This article defines the metrics and attributes you can use to create Explore reports based on agent state. These datasets are also used for the agent state prebuilt dashboards. See Analyzing agent state and activity.
This article contains the following topics:
Agent state dataset
The Agent state dataset contains metrics and attributes that relate to when an agent enters a given state and how long they’re in that state. Supervisors can use this information to create timestamp reports that give them an in-depth look at how their agents spent their time.
In this dataset, an agent’s time in a given state is recorded only after their state changes. This means that if an agent is in a single state for the entire duration of the report’s time period, their state is not reported, as there have been no changes to it. To report on agent states with or without changes, use the Agent state daily dataset instead.
This section contains the following topics:
Agent state metrics
This section lists and defines all metrics available in the Agent state dataset.
Metric | Definition | Explore formula |
---|---|---|
Agent time in state / seconds | The time spent in seconds by the agent in a
given state per channel. When you add this metric to a report and slice the results by an agent attribute, you must also add the Group attribute. If you don't, an agent's total time is counted for each group they're in. When you add this metric to a report and slice the results by the State attribute, you must also add the Channel attribute. If you don’t, the time spent in per-channel states is counted multiple times. See Why is the per-channel agent status time different than the unified agent status time? |
Agent state attributes
This section lists and defines all attributes available in the Agent state dataset.
Attribute | Definition |
---|---|
Group ID | The ID number of the group. |
Group name | The name of the assigned group. |
Channel | The channel that an agent state relates to. Possible values include Support, Talk, Messaging, Chat, and Unified. |
State | The state of the agent. Possible values
include:
When you add this attribute to a report, you must also add the Channel attribute. If you don’t, the time spent in per-channel states is counted multiple times. See Why is the per-channel agent status time different than the unified agent status time? |
Agent name | The name of the agent. |
Agent role | The role of the agent. |
Agent ID | The ID of the agent. |
Agent email | The email address of the agent. |
Agent external ID | The external ID of the agent. |
Agent locale | The locale of the agent. |
Agent is moderator | Whether the agent is a moderator. |
Agent status | The status of the agent. Possible values are Active or Suspended. |
Agent active | Whether the agent is active. |
Agent suspended | Whether the agent is suspended. |
Agent organization ID | The organization ID of the agent. |
Agent tags | Tags added to the agent. |
Agent timezone | The time zone of the agent. |
State start time | A collection of attributes in different units of time that record the time at which an agent entered a state. |
State end time | A collection of attributes in different units of time that record the time at which an agent exited a state. |
Agent state daily dataset
The Agent state daily dataset contains metrics and attributes that relate to how groups and agents spend their time across channels. This dataset helps you understand, at a daily level, agents’ total time spent in certain states.
Data in this dataset is aggregated at the end of each day and reported in the viewer's time zone. (See Which time zone does Zendesk use?)
Because the data is aggregated daily, any group membership changes are reflected only in the following day’s data. For example, let’s say an agent is moved from Group A to Group B at noon on January 20. The data for January 20 shows them in Group A (even after they were moved), while the data for January 21 shows them in Group B.
This section contains the following topics:
Agent state daily metrics
This section lists and defines all metrics available in the Agent state daily dataset.
Metric | Definition | Explore formula |
---|---|---|
Agent daily time in state |
The total time per day spent by agents in a given state per channel. This metric is aggregated daily and cannot be broken down into smaller time increments. When you add this metric to a report and slice the results by an agent attribute, you must also add the Group attribute. If you don't, an agent's total time is counted for each group they're in. When you add this metric to a report and slice the results by the State attribute, you must also add the Channel attribute. If you don’t, the time spent in per-channel states is counted multiple times. See Why is the per-channel agent status time different than the unified agent status time? |
Agent state daily attributes
This section lists and defines all attributes available in the Agent state daily dataset.
