Browse through the most frequently asked questions on Talkdesk Explore. For additional in-depth guidance, we recommend exploring the step-by-step articles on Talkdesk Explore and completing the Reporting & Analytics certification in Talkdesk Academy.
- Data and Permissions
- Dashboards and Reports
- How do I create a report in Explore?
- Can I create Explore reports with more than one dataset?
- How many rows can be displayed on a report accessed in the Explore application?
- Is there a limit to the number of rows in a report scheduled from Explore?
- What is the start day of the week for Explore reports?
- Can I export a spreadsheet with calls, notes, and dispositions?
- Can I create a custom dashboard?
- How many widgets can be used on a single Explore dashboard?
- How fast can Explore run reports and dashboards?
- What is the minimum refresh rate for Explore dashboards?
- How do I replicate the Metrics shown in Talkdesk Home Legacy in Talkdesk Explore?
- Which dataset should you use to get more insights into an agent’s activity?
- Why does an agent's time in the "On a Call" status doesn’t match the agent's handle time in “Contacts”?
- Why are there some differences between the time metrics from a Calls report and the time metrics from a Contacts report?
- What is the difference between Ring Time, in the Contacts Report, and Speed to Answer, in the Calls Report?
- Why do callbacks count at the Service Level?
- Why are callbacks showing up as abandons in reporting?
- What’s the difference between the Inside Service Level and the Within Service Level metrics?
- Do pre-queue calls generate a contact for the contacts data model?
- Why is the Ring Group null (blank) on my contacts report/dashboard?
- Are short abandoned contacts included in the standard abandoned contacts and abandon rate metrics?
- Why does the agent field show "If-No-Answer Agent" in a report?
- Why does the agent field show "External Phone Number" in a report?
- Why do deactivated/deleted agents still show up in report filters?
- Why doesn't the Inbound Contacts report have a filter or a scope for Team?
- If a Studio flow call is routed to an external number, does it create a new contact?
- How are outbound tags for Classic Routing calculated?
- How can I know how a call was routed using Explore reports?
- What are possible hangup causes?
- What are the differences between Waiting Time (WT), Average Speed of Answer (ASA,) and Average Agent Speed of Answer (AASA)?
Data and Permissions
How can I control user access to Explore?
To manage the users who have access to Explore, go to My Apps > Explore > Users. To configure the permissions for the different areas in Explore, including the “Create” feature (to create custom reports and dashboards), go to Roles and Permissions
What is Explore’s data retention period?
By default, there’s no data retention period in Explore (both UI and API). Although there are no data limitations, it is only possible to extract a maximum of 13 months in each extraction. This applies to reports extracted from both the Explore app and the Explore API. Please note that Studio Flow Execution Report data is only available from August 25th, 2019 onward. You can define a shorter “Reporting Data Retention Period”, in Admin > Preferences > Metrics Settings.
What happens to data when a phone number is deleted from my account?
Data from numbers that have been deleted from an account still exist for Explore reporting.
Why can’t I see data before a specific date in reports?
If the data is within the last 13 months, please review the data retention settings on your account. The data may no longer be available based on the defined retention period.
Dashboards and Reports
How do I create a report in Explore?
Learn here how to create a report. Please note that the “Create” feature is not available by default.
Can I create Explore reports with more than one dataset?
At present, multiple datasets within the same report for custom report creation are not supported. However, Explore dashboards allow you to utilize more than one dataset.
How many rows can be displayed on a report accessed in the Explore application?
By default, Explore returns the first 5,000 rows of a report. This is a system definition designed to prevent poor performance resulting from a larger dataset being returned automatically. This limit cannot be modified.
Is there a limit to the number of rows in a report scheduled from Explore?
There is no row limit for scheduled reports because they are provided in CSV format. Please note that there is a row limit in Microsoft Excel.
What is the start day of the week for Explore reports?
The week starts on Sunday in Explore reports.
Can I export a spreadsheet with calls, notes, and dispositions?
Yes. You can download or send a Calls Report. Learn here how to do this. When sending, we recommend selecting “All Results” in “Advanced Options”. The recipient will receive an email with a link to download the report.
