Over the years routing strategies have been developed to improve productivity and quality of call centre operations in a structured way.
Productivity is improved by routing incoming calls to the longest available agent, while quality is improved by routing incoming calls to that agent who has the best profile to answer the call from a knowledge perspective: the call is routed to the agent with enough knowledge to be able to answer the call , and who is available in a reasonable time.
Intelligent routing strategies will aim to combine both quality and productivity aspects in the routing decisions.
The first routing algorithms were based on “group based routing”. In “group based routing”, incoming calls are presented to a defined group of agents and any available agent in that group can take the call.
This concept has a major limitation because an agent either belongs to a group or does not. It is very difficult to indicate whether a particular agent has more knowledge on a certain topic than another.
The only way around this, is to create a complex mechanism of overflow from the initial group to another group (e.g. to a group of less trained agents), but once the overflow is activated, an agent from the initial group is no longer able to handle the call.
The introduction of “skill-based routing” improved the ability of the call centre manager to organize the call centre by enabling him/ her to add knowledge aspects to the agent definitions.
On the one hand calls can be presented to more than one group at the same time, hence overcoming weakness of the obsolete overflow mechanism.
On the other hand, an agent is allowed to be available in more than one skill (group) at the same time.
Furthermore, parameters can be set to define the degree of knowledge that every agent has for a particular skill. This enables a much more accurate decision during the routing process.
“Service Level Agreements” are implemented in call centres in line with the possibilities of the routing mechanisms available on the installed equipment.
Typical SLA’s define the percentage of calls that need to be answered within a predefined number of seconds. Other SLAs define both the average and maximal waiting times allowed.
Most SLA’s implemented in today’s call centres have a quantitative nature.They control time-related aspects for answering the incoming calls. Qualitative measure points, e.g. did the customer get the correct answer or was it necessary to transfer the customer to a more experienced agent, are rarely found in SLA’s.
Today’s customers are no longer satisfied if a contact centre only respects those traditional quantitative SLAs. Not only do they expect to be answered timely, but they also expect to get an accurate answer on the first contact.
Call centre management requires cost reductions and more value out of each contact.
Both imperatives call for increasingly complex routing algorithms. Today one wants a call to be measured against estimated value, and routed to that agent who can yield the correct outcome out of the call in the most cost efficient way. This becomes even more important when adding new aspects such as agents’ delocalization to low cost countries, or allowing for new media that require to be treated and measured in the same way as traditional phone calls.
In some cases there is even a need to blend the different media. Value becomes the decision driver and one wants to have the best and most cost efficient agent to handle a particular demand.
Taking someone who is overqualified not only represents a higher cost, but can also result in the non-availability of that person for another communication with a higher potential value.
Implementing “profit based routing” by using traditional skill based routing means that a skill has to be defined for each value-range. This will result in an enormous number of skills and waiting queues making the life of the contact centre manager and supervisors simply impossible. It will even be less cost efficient because a dramatic increase in the number of skills will cause an increase in the required number of agents with the ability to handle all defined skills.
Profit based routing therefore has to be implemented as an extension to skill based routing. Every demand is then evaluated on value. This can be done by a set of interactive questions using an IVR, or by customer identification technology in combination with service numbers, customer contact history or IVR sessions. Once this procedure is completed, the demand can then be transfered to the appropriate skill.
In addition, the cost of every agent who can take the demand is calculated (it is the sum of the cost of the agent and the cost of transferring the demand to the particular agent). Both figures are put together and the most appropriate, cost efficient agent is chosen to handle the demand.
Profit based routing responds to both the customers’ and the contact centre management’s expectations. As it remains easy to administer, profit based routing will change the way contact centres are operated, resulting in lower costs and more satisfied customers.
