System and methods for intelligent routing of customer requests using customer and agent models

Description Submit all comments and votes

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to routing requests from customers to agents, and more specifically to routing requests from a customer to an agent best matched for that customer based on a model of the customer and models of the agents.

2. Discussion of Background Information

A business or company that provides services and/or products to clients or customers may provide their customers with customer service in the form of a customer service center that handles customer requests. Customer requests may comprise requesting new products or services, getting support for a product or service, asking questions about a product or service, etc. If a customer made a call to the service center with a request, the service center would route the call to an agent or other capability that would service the customer's request.

The ideal incoming call center would route incoming calls so that each call is handled to the complete satisfaction of the customer, and also supports the business performance objectives of the center. The routing of incoming calls is accomplished with full and complete knowledge of the functionality, technology, and customer and agent requirements.

While functionality and technology issues are well represented in the management of a call center, a full rich understanding of the behavioral characteristics of the customer population, agent population, and the customer/agent interactions are not well understood. Presently, incoming calls are routed on a random "first come, first served" basis. The characteristics of both the customer and agent are simplistic by assuming a single view of both populations. This single view approach limits management's ability to influence the center's performance because not all of the customer's or agent's behaviors are accounted for during call routing. Thus the implemented routing system may function well for a certain set of customers or agents, while not being well matched to other sets within the population. When overall performance is critical, these mismatches and resulting reduced performances can cost organizations time and money.

The common practice in the industry is to route incoming calls to an available agent. There are two variations of this "available agent" routing in current practice. One variation is to route incoming calls to an available agent who has been idle, or not handling a call, for the longest period of time. The second variation is to route incoming calls to an available agent that is randomly chosen among agents of whom are available to take a call (no consideration for how long each individual agent has been idle). The current practice represents and treats the customer call population with a single set of characteristics and behaviors, i.e. any agent can take any customer call.

Some call centers may split off incoming customer calls to a dedicated agent group when the customer has a specific language preference. Some call centers also split off customer calls that originate from residences separate from customer calls that originate from businesses. In addition, some call centers split customer calls where the customer wishes to place an order for a service separate from the customer calls where the customer wishes additional information on their bill or similar information. For call centers that do separate incoming calls based on some criteria, the result of the actual call routing to an agent is still a randomly based transaction. Whether a call center performs any preliminary routing or not, currently, none of the call centers route customer calls to agents using a performance optimizing calculation.

Presently, if agents are categorized, they are done so on an informal basis based primarily on the opinion and judgment of the local operating management of the call center. Even though these agents may be identified by category, calls are not routed dependent upon those categories. Another aspect of current practice is that the description of behavior is done anecdotally, not statistically. Quantitative performance results are not incorporated into the behavioral descriptions. Customer and agent models are generally not constructed, primarily because there is only one agent representation and both management and agents accept that single view.

SUMMARY OF THE INVENTION

Accordingly, the present invention is directed to a method for intelligent optimized routing of requests from customers to agents based on behavioral models of the customers and the agents that substantially obviates one or more of the problems arising from the limitations and disadvantages of the related art.

It is an object of the present invention to provide a system and methods that use customer and agent models for matching a customer with an optimal agent.

It a further object of the present invention to provide a system and methods for routing requests from customers to optimal agents that increases customer satisfaction.

Another object of the present invention is to provide a system and methods for routing requests from customers to optimal agents that improves performance at a call center.

Accordingly, one aspect of the present invention is directed to a method for intelligent routing of requests from a customer to an agent that includes receiving a request from a customer, creating a behavioral model of the customer, creating behavioral models of at least two agents, matching the behavioral model of the customer with each of the behavioral models of the at least two agents, and routing the request from the customer to a best match agent of the at least two agents based on the matching.

According to another aspect of the present invention, the behavioral model of the customer is calculated from a detailed profile of the customer's needs, task objective, sales preferences, and expectations for satisfaction.

According to yet another aspect of the present invention, the behavioral model of the at least two agents is calculated from a detailed profile of the at least two agents' sales strategies, customer service behaviors, and sales performance.

In a further aspect of the present invention, the matching includes performance optimization calculations.