Can I create a custom dashboard?
Yes - learn more here. Please note that the “Create” feature is not available by default.
How many widgets can be used on a single Explore dashboard?
Each Explore dashboard can have up to 16 widgets.
How fast can Explore run reports and dashboards?
The speed at which the reports and dashboards return data is dependent on a few factors, including the timespan and volume of data that you are requesting.
You should also keep in mind that software and applications running on your computer can negatively impact the speed of Explore. Applications such as video chat can take considerable processing power and memory.
What is the minimum refresh rate for Explore dashboards?
Data in Explore dashboards can be refreshed at a minimum of every 5 minutes.
How do I replicate the Metrics shown in Talkdesk Home Legacy in Talkdesk Explore?
Go here to learn how to do this.
Which dataset should you use to get more insights into an agent’s activity?
That depends on what you are trying to accomplish. If you are trying to assess an agent’s availability, idle time, unavailable time, or utilization, you should use the Agents Activity Analysis or User Status datasets. These datasets have different granularity but they both refer to the agent and their activity. If you are trying to assess an agent’s handle time, talk time, or ACW performance, you should use the time metrics in the Contacts dataset. Learn more about Datasets.
Why does an agent's time in the "On a Call" status doesn’t match the agent's handle time in “Contacts”?
These two metrics read from different systems. The status information is fed from the presence system. This data is used for agent status because if it were based on the Contacts system, you would not have insight into how an agent spent time while they were not handling contacts. The Handle Time for Contacts reads from the contacts system. Because two different systems with different events are used, you may see some minor variances between the two datasets. Also, the Handle Time includes not only the “On a Call” status duration but also the “ACW” status duration.
Why are there some differences between the time metrics from a Calls report and the time metrics from a Contacts report?
There are a couple of reasons this may occur:
- Some metrics in the Contacts data model were refactored. Examples of these include Talk Time and Abandon Time.
In the Calls data model, Talk Time includes both time spent talking to the caller and time spent on hold. For the Contacts data model, these two time values are reported separately as Talk Time and Hold Time.
In the Calls data model, abandon time includes both time in IVR and time in queue. In the Contacts data model, it only counts the time spent in the queue before the person was abandoned.
- The Contacts and Calls Analysis datasets are based on different back-end systems and, because of this, sometimes they are not triggered by the same events (specific events exist in one system but do not exist in the other).
- The Calls measurements assign all-time values to a single agent, even if more than one agent handled the call. The Contacts data apportions the correct time to each agent involved in a call.
What is the difference between Ring Time, in the Contacts Report, and Speed to Answer, in the Calls Report?
The Ring Time and Speed to Answer metrics measure the same thing — the time between when the call began ringing and when it was answered. However, these metrics originate from two different back-end systems: Ring Time comes from the Contacts data model and Speed to Answer comes from the Calls data model. This means that each system may be listening for different events, resulting in minor variances in the metrics.
Note: None of these metrics will aggregate the total ring time if the same call rings more than one agent at the same time. The metrics represent only the linear time from ‘ringing’ to ‘answering’.
Why do callbacks count at the Service Level?
Like other inbound contacts, the wait time for callbacks from the queue is measured between entering the queue and being connected with the agent. The logic behind this is simple: If the customer didn’t choose the callback option and instead waited on the phone for an agent, this time would be counted as wait time toward the Service Level (SL%) calculation. The callback is simply a proxy for the caller’s physically waiting on the phone.
Why are callbacks showing up as abandons in reporting?
When you place a Reporting component before a Callback component, in Talkdesk Studio, you are telling the system that you want a new contact reported for a new queue. Since the existing contact must be ended to start the new contact, it needs an end reason. Given that the first contact was in the queue and the customer then physically hung up while in the queue, after requesting the callback, the system reads this as an abandoned call. To remedy this, remove the Reporting component before the Callback component.
What’s the difference between the Inside Service Level and the Within Service Level metrics?
Inside service level: Indicates whether or not a contact was within the wait time threshold set in the Admin settings. The Inside Service Level applies to answered contacts only. This is the metric used to calculate the service level metric for the contacts data model.
