Agent Groups

Agent groups are a collection of users who include a number of user accounts. Distribution groups or individual user accounts might be added to an agent group. After group membership has been defined, administrators can select how calls are routed within the agent group, such as round robin or in parallel. The last option administrators can define is whether users are automatically signed in to the agent group when signed in to Lync or whether they can manually participate or leave the group. The following options are available when creating an agent group:

• Participation Policy—Determines whether agents need to sign in or out of the group manually. Selecting formal here means users have to manually enter and leave the agent group. Informal means the agents are automatically included in the group as long as they are signed in to Lync.

• Alert Time—The number of seconds a call rings an agent before attempting to ring the next agent.

• Routing Method—How the calls are routed among agents in the group.

• Agents—User accounts or a distribution group used for the agent group membership. Keep in mind that distribution groups do not recognize nested groups and that only one distribution group can be specified.

The routing methods are a key part of defining how agents take calls. These options are separated here for some additional clarity on behavior.

• Longest idle—The call is routed to the agent who has had a presence status of Available the longest without taking a call. For example, if three agents are part of the agent group and one agent is Busy while two are Available, the call is routed to the user with the Available presence the longest.

• Parallel—Rings all agents at the same time. The agent who accepts the call first is placed in a conversation with the caller.

• Round Robin—Call requests are evenly sent to agents. Assuming three agents exist, the first call goes to Agent A, the second to Agent B, and the third to Agent C. The fourth call rings Agent A again.

• Serial—Calls are sent to agents in the order defined in the agent list. Assuming three agents exist, the first call goes to Agent A. The next call again attempts to ring Agent A, and if Agent A is unavailable, the call then goes to Agent B. The difference from round robin distribution is that the next call will follow the same order, starting with Agent A again.

• Attendant—Calls are routed to all agents just as in parallel fashion, but includes agents who are busy or currently in a call. Calls are not routed to agents with a status of Do Not Disturb.

To create a new agent group, use the following steps:

1. Open the Lync Server 2010 Control Panel.

2. Click Response Groups.

3. Click Group.

4. Click New.

5. Select an application server and click OK. This is typically just a Front End pool.

6. Enter a Name for the group.

7. Enter a Description for the group.

8. Select a Participation policy for the agents.

9. Specify the Alert time in seconds for how long a call will ring an agent.

10. Select a Routing method for the group.

11. If using a distribution list for the agent list, select Use an existing email distribution list and then enter the SMTP address of the list.

12. If manually adding agents to the group, click the Select button.

13. Enter a search for users and click Find.

14. Highlight the selected user and click OK.

15. Click Commit after all agents have been added.

Alternatively, the Lync Server Management Shell can be used to create a new agent group:

New-CsRgsAgentGroup –Parent ApplicationServer:<Application Server ID> -Name
<Agent Group Name> -AgentAlertTime <Seconds a call will ring agents>
-AgentsByUri <Comma separated SIP URIs to add as agents including sip:
prefix> -Description <Agent Group Description> -DistributionGroupAddress
<Distribution Group SMTP Address> -ParticipationPolicy <Informal | Format>
-RoutingMethod <LongestIdle | RoundRobin | Serial | Parallel | Attendant>

After completing the agent group configuration, continue the Response Group setup process by creating queues.