Agent Insights

Overview

Agent Insights provides a way to create highly customized views into your agent performance. Metrics can be aggregated by individual human agent, or groups of agents, for a specified period of time.

Admins can create and save different Agent Views that highlight different groups of agents, and/or different metrics, that can then be shared with Admins, Staff Admins & Managers within your organization.

📘
For help creating common views, reference the Agent Insights Playbooks section.

Views

A View is a version of the Agent Insights report that can be modified and saved to show a specific configuration. The first thing users will see when clicking into Agent Insights will be all the existing Views they have access to in their Quiq site:

Administrators can create Views in order to:

Only include certain agents
Select additional metrics to display, including survey scores
Hide any metrics that are not applicable to the audience
Customize how the data is formatted, like the time units for duration columns, the decimal precision for number columns, or rather to show a percentage as a progress bar.
Allow a View to be available to specific Manager users by Role

Opening a View

Navigating to the Agents tab within the Insights tab in Reports will show a list of all the Views the logged in user can access.

Admins will see all the Views that have been created. Staff Admins and Managers will only see Views they have been granted access to by Role:

Administrators can Copy or Delete a View from the list by choosing the buttons on the right. Only Administrators can create and edit Views.

In the beginning, Administrators will see messaging that no Views have been created. Simply click the get started link to create your first View. Staff Admins and Managers will not see the Agents tab until they have been granted access to at least one View.

Create a New View

The +New View button will create a new View that will have the base configuration by default. From here, an Admin can modify and save the View as desired.

📘
Learn how to get started creating some common Agent Insight Views in Agent Insights Playbooks.

General Settings

Select the pencil icon to access the General Settings:

Within General Settings an Admin can manage the following:

1. Label: Change the label name of the View

2. Available for: Grant viewer permission to sets of Users.

Admin users will always see all Views.
Staff Admin users and Managers will only see the View if it is marked as Available for Admin and Staff Admin or Admin, Staff Admin, and Manager.
If the Restrict Viewership by Role toggle is disabled, all users for the selected permission set will see the View. If the toggle is enabled, then a non-Admin user must be a member of one of the selected routing roles in order to have access to the View.

3. Delete View: Delete the View. Once a View has been deleted, it is no longer accessible to any user.

Edit Columns

Select the Edit Columns button to customize which metrics are displayed as columns in the table, and how they are displayed:

The left side of the dialog will show all of the available metrics by category:

An Admin can check and uncheck available metrics to display or hide them as desired.

All the metrics for a section can be selected (or unselected) by checking the box in the the Category heading.

When a column is selected on the left, the right side shows the display options for that column:

The types of display options differ by the type of field. Number and Percent fields allow visualization to be set to a number or a bar, and the decimal precision can be changed. Duration fields can have the displayed time units changed.

Reset Display Preferences will change all column visibility and display options back to that of the default base view.

📘
Each View can have a maximum 50 visible metrics columns.

Agent Segments

Agent Segmentation can be applied to the Agents table much the same way as other Segments work in other Insights sections. This allows the data to be filtered to a subset of the agents, and also allows the data to be aggregated by a group of agents rather than each agent individually.

Select the Agent Segments button to apply or configure Agent Segments:

A Create New Agent Segment button is available alongside any existing Agent Segments:

Creating a new Agent Segment

Select the Create New Agent Segment button to open the segment editor:

Each segment can be defined to select a group of users by:

Constraints defining a constraint, or group of constraints, to select a group of users by filtering on their e-mail address, username, or the role to which they are a member
Always include allows for the individual selection of one or more users to always be included in the segment
Always excludeallows for the selection of individual users to remove from the segment when they are included with the defined constraints.