The present invention also includes obtaining identifying information from the customer where the identifying information may include the customer's telephone number, e-mail, or account number.

According to another aspect of the present invention, the identifying information is obtained by a human operator, an automatic caller ID system, an interactive voice response (IVR) interface, e-mail, Internet, or other communications channel.

According to yet another aspect of the present invention, the invention includes retrieving historical background information regarding the customer based on the identifying information.

In a further aspect of the present invention, the invention includes searching for historical background information regarding the customer's account based on the identifying information.

In the present invention, the customer's account information may include the customer's billing history, products and services currently being provided, or household information.

According to another aspect of the present invention, the invention includes obtaining information regarding the customer's current task objective and the customer's current expectations for satisfaction, the customer's current task objective and the customer's current expectations for satisfaction each including one or more attributes.

According to yet another aspect of the present invention, the customer's expectations for satisfaction attributes including the customer's willingness to be up-sold additional products or services, the customer's preference for a lengthy or brief negotiation, or the customer's desire to have questions answered.

According to a further aspect of the present invention, the invention includes upgrading the historical background information based on the customer's current task objective and the customer's current expectations for satisfaction.

According to another aspect of the present invention, the task objective attributes and the expectations for satisfaction attributes may be quantified and transformed into numerical values.

According to yet another aspect of the present invention, the invention includes calculating and constructing the customer model using the task objective attributes values and the customer expectations for satisfaction attributes values.

According to a further aspect of the present invention, the at least two agents' sales strategies, customer service behaviors, and sales performance each include one or more attributes.

According to another aspect of the present invention, the sales strategies attributes, customer service behaviors attributes, and sales performance attributes are quantified and transformed into numerical values.

According to yet another aspect of the present invention, the invention includes calculating and constructing the agent model using the sales strategies attributes values, customer service behaviors attributes values, and sales performance attributes values.

According to a further aspect of the present invention, each task objective attribute and each customer expectations for satisfaction attribute are assigned a weighting value based on a relative importance of the each attribute.

According to another aspect of the present invention, each sales strategies attribute, customer service behavior attribute, and sales performance attribute is assigned a weighting value based on a relative importance of each attribute.

According to yet another aspect of the present invention, the weighting value for each attribute is used in the calculating and constructing of the customer model.

According to a further aspect of the present invention, the weighting value for each attribute is used in the calculating and constructing of the agent model.

According to another aspect of the present invention, the performance optimizing calculations are used to generate a match score for each of the at least two agents.

According to yet another aspect of the present invention, the best match agent is the agent with the highest match score.

According to a further aspect of the present invention, a list of optimal agents is generated based on the match scores of the at least two agents that are above an optimal threshold.

According to another aspect of the present invention, the request from the customer is routed to an available agent on the list of optimal agents.

According to yet another aspect of the present invention, the request from the customer is placed in a wait queue until an agent on the list of optimal agents becomes available.

According to a further aspect of the present invention, additional agents are added to the list of optimal agents the longer the request from the customer remains in the wait queue. The additional agents are added after reducing the optimal threshold. A wait time is increased while the request from the customer is in the wait queue.

According to another aspect of the present invention, the request from the customer is routed to an available agent with the highest match score when the wait time equals a maximum wait time.

According to yet another aspect of the present invention, the optimal threshold and the maximum wait time are set by a call center controller.

According to a further aspect of the present invention, the invention includes a method for intelligent routing of requests from customers to agents that includes: receiving a request from a customer, accessing identification information related to the customer, retrieving background information on the customer, gathering task and attitude information about the customer, calculating a model of the customer, retrieving models for at least two agents, performing performance optimizing calculations using the customer model and the models for the at least two agents, identifying an optimal agent from the at least two agents based on said performing, and routing the request to the identified optimal agent.