Attribute | Definition |
---|---|
Group ID | The ID number of the group. |
Group name | The name of the assigned group. |
Channel | The channel that an agent state relates to. Possible values include Support, Talk, Messaging, Chat, and Unified. |
State | The state of the agent. Possible values include:
When you add this attribute to a report, you must also add the Channel attribute. If you don’t, the time spent in per-channel states is counted multiple times. See Why is the per-channel agent status time different than the unified agent status time? |
Agent name | The name of the agent. |
Agent role | The role of the agent. |
Agent ID | The ID of the agent. |
Agent email | The email address of the agent. |
Agent external ID | The external ID of the agent. |
Agent locale | The locale of the agent. |
Agent is moderator | Whether the agent is a moderator. |
Agent status | The status of the agent. Possible values are Active or Suspended. |
Agent active | Whether the agent is active. |
Agent suspended | Whether the agent is suspended. |
Agent organization ID | The organization ID of the agent. |
Agent tags | Tags added to the agent. |
Agent timezone | The time zone of the agent. |
Time - Agent time in state | A collection of attributes in different units of time that return when the agent’s state was recorded. |
Agent productivity dataset
The Agent productivity dataset contains metrics and attributes that relate to work items offered and assigned to agents and how agents used their capacity. Supervisors can use this information to create reports that provide an in-depth look at the productivity of their agents.
This dataset reports on the number of work items newly assigned to an agent during a given time period. It doesn’t report on the total number of existing work items where the agent is listed as the assignee. To report on existing work items instead, use the dataset for the channel you’re interested in. For example, use the Support - Tickets dataset to see how many total tickets are assigned to an agent.
Data in this dataset is aggregated hourly, meaning reports can’t be broken down into smaller time increments.
If an agent has no new work items assigned to them for a given hour, no data is recorded for that timeframe. This means reports will include gaps for hours during which an agent wasn’t assigned any work items. In the example report below, no work items were assigned to the agent during the 15:00 hour, so no data was returned.
Similarly, if an agent wasn’t in the Online state for a given hour, no data is recorded for the Agent average used capacity metric.
Agent productivity metrics
This section lists and defines all metrics available in the Agent productivity dataset.
Metric | Definition |
Agent average used capacity | The average capacity used by an agent for a
channel during a period (while they are in the Online state
only). A null value means the agent was not online. A
0.00 value means the agent was assigned no work
items during that period.
Note: The
metrics omnichannel_agent_productivity_capacity_seconds
and omnichannel_agent_productivity_agent_online_seconds
shown in the formula are used only for calculating
average used capacity and can't be added directly to a
report.
Formula: SUM(omnichannel_agent_productivity_capacity_seconds) /SUM(omnichannel_agent_productivity_agent_online_seconds) |
Agent max used capacity | The maximum capacity used by an agent for a channel during a period. |
Agent min used capacity | The minimum capacity used by an agent for a channel during a period. |
Max capacity limit | The maximum capacity assigned to an agent by the admin per channel. |
Offer count |
The number of work items offered to the agent during a specific period. Available only for web and mobile messaging, social messaging, and live chat. Acceptance rate is not available if auto-accept or broadcast mode is used for routing. Offer count is measured at the agent level, meaning this metric returns values for the agent regardless of group filters. |
Accepted count |
The number of work items accepted by the agent during a specific period. Available only for web and mobile messaging, social messaging, and live chat. Acceptance rate is not available if auto-accept or broadcast mode is used for routing. Accepted count is measured at the agent level, meaning this metric returns values for the agent regardless of group filters. |
Acceptance rate |
The ratio of accepted work items to the total work items assigned to an agent during a specific period. Available only for web and mobile messaging, social messaging, and live chat. Acceptance rate is not available if auto-accept or broadcast mode is used for routing. If no work items are offered to or accepted by an agent in the given time period, this metric returns a null value. Acceptance rate is measured at the agent level, meaning this metric returns values for the agent regardless of group filters. Formula: IF (ATTRIBUTE_ADD(SUM(Offer count),[Agent productivity start time - Date]) > 0) THEN 100*SUM(Accepted count)/SUM(Offer count) ELSE 0 ENDIF |
Agent productivity attributes
This section lists and defines all attributes available in the Agent productivity dataset.