Once a segment contains at least one constraint, or at least one user has been identified to always include, and a name has been given, the segment can be saved and can be applied to the Agents table:

Applying Agent Segments

An Agent Segment will not be automatically applied to the data after it is saved. To apply an Agent Segment, select the checkbox to the left of the name in the selection list:

Unlike other Segments, multiple Agent Segments can be applied at once. This will have the overall effect of returning the agents that match any of the selected segments, and each selected segment will be listed as a separate grouping in the returned results:

Any segment can be expanded to view the data for the individual agents who belong to that segment:

Resolving Agent Segment Conflicts

Sometimes when multiple segments are selected, there will be conflicts if the same agent matches more than one of the defined groups. If at least one agent matches one or more of the selected segments, the Agent Segmentation Conflicts dialog is automatically opened:

Every agent who matches more than one segment will be listed. By default, each agent is listed with the Unassigned segment.

Any agent can be left associated with the Unassigned segment, and the results will show an Unassigned grouping along with those for the selected segments. Each agent will only be listed with one selected segment.

Any segment can be expanded to view the data for the individual agents who belong to that segment:

Save the View

When changes have been made by editing the View, or by applying segmentation, Save buttons become available in the upper right.

Save will apply the changes to the currently selected View.

Save Copy allows you to create a new View with the current changes saved, leaving the current View unchanged.

Cancel will discard any changes.

Reviewing Conversations for a Metric

To help analyze and review the data, any column that is a count of some conversation event supports drill-down. The columns are marked with a message bubble indicating that drill-down is available:

Selecting the bubble will open a drill-down pane that shows the conversations that were counted for that agent for the selected metric.

Each conversation can be selected to show the transcript details that occurred around the time of the event.

For example, if it was observed that some agent had a higher than average Missed Response Rate, the manager could review the conversation transcripts to gain insight into what was causing the agent not to respond in time:

Exporting Data

You can export the data into a CSV formatted file by selecting Export Data > Agent Performance Data from the top toolbar. Only the columns available in the view are included in the CSV file. Display options are ignored and data will be returned in raw format.

Agent Insights Metrics

Below is a list of some of the metrics available when creating and analyzing agent performance in agent insights.

Productivity

A list of available metrics related to agent productivity.