Within service level: Applies to all contacts, rather than just answered contacts. This dimension exists to enable you to create a different service level metric than the standard definition if desired.
Do pre-queue calls generate a contact for the contacts data model?
Because the call has not yet reached a queue, a contact is not created. This means pre-queue activity (IVR) is not available in the contacts data model. However, this information can be obtained using Calls data.
Why is the Ring Group null (blank) on my contacts report/dashboard?
The ring group field in the contacts data will be blank in the following situations:
- The contact was for a direct line (ring groups are not used for direct lines).
- An agent transferred directly to another agent, meaning that there wasn’t a queue time.
- You are using attribute-based routing, and you have not included a Reporting component in the flow to identify which queue (ring group) to report for the contacts.
Are short abandoned contacts included in the standard abandoned contacts and abandon rate metrics?
By default, the abandoned contacts and abandon rate metrics exclude short abandons.
Why are there calls in the queue while agents are showing as available?
The ring group (sometimes referred to as the “contact ring group” or “queue”) is the ring group under which contacts are reported. By default, this ring group is the ring group from the first Assignment & Dial (A&D) component in the branch of the flow. However, if a Reporting component is used in the flow, it will override the default designation with the value you set in the Reporting component.
As a contact moves through a flow, it may go through several A&D components seeking eligible agents to whom the contact can be routed. If the flow admin did not add the ring group(s) from the first A&D component to all subsequent A&D components in the branch, the routing engine ceases trying to assign the contact to the group of agents from the first A&D component.
As a result, a contact may have moved past the first A&D component to attempt assignment to agents with other ring group assignments, leaving the agents from the first A&D component idle while contacts are showing in the queue.
A call enters the flow and reaches the first A&D component, which is looking for agents with the “Billing” ring group assignment. No Billing agents are currently available, so the call continues in the flow. This call will show as a contact in the Billing queue because that’s where it arrived -the first A&D component (or the ring group specified in the Reporting component).
Eventually, the call reaches an A&D component looking for agents with the “Support” ring group assignment as an overflow step. However, the flow admin did not include the Billing ring group in this component, so the flow is no longer looking for Billing agents.
In the meantime, the Billing agents have finished the calls they were handling. The call that is waiting in the Billing queue won’t be routed to them because the current A&D component is looking for agents with the “Support” ring group assignment; it’s not looking for agents with the “Billing” ring group assignment.
The result is a call waiting in the Billing queue while Billing agents sit idle.
Why does the agent field show "If-No-Answer Agent" in a report?
When a call is forwarded to a Talkdesk number or external number via the Studio Forward to External component, the agent name will be populated with “if-no-answer”.
When the “if-no-answer setting” is enabled and the call is routed to an external number because no agents are available, the agent name will be populated with “if-no-answer”.
Why does the agent field show "Forwarding Phone Number" in a report?
If the agent has the External Phone Number setting populated in their profile, the system will ring that agent on the designated phone number. This will populate “Forwarding Phone Number” in the agent field on reports.
Why does the agent field show "External Phone Number" in a report?
The Agent field on the Calls report will display “External Phone Number” if the call is transferred to an external phone number. If a transfer is made to another Talkdesk number, it is still considered an external transfer.
Why do deactivated/deleted agents still show up in report filters?
When any record is deleted in Admin (e.g., users, ring groups, numbers) the data is retained for historical reporting purposes. This means that records that have previously been deleted will continue to be visible in reports and filters.
Why doesn't the Inbound Contacts report have a filter or a scope for Team?
The Inbound Contacts widget is meant to show all inbound contacts for a ring group (queue), regardless of any agent involvement and how the contact ended (answered, abandoned, short abandoned, voicemail, or missed).
When a team filter is applied to the widget, it will be filtered by the agents assigned to the team. Since agents are only present in contact records if those contacts were answered, filtering by team would remove visibility of all the other inbound contacts where no agent was involved, i.e., arrived contacts, abandoned contacts, and missed contacts, as well as their related metrics such as abandon rate and service level %.