Attribute | Definition |
Group ID | The ID number of the group. |
Group name | The name of the assigned group. |
Channel | The channel that a capacity value relates to. Possible values include Support, Talk, Messaging, and Chat. |
Agent name | The name of the agent. |
Agent role | The role of the agent. |
Agent ID | The ID of the agent. |
Agent email | The email address of the agent. |
Agent external ID | The external ID of the agent. |
Agent locale | The locale of the agent. |
Agent is moderator | Whether the agent is a moderator. |
Agent status | Whether an agent is active or suspended. |
Agent active | Whether the agent is active. |
Agent suspended | Whether the agent is suspended. |
Agent organization ID | The organization ID of the agent. |
Agent tags | Tags added to the agent. |
Agent timezone | The time zone of the agent. |
Time - Agent productivity start time | A collection of attributes in different units of time that return when work item capacity was recorded. The start time is recorded when the work item is assigned to the agent. |
Time - Agent productivity end time | A collection of attributes in different units of time that return when the end time when work item capacity was recorded. The end time is recorded when the work item is solved or assigned to another agent. |
16 comments
Casey Eisenberg
Hi Erin,
Thanks for this information, I have been seeing some issues with the data but I understand this is still in EAP. Is there an estimate for when this will be out of EAP and the dataset will be reliable?
Thank you.
0
Karen Hynes
Hi Casey Eisenberg,
Do you mind creating a ticket/sharing what issues you are seeing with the data?
Thank you!
Karen
0
Casey Eisenberg
Hi Karen Hynes,
Yes I had created a ticket where the team told me since this was still in EAP that was to be expected. For daily totals in agent state I am seeing multiple instances of agents having greater than 24 hours of a certain state in a day.
Thank you,
Casey
0
Stephen
Hi Karen Hynes,
How do you sign-up for this EAP?
Best regards,
Stephen
0
Brett Bowser
Let me know if you have any other questions!
0
Stephen
Hi Brett Bowser,
If I not mistaken, I think the Agent Capacity is a new dataset. I don't see it available when I select Omnichannel:![](/hc/user_images/01HQQPAKGV7YP7WN5EA53XQMBG.png)
I see a report mentioned here that suggests using this newer dataset: https://support.zendesk.com/hc/en-us/articles/6446747033626-Explore-recipe-Reporting-on-maximum-capacities-with-omnichannel-routing
0
Elaine
The article is accessible to customers who successfully signed up for the EAP (Early Access Program). However, the EAP registration period has ended. The good news is that the feature will soon be released to all users. Please visit our announcement page for further updates and information.
0
Cécile Gao
Hello,
I created a support ticket but was advised this was an EAP and was redirected here to get an answer to my queries.
I understood from your documentation that "unified away" collects all times spent in away states across custom statuses. When enabling custom agent status I noticed the regular "Away" status still exists in the status selection for the agent. Does this mean that :
-when agent uses the regular (non-custom) "Away" status, that time is available in "Agent state" = "Away"
-when using custom status "On Break" (if defined as away on all channels), that time will be in "Agent state" = "On Break" BUT will also be in "Agent state" = "Away" (so we shouldn't use both states in same report otherwise time will be doubled) ?
If I am away 1h on regular "Away", and away 1,5h on custom status "Lunch break", will my query with agent state = "Away" (NOT unified away) give me 2,5h or only 1h ?
Is the time spent in a custom status also included in the time spent in the regular status it is built on ?
I'm asking because I used the "Pause déjeuner" (lunch break) status at 12:02 in my account, but noticed the time spent in that custom status also popped up in "away" as well as in "unified away" status details.