Metric Name	Definition
Availability	The total time the agent was in the Available status. This includes any time spent in any custom Available sub-statuses. When the current date/time is used in the selected date interval, the value returned will be near real-time.
Time Online	The total time the agent was logged into the Quiq application, regardless of status. When the current date/time is used in the selected date interval, the value returned will be near real-time.
Time Offline	The total time the agent was not logged into the Quiq application during the selected date interval. When the current date/time is used in the selected date interval, the value returned will be near real-time.
Time Current Conversations	The total time the agent was in the Current Conversation status. This includes any time spent in any custom Current Conversation sub-statuses. When the current date/time is used in the selected date interval, the value returned will be near real-time.
Time Unavailable	The total time the agent was in the Unavailable status. This includes any time spent in any custom Unavailable sub-statuses. When the current date/time is used in the selected date interval, the value returned will be near real-time.
Received	The number of conversations received. Agents may receive the same conversation more than once. Reactivations are not considered received unless the owner changes. A received conversation is counted when the agent declines an invitation or missed an invitation time such that the conversation is re-queued.
Received / Available Hour	The average number of conversations received per hour of availability. This is calculated with Received/Availability (in hours). This will always be 0 if there were 0 Received, and blank if there is no time Available.
Completed	The number of conversations completed. A conversation’s completed time may be when an agent ends the conversation, or when it goes inactive. NOTE: This metric will be delayed when viewing recent data because current conversations are not yet completed. If a conversation was Ended from an Active status by an agent, the completed value will be counted with the closed time. If the conversation was Inactive when it ended, either manually or from a timer, the completed value will be counted with the time it last went Inactive. This means the metric can change when viewing data for a date interval in the recent past since inactive conversations may have since closed. A conversation that is transferred or re-queued is not considered completed by the agent. Conversations Marked as Spam are counted as Completed.
Completion Rate	The percent of conversations completed per conversation received. NOTE: This rate will be artificially low when viewing recent data because current conversations are not yet completed. This is calculated by Completed/Received. This will always be 0% if there are 0 Completed but more than 0 Received, and blank if there are 0 Received.
Completions / Hour	The average number of conversations completed per hour of availability. NOTE: This rate will be artificially low when viewing recent data because current conversations are not yet completed. This is calculated by Completed/Availability (in hours). This will always be 0 if there were 0 Completed, and blank if there is no time Available.
Utilization	The degree to which agents’ routing slots were filled with conversations while they were available. May exceed 100% if agent is consistently above their soft routing limit. If there are active assigned conversations while the agent is in Current Conversations or Unavailable, the utilization is not affected. Being Available and having no assigned Active conversations would count as 0% utilization. While being not-Available isn’t counted toward utilization - or will show as a blank utilization. Conversation weights are taken into account so some conversations may count more than others. Utilization is near real-time. That is, Active conversations (where the assignment hasn’t ended) are taken into account when using an interval that includes the current time.
Avg. Workload	The average weight of the assigned Active conversations when the agent is Available. Any conversations being worked while the agent is not Available, do not count toward workload. Conversation weights are taken into account so some conversations may count more than others, and soft limits are not relevant to the metric. When the agent is Available and has no conversations, he has a workload of 0. This is basically the same as concurrency - only with conversation weight applied.
Obligated	The percentage of time the agent was Available and had at least one conversation requiring a response. Calculated with the Obligated Total/Availability. The Obligated and Waiting percentages will always add to 100%.
Obligated Total	Total duration of time when the agent was Available and had at least one conversation requiring a response. If the agent has no active assigned conversations that required a response - that is, have the Response Timer running - then he is Waiting, not Obligated.
Waiting	The percent of time the agent was available but did not have at least one conversation requiring a response. Calculated with the Waiting Total/Availability. The Obligated and Waiting percentages will always add to 100%.
Waiting Total	The total duration of time when agents did not have at least one conversation requiring a response. If the agent isn't Available, he isn't Waiting. If the agent is Available, and has no assigned conversations, he is always Waiting. If the agent is Available and has one or more assigned conversations and there is response obligation for at least one of them, then he is not Waiting. If the agent is Available and has one or more assigned conversations, but there is no response obligation for any of them, then he is Waiting.
Total Work Time	The total work time logged by this agent. The accumulated work time will not be counted in the report until the assignment interval is completed. That is, the completed work time intervals for an active conversation still assigned to the agent, aren’t counted when the report interval includes the current time. Work time for both completed and non-completed conversations is counted.
Work Time / Completion	The average work time per conversation. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated from the Total Work Time/Completed
Work Time / Received	The average work time per conversation received. This is calculated from the Total Work Time/Received.
Adj. Handle Time Total	The total concurrency-adjusted handle time accrued by agents. This measure of handle time has built in account for concurrency. The handle time will not be counted in the report until the assignment interval is completed. That is, the handle time accumulated for an active conversation still assigned to the agent, isn’t counted when the report interval includes the current time. All the handle time done during the assignment interval will be counted when the assignment interval ends (conversation counted as Received), even if some of it exists outside the report interval. Handle time for both completed and non-completed conversations is counted.
Adj. Handle Time / Completion	The average concurrency-adjusted handle time per conversation completed. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated from the Adj. Handle Time Total/Completed. The handle time done for non-completed conversations is counted towards this value.
Adj. Handle Time / Received	The average concurrency-adjusted handle time per conversation received. This is calculated from the Adj. Handle Time Total/Received
Basic Handle Time Total	The total duration of routing slot occupancy for agents. The handle time will not be counted in the report until the assignment interval is completed. That is, the handle time accumulated for an active conversation still assigned to the agent, isn’t counted when the report interval includes the current time. Handle time for both completed and non-completed conversations is counted.
Basic Handle Time / Completion	The average basic handle time per conversation completed. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated from the Basic Handle Time Total/Completed. The handle time done for non-completed conversations is counted towards this value.
Basic Handle Time / Received	The average basic handle time per conversation received. This is calculated from the Basic Handle Time Total/Received.
Avg Initial Response per Assignment	The average time it took for the agent to send a response to the customer after assignment. If the conversation is assigned to multiple agents (for example, in transfers or queue changes), then the first response for each assignment will be included in the calculation. Assignment begins when the conversation is offered to the agent, when using invitations. The time spent waiting in queue is not included.