If a Studio flow call is routed to an external number, does it create a new contact?
How are outbound tags for Classic Routing calculated?
Outbound tags (ring groups) for classic routing are calculated as follows:
- Intersection of IVR Tags with Agent ring groups for Legacy IVR Ring groups.
- If the previous is not considered, the intersection between phone ring groups and Agent ring group is applied.
- If the second is not considered, Phone Number ring groups will be taken into account.
How can I know how a call was routed using Explore reports?
1. Identify the call and the respective Interaction ID. Learn how here.
2. Create a Studio Flow Execution Report to understand the call path in the Studio Flow:
- Filter the report by the interaction ID.
- Check the "Exit" row for any “Assignment and Dial” or “Callback” components.
- Look for entries whose Exit has “no_match” or “no_answer”.
3. Identify which agents were available at a certain time:
- Pull an "Agent Status Report" for the same time frame as the call you are analyzing.
- Check if any agent with the same ring group as the "Assignment and Dial" or "Callback" components was available.
- Considerations about outbound dialer reservations and available status may contain reserved ringing agents, i.e., agents assigned to a campaign on an outbound dialer that may show as "Available but be reserved by the campaign for an outbound call and unavailable to receive this call.
- If the agent was on an available status but has received a different call with a shorter waiting time, review the priorities inside the "Assignment and Dial" and "Callback" components of your Studio Flows.
4. Find the result for each ring attempt:
- If the agent was on on an available status, run a "Ring Attempts Report" and check if the call was offered to that particular agent.
- Results can range from “Device unavailable” to “Rejected”. You can find all those definitions in the Data Dictionary.
What are possible hangup causes?
When a call ends in Talkdesk it can be classified according to its hangup cause: the reason why the call was finished.
The following are the possible values for this field:
- IVR: the caller hung up while in the IVR.
- Waiting: the caller hung up while in the waiting queue.
- Hold: the caller hung up during hold.
- Transfer: the caller hung up while being transferred.
- Disconnected: all other hangup causes.
- Abandoned Disconnected:
- Example 1 - On a classic routing scenario (legacy), when the call is abandoned while on the initial greeting (and there is no IVR configured and only a greeting is associated with the number), the hangup cause is set as “disconnected”. In Studio, a call is considered “abandoned disconnected” when the call is abandoned before the IVR.
- Example 2 - There are other exceptions where a call can be classified as “abandoned disconnected”, for example: unsuccessful Intelligent reconnect (see the article Intelligent Reconnect); a call is abandoned while on the voicemail prompt.
- Abandoned Disconnected:
What are the differences between Waiting Time (WT), Average Speed of Answer (ASA,) and Average Agent Speed of Answer (AASA)?
Waiting time (WT): The time a call remains in the queue and ringing until an agent answers it. If the customer is sent to the waiting queue multiple times as a result of no answers or transfers, the waiting time for the call will be the total time spent in the waiting queue across all instances. Time spent in IVR / Voice Prompts is not considered. It also includes time spent waiting for a callback, if the callback feature is enabled in your IVR. Talkdesk provides Waiting Time, Maximum Waiting time, and Average Waiting time metrics.
Average Waiting Time (AWT): The average time a call remains in the queue until an agent answers it. Also known as Average Speed of Answer (ASA) or “Average Delay”. This metric can be calculated for the whole account, per ring group, and phone number. Aggregated Waiting Time data, such as total or average wait times for a ring group, agent, or phone number, will only include waiting time for calls that were handled. They will not include missed or abandoned calls.
Average Agent Speed of Answer (AASA): Calculates the average time it takes an agent to answer the call after it starts ringing. This metric is available per agent but may include the time the call spent ringing to previous agents, in case of sequential dial attempts for the same call. For example, if a call rings Agent A then Agent B, it will capture the ring time for both agents.
For calls that include multiple answers as a result of transfers, the speed-to-answer for the call will be the sum of time it took for each agent to answer their leg of the transfer. As with Waiting Time, Speed-To-Answer is only measured in instances where the call is answered by an agent and does not include Missed or Abandoned Calls, or instances in which an agent did not answer the call when presented.