E-Handbook: Contact center workforce management keeps pace with consumers Article 4 of 4

auremar - Fotolia


3 call routing methods to consider in your contact center

Contact center call routing methods include next-available agent, skill-based and predictive; however, each strategy has pros and cons. Learn what is best for your business.

A basic function of any contact center is routing customer calls to agents for resolution.

Over time, call routing methods evolved as businesses sought ways to increase call center efficiency. Here are three of the most common routing practices, still in use in contact centers today:

1. Next-available agent routing

In this original call routing method, businesses use an automated call distributor (ACD) to send incoming calls to agents who have been waiting the longest. If no agents are available, calls wait in queue until someone is free.

In this model, agents receive all types of requests. If the employee does not have the proper training or skills to handle the specific inquiry, agents transfer calls to the appropriate person or department, or schedule a callback.

The downside to this method is that as inquiries become more complex, the likelihood of all agents being able to resolve every inquiry is lower.

2. Skill-based routing

The next step in the evolution of call routing methods is skill-based routing, where an ACD routes calls to the next available agent within a group of employees with a specific skillset.

There are two requirements to support skills-based routing. First, contact centers must establish distinct call queues composed of specially trained agents to perform specific functions. Second, contact centers must program routing systems to use specific caller information when customers dial in to route calls to the proper employees. Such information includes:

The latest call routing method is predictive routing, which uses AI and machine learning to determine how calls should be routed to agents.
  • caller data, such as a dialed number identification service number;
  • customer entered data, such as a billing dispute; and
  • customer data from a database, such as past due account information.

As an example, a contact center may set up a queue with a group of agents to work with customers who are past due on their accounts. When a customer calls into the contact center and enters a unique ID, the system might determine that the customer is past due and automatically route that customer to the "past due" queue.

This enhanced routing capability reduces the number of dial transfers that next-available agent routing would have produced. However, this call routing method can also create inefficiencies when too many specialty queues are set up with a small number of agents.

Call routing methods chart

3. Predictive routing

The latest call routing method is predictive routing, which uses AI and machine learning to determine how calls should be routed to agents.

Predictive routing focuses on achieving a specific outcome from a customer interaction by analyzing large amounts of data from both a caller and agent perspective. Desired outcomes can include increases in sales, retention or customer satisfaction.

Predictive routing also uses a feedback loop, where the actual outcome from a specific interaction is fed back into the model to continue to enhance the accuracy and success of the routing model.

Two things need to happen to support predictive routing. First, similar to skill-based routing, predictive routing analyzes customer data and behaviors by using additional technologies such as speech analytics, which can identify the frustration level of a customer in real time. Second, contact centers need to develop individual agent profiles that capture specific attributes for each person -- such as average handle time, first contact resolution, experience level and customer feedback.

For example, if a caller has a technical support issue and is calling for the second time on a given day, it may be more effective to automatically route the call directly to a tier 2 technical support agent. And if speech analytics determines that the customer is frustrated, it may route the call to a tier 2 technical support agent who has a strong reputation for defusing angry customers.

The major goals of this enhanced routing process are to improve the customer experience and attain specific outcomes that are important to the customer and the organization.

There are two potential downsides to consider with predictive routing. First, customers may have to wait a little longer in queue before being routed to a specific agent; however, this wait is often offset by the improved experience with the agent. And second, top agents may get overburdened with calls as a result of the superior service they provide while less effective agents remain idle.

Dig Deeper on Customer service and contact center

Content Management
Unified Communications
Data Management
Enterprise AI