Quality

A list of available metrics related to quality.

Metric Name	Definition
Reopen Rate	The rate at which completed conversations reopen and escalate to live agents within 72 hours. This is calculated as the Reopens/Completed.
Reopens	The number of completed conversations that were reopened and escalated to live agents within 72 hours. The re-open is credited to the agent who is credited with closing the original conversation, no matter which agent handled the re-open. The re-open is counted at the time the original conversation was completed - but not until the reopen occurs. Thus, this metric may change over time for a report window in the past 72 hours.
Missed Response Rate	The percent of response timer misses per conversation received. This is calculated as the Missed Response/Received.
Missed Responses	The number of times the agent missed the response timer and the conversation was taken away.
Invitations	The number of invitations offered to the agent - or the number of conversations routed to the agent from the queue. When invitations are disabled such that conversations are directly assigned to agents without invitation, an Invitation is still counted.
Missed Invitation Rate	The percent of invitation misses or rejections per invitation offered. This is calculated as the Missed Invitations/Invitations.
Missed Invitations	The number of invitation misses or rejections. This also includes the number of times the agent manually rejected the invitation using the Decline button, but does not include Transfer request Declines.
Accepted Invitation Rate	The percent of invitations accepted per invitation offered. This is calculated as the Accepted Invitations/Invitations.
Accepted Invitations	The number of invitations accepted by the agent. This does not include Transfer requests that were accepted. When invitations are disabled such that conversations are directly assigned to agents without invitation, an Accepted Invitation is not counted.

Feedback

A list of available metrics related to Surveys & Feedback.

Metric Name	Definition
Avg. Score surveyLabel questionLabel	The average score for the survey question surveyLabel questionLabel. Only survey questions that are of type Score choice or Number rating will be available. The survey score is attributed to the last agent assigned the conversation prior to survey invitation.
Count surveyLabel questionLabel	The count of responses provided for the survey question surveyLabel questionLabel. Only survey questions that are of type Score choice or Number rating will be available.
Total surveyLabel questionLabel	The total score for the survey question surveyLabel questionLabel. Only survey questions that are of type Score choice or Number rating will be available.

Transfers & Requeues

A list of available metrics related to transfers and requeues.

Metric Name	Definition
Transfer Rate	The percent of Transfers per conversation received. This is calculated with the Transfers/Received.
Transfers	The number of times agents transferred conversations to another agent. Declined Transfer requests are ignored. Manager take-overs are not counted.
Transfers Received Rate	The percentage of received conversations that began as Transfer from another agent. This is calculated with the Transfers Received/Received
Transfers Received	The number of times agents accepted Transfers from other agents. Declined Transfer requests are ignored. Manager take-overs are not counted.
Queue Return Rate	The percent of re-queued conversations to the same queue per conversation received. This is calculated with the Queue Returns/Received.
Queue Returns	The number of times agents re-queued conversations to the same queue. Manual queue changes to a different queue are not counted and changing queue, remaining in conversation are not counted. Re-queues due to Missed Response Timers or Missed Invitations are not counted.
Queue Change Rate	The percent of agent queue reassignments per conversation received. Does not include cases where the agent changes the queue but remains the owner. This is calculated with the Queue Changes/Received.
Queue Changes	The number of times agents reassign queues. Does not include cases where the agent changes the queue but remains the owner. Manual queue changes to the same queue and changing queue, remaining in the conversation, are not counted.
Manager Takeover Rate	The percentage of received conversations that were taken over by a manager.
Manager Takeovers	The number of times a manager did a take-over of a conversation that was assigned to the agent.

Asynch Handling

A list of available metrics related to handling asynchronous conversations.

