?
Confetti only drops on one leaderboard per result day (the first leaderboard the agent visits where applause has been received). A toast helps the user understand why they are receiving confetti.
Confetti only appears for users with the “Anonymized” leaderboard view only, not users with “Full” leaderboard view permissions.
Measures and Results
A measure is Talkdesk’s term for what is traditionally referred to as a Key Performance Indicator (KPI).
Can I add new Measures? Where are these configured?
Performance Management does not offer the ability to create or add new measures. In the future, this capability will be offered in the “Configurations” menu.
I always see no results for my selection on some Measures. Why?
If the Measure is powered by a separately licensed Talkdesk module, such as Quality Management or Feedback, then no results will be displayed. For voice metrics, contacts must be answered on inbound ring groups to populate the measure results. See Measure Dictionary for more details.
Can I turn off a Measure?
Measures not desired may be deactivated by an administrator on the “Configurations” menu.
What happens when I turn off a Measure?
The measure will no longer be available in the Leaderboard drop-downs (past dates and going forward), nor can the measure be coached on. Performance Exception Notifications for the measure will be suspended. The measure chosen on existing coaching sessions will remain.
Is there an overall score across all the Measures?
Not today. In the future, attainment will create the ability to calculate and display an overall score across measures while allowing each measure to retain its own unit (duration, percent, rate, among others).
How many decimal places are calculated and shown on the results?
Results are calculated precisely and rounded, when applicable, to two places for both storage and use in the user interface.
How are ties ranked?
Results are calculated precisely and rounded to two places, when applicable, for both storage and use in the user interface. After rounding, any tied results will be assigned the same rank. There are no further sub-sort criteria for the tied performers, so the order within tied groups of performers is random.
What is “Volume”?
“Volume” is a count of the unique source events that contribute to a result and help inform confidence around the result displayed. For a measure like “Inbound Contacts per Online Hour”, “Volume” is the number of contacts handled. For “Average CSAT”, “Volume” is the number of surveys returned. The in-app description describes the “Volume” for each measure.
Why is taking more calls or having a lower Average Handle Time (AHT) always considered “better”?
In its first iteration, handling more calls, or having lower AHT is considered better because this is generally favorable behavior. In the future, target-style goals will be available and Measures may be configured to indicate if “lower values are better” or if “higher values are better.” For example, on sales queue, longer AHT may be favorable for forging stronger customer relationships, generating greater revenue that offsets the increased personnel time cost.
Why is the Measure “Inbound Contacts per Online Hour” not an easy-to-understand count of inbound calls?
Simply counting inbound calls does not scale well. For example, it would be inappropriate to compare part-time with full-time employees, or employees who had taken a one-week vacation with those who hadn't. Measuring the rate at which calls are handled establishes an instant apples-to-apples comparison.
What is an Online Hour? Can I change it to exclude certain “away” and “not ready” statuses?
Online Hour is a new concept for “Performance Management” and cannot be found in the standard Talkdesk Data Dictionary. Online Hour is the logged-in time for an agent, minus the time spent in the Offline status. So, all green, yellow, and red statuses are considered “Online,” but the gray status is not.
Note: Online hours can be expressed in hours, minutes, and/or seconds. It is not limited to, nor rounded to, whole hours.
Will the results update if data is changed in the past, for example, if I delete Quality Evaluations, or make days-old callbacks?
Flagging changed data and recalculating prior leaderboards is not a function of the software today, but this is on the roadmap.
Are AI-scored Evaluations included in the Quality Evaluation Score? What about evaluations not associated with an inbound, ring group-routed voice call?
Yes and yes. “Performance Management” makes no distinction between human and “AI Scored” evaluations. Evaluations not associated with an inbound, ring group-routed voice call may be viewed by choosing the “Non-queue” in the queue filter.
How long is data retained in the Performance Management application?
The data is retained for 14 months in the Performance Management application. This includes leaderboard results, coaching sessions, and performance exceptions. The raw data needed to calculate leaderboards are stored for 3 months in anticipation of adding the ability to update results as the result of a data change (e.g., changing the score on a Quality Evaluation completed in the past).