Also, the prerequiqites of getting agent state EAP were to have agent workspace and omnichannel routing in place. Now I'm considering stepping back from omnichannel routing to the old talk way, but would like to know if I do so if the agent state dashboards / metrics will still be in my account with accurate data ?
Thank you,
rgds
0
Karen Hynes
Hi Cécile,
Happy to help! Please refer to the following documentation for more information on unified agent status and creating custom unified agent statuses.
There are four default unified agent states: online, away, transfers only, and offline, as well as a custom states. Therefore, regardless of creating custom states, the four default unified states will always be available.I think the confusion here may be as a result of the custom states mapping to per channel states and the impact this has on reporting. We have an article to help with this available here.I will elaborate on this further below:
Reports can be filtered by CHANNEL. When OCR is enabled, by default the pre-canned dashboard is filtered to display UNIFIED channel. UNIFIED channel reports on default unified and custom states. For custom states, as a result of their configuration by channel, there would also be per channel reporting available. This can be viewed by filtering for all channels (screenshot 1).
Pause déjeuner would be reported on under the UNIFIED channel for 1.5 hours. If this is mapped to AWAY for all channels, you would see this time reported as Away when you filter by Away for Support, Talk, Messaging channel (screenshot 2 - filter includes channel Away state and Custom state (eg: pause déjeuner)).
If you were in UNIFIED Away state prior to Pause déjeuner, this would also be included in the per channel Away duration. Take the following example for the Messaging channel:
I hope this helps clarify your question. These datasets have also moved to GA, as per announcement here.
This is also available for non OCR customers on Professional+ plans who have agent workspace enabled but please note, custom state data would not be available and therefore you would just see per channel mapping.
0
Benjamin David John Clough
Hi,
It will be extremely useful to incorporate tickets updated, tickets assigned, ticket solved and other ticket metrics into this report.
We focus heavily on productivity and currently I'm having to extract from the Updates History Dataset or Support Tickets Data Sets, then offline in an excel sheet incorporate the worked hours.
Please let me know if this is possible?
1
Shai Eyal
Can we customize the "time in state" and add "Start time" and "end time" so we can specifically report on these?
0
Karen Hynes
Hi Benjamin David John Clough,
This feedback is noted and we will be releasing another dataset shortly which contains further information regarding agent capacity. After this release, we hope to enhance this capacity dataset with further agent level productivity metrics in H2.
Thanks,
Karen
0
Karen Hynes
Hi Shai Eyal,
If you are utilising the 'agent state' dataset, you can report on state start time, state end time and also duration in state.
Thanks,
Karen
0
Farid
Hello, i am currently trying the new dataset : Omnichannel : Agent Productivity.
When im doing Agent Avg Used Capacity here, the calculation commented there is
SUM(omnichannel_agent_productivity_capacity_seconds) / SUM(omnichannel_agent_productivity_agent_online_seconds)
I can see my lower performing agent scores higher at this, which is weird.
However, i notice on the agent, she will only set online when the traffic is busy and sets away again once the queue is less.
I would assume based on the formula above, it calculates the capacity of the agent at the time, over the unified online time.
So as soon as she sets away, the counter of agent_online_seconds stops counting.
With this , it doesn't really give us a full picture of the agents are having. Could you advise where should i look on for this context? Thank you.
0
D.Fitz
Is there any plan to incorporate ticket data here? We're using our own custom Tickets Solved Per Hour metric within the Update History dataset, but if we could combine ‘Tickets Solved’ with some of the agent productivity metrics that would be a gamechanger.
1
Jahn
Karen Hynes Erin O'Callaghan - is there a chance we can incorporate ticket data here as what D.Fitz mentioned above?
We want to check the validity of the metrics this dataset is providing by adding drill in function to see the ticket ID whether it is really missed by an agent. So far, there's no other way to validate whether the data on this dataset is 100% accurate.
0