Metric Name	Definition
Inactive Workload	The average workload of inactive conversations assigned to agents while they’re available. The time spent Offline, Unavailable or in Current Conversations is not counted - only the Available time periods are counted. The weights of the conversations are taken into account so the value could be higher than the number of conversations that were shown to the agent as Inactive. If an agent was Available, but had no Inactive conversation the entire period, a value of 0 is shown. If an agent wasn’t Active, then no value is shown.
Manual Completion Rate	The percent of conversation completions in which agents synchronously closed the conversation via the End button. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated with Manual Completions/Completed.
Manual Completions	The number of conversation completions in which agents synchronously closed the conversation via the End button. Conversations that were closed by a timer are not counted.
Manual Inactivation Rate	The rate at which agents manually inactivate conversations per conversation received. This is calculated with Manual Inactivations/Received.
Manual Inactivations	The number of times the agent manually inactivated a conversation by selecting the Inactive button. Conversations that became Inactive based on a system timer are not counted. Each time the agent selects the ‘Mark Inactive’ button is counted so the same conversation/assignment could be counted multiple times if the ‘Mark Inactive’ button is selected again after the conversation is reactivated.
Manual Reactivation Rate	The rate at which agents manually reactivate conversations by sending a message per conversation received. This is calculated with Manual Reactivations/Received.
Manual Reactivations	The number of times agents manually reactivate conversations by sending a message. This is counted when an agent sends a response on an Inactive conversation, which re-activates it. A customer message which re-activated a conversation is not counted. If an agent does some other action on the conversation that re-actives it (like transferring), it will not be counted.
Unnecessary Reactivation Rate	The rate at which agents sent a message the system determined was unnecessary to inactive conversations, thereby reactivating them, per conversation received. Calculated with Unnecessary Reactivations/Received.
Unnecessary Reactivations	The total number of times agents sent a message the system determined was unnecessary to Inactive conversations, thereby reactivating them. This will only be counted when the agent sends a message to an Inactive conversation where there was no obligation to do so. Often, this is something like “Are you still there?” or “If you still need help…”. If the agent sends such a message (no obligation to send one) while the conversation is Active, it will not be counted. If the agent sends a message that was required that reactivates a conversation (Inactive, but still have obligation to respond), then it is not counted.

Message Authoring

A list of available metrics related to how Agents author messages.

Metric Name	Definition
Messages / Completion	The average messages sent by the agent per conversation completed. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated with Messages Total/Completion.
Messages / Received	The average messages sent by the agent per conversation received. This is calculated with Messages Total/Received.
Messages Total	The total number of messages sent by agents. Only human agent messages are counted - customer, system and bot messages are not counted.
Typing / Completion	The average amount of time spent typing per completion. NOTE: This rate will be artificially high when viewing recent data because current conversations are not yet completed. This is calculated with Total Typing Time/Completed.
Typing / Received	The average amount of time spent typing per conversation received. This is calculated with Total Typing Time/Received.
Typing / Work	The percent of work time that was spent typing. This is calculated with Total Typing Time/Total Work Time.
Total Typing Time	The total time spent typing message text. This is the time spent actively entering characters from the keyboard. It does not count time pasting in characters from clipboard or using Snippets, nor does it include time sitting with focus in the Message text box when not typing.
Keyboard	The percent of agent message characters manually typed from the keyboard. This is calculated with the total number of characters typed from the keyboard/total number of characters entered in the message. If any typed characters are removed before sending, they are still counted.
Clipboard	The percent of agent message characters pasted from the clipboard. This is calculated with total number of characters pasted in from the clipboard/total number of characters entered in the message. If you remove any of those characters before sending, they are still counted.
Snippets	The percent of agent message characters inserted from Snippets. This is calculated with the total number of characters entered from a Snippet/total number of characters entered in the message. If you remove any of those characters before sending, they are still counted as entered.
Compose	The percent of agent message characters accepted from Compose suggestions. This is calculated with the total number of characters added by accepting Compose suggestions/total number of characters entered in the message. If you remove any of those characters before sending, they are still counted as entered
Other	The percent of agent message characters injected from other sources like right-hand-side extensions. This is calculated with the total number of characters added through RHS extensions/total number of characters entered in the message. If you remove any of those characters before sending, they are still counted as entered.