Preferred agent routing behavior
This article describes how preferred agent routing works for various routing and agent scenarios.
- Similar to bullseye routing, when the system matches more than one preferred agent, it initially evaluates the top agent first, then the next highest agent.
- You can choose the evaluation rules. For example, you can ring the top two agents for 30 seconds and the top five agents for the next 30 seconds. The pool of agents only expands; the system does not reduce it.
When an agent blind transfers or consult transfers an interaction set with preferred agents, the system no longer considers those preferred agents during the transfer.
If you use bullseye routing along with preferred agent routing, the system considers preferred agents first, regardless of whether they have the required skills.
Score preferred agents in Architect. You can score agents in a range from 0 to 100. The higher the score, the more preferred the agent. Multiple agents can have the same score.
Preferred agent routing is not compatible with workforce management scheduling and forecasting.
Last agent routing does not work for email if you use email flows. It works only if you send the email directly to a queue. For more information, see Email routing and preferred agent routing in ACD email overview.
For email and messaging interactions, and inbound callbacks, with preferred agent routing the system no longer attempts to route the interaction to the last agent who handled it. Scheduled callbacks, however, are unaffected by preferred agent routing.
For more information, see Advanced routing overview.
- Preferred agent routing only routes an interaction to preferred agents who belong to the queue in which the system routes the interaction. Agents who do not belong to the queue do not receive routed interactions, even if you designate them as preferred agents.
- Architect provides a way to create a collection of agents and scores. However, during flow authoring, make sure to provide the appropriate agents that are members of the queue.
- If all preferred agents in the ring are not a member of the queue, the system skips configured timeout for preferred agent routing rings.
Single preferred agent specified (timeout ignored)
- Specify one preferred agent as part of a flow. The timeout is 900 seconds, or 15 minutes. An interaction routes to a queue to which the preferred agent does not belong. The system does not follow the 15-minute timeout, and instead immediately routes the interaction to the next available agent.
Multiple preferred agents specified (timeout not ignored)
- Specify 10 preferred agents as part of a flow. The timeout to reach the first agent with a score 100 is 300 seconds, and the timeout to reach all preferred agents is another 300 seconds. The total timeout is 600 seconds, or 10 minutes, for the system to attempt to route the interaction to one of the preferred agents specified.
- The first agent belongs to the queue to which the system routed the interaction. Of the nine remaining agents, only five belong to the queue.
- In this case, routing works as follows:
- The system attempts to route to the first preferred agent for 300 seconds.
- For the next 300 seconds, the system attempts to route the interaction to the first preferred agent and the five other preferred agents that belong to the queue.
The system only routes interactions to agents who are active on a specific queue, and ignores inactive agents. If an agent becomes active on a specific queue while the system attempts to route an interaction to preferred agents, then the system includes the newly activated agent in the pool. In this case, the system still considers timeout, because an agent is active on the queue. For example, if a preferred agent returns from lunch and then becomes active on the queue, the system immediately routes an email intended for that agent.
The system only routes to agents who are on queue. For example, it does not route interactions to preferred agents who are on break.
The system routes interactions to preferred agents first, regardless of whether they have the required ACD skills and language skills.
When a preferred agent is fully utilized, the system does not route interactions to that agent. If the agent is not fully utilized during the configured preferred agent timeout, then the system considers them for routing. For more information, see Agent utilization.
Interactions with higher priority take precedence over preferred agent routings. For example, if an agent is available and has an interaction that lists them as a preferred agent in the queue, but another interaction in the queue has higher priority, then the system first routes the higher priority interaction to the agent.