According to another aspect of the present invention, the invention includes an intelligent routing system for intelligent routing of requests from customers to agents using models of the customers and models of the agents that includes a first interface where the first interface receives requests and information from at least one customer. The first interface communicates customer data gathering information to the at least one customer from the intelligent call routing system. A second interface receives agent availability information from at least two agents. The agent interface communicates agent data gathering information to the at least two agents from the intelligent call routing system. A database contains behavior models of the at least one customer and behavioral models of the at least two agents. A first processor is operatively connected to the first interface and the database where the first processor generates customer models based on the customer data gathering from every at least one customer and stores the customer models in the database. The first processor gathers customer identification, customer task, and customer attribute information from the at least one customer. The first processor retrieves one of the customer models from the database based on the customer identification, the customer task, and the customer attribute information. A second processor is operatively connected to the second interface and the database where the second processor generates an agent model based on the agent data gathering from every at least two agents and stores the agent models in the database. The second processor gathers agent identification and agent attribute information from the at least two agents, and retrieves the agent models from the database. A third processor is operatively connected to the first processor and the second processor where the third processor compares the one of the customer models with every agent models associated with each at least two agents and calculates a match score for each at least two agents. The third processor generates and stores a list containing best matching agents where the best matching agents include all the at least two agents with a match score at or above a threshold value. A timer is operatively connected to the first interface where the timer tracks the length of time from receipt of the customer request. A fourth processor is operatively connected to the third processor, the first interface, the second interface, and the timer. The fourth processor monitors the availability of each best matching agents. The fourth processor routes the customer request to the first available best matching agent.

According to yet another aspect of the present invention, the fourth processor uses a match graph associated with the customer, the match graph is generated by the first processor based on the customer model and parameters from a service center, the match graph defining times during the length of time from receipt of the customer request and associated lower threshold values when additional best matching agents may be added to the list based on the lower threshold values.

Other exemplary embodiments and advantages of the present invention may be ascertained by reviewing the present disclosure and the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is further described in the detailed description which follows, in reference to the noted plurality of drawings by way of non-limiting examples of preferred embodiments of the present invention, in which like reference numerals represent similar parts throughout the several views of the drawings, and wherein:

FIG. 1 is a flow chart of an exemplary method for intelligent call routing according to the present invention;

FIG. 2 is a flow chart of initial customer inquiry and identification according to the present invention;

FIG. 3 is a flow chart of the query of additional customer information according to the present invention;

FIG. 4 is a flow chart of customer model calculation and information update according to the present invention;

FIG. 5 is a flow chart of inputs and computation of the performance optimization calculation according to the present invention;

FIG. 6 is a block diagram of an exemplary intelligent routing system according to the present invention;

FIG. 7 is an exemplary performance optimization calculation according to the present invention;

FIG. 8 is a table of exemplary attributes and weights according to the present invention;

FIG. 9 is a table of exemplary attribute scores for two agents according to the present invention;

FIG. 10 is a table of exemplary attribute scores for an exemplary Customer 1 according to the present invention;

FIG. 11 is a table showing exemplary Agent 1 and Customer 1 match calculation according to the present invention;

FIG. 12 is a table showing exemplary Agent 2 and Customer 1 match calculation according to the present invention;

FIG. 13 is a table of exemplary attribute scores for an exemplary Customer 2 according to the present invention;

FIG. 14 is a table showing an exemplary Agent 1 and Customer 2 match calculation according to the present invention;

FIG. 15 is a table showing an exemplary Agent 2 and Customer 2 match calculation according to the present invention;

FIG. 16 is a flow chart of an exemplary acceptable agent availability question and results according to the present invention;

FIG. 17 is a graph of an exemplary call wait function according to the present invention;

FIG. 18 is a graph of an exemplary call wait function with an acceptable match score of 80 and a maximum 30 second wait according to the present invention;

FIG. 19 is a graph of an exemplary call wait function with an acceptable match score of 60 and a maximum 30 second wait according to the present invention;

FIG. 20 is a graph of an exemplary call wait function with an acceptable match score of 90 and a maximum 60 second wait according to the present invention;

FIG. 21 is a graph of an exemplary call wait function with an acceptable match score of 70 and a maximum 40 second wait, then a decreasing match score up to a 70 second wait according to the present invention; and

FIG. 22 is a graph of an exemplary call wait function with an acceptable match score of 80 and a maximum 30 second wait, then a decreasing match score up to a 60 second wait according to the present invention.

DETAILED DESCRIPTION OF THE PRESENT INVENTION

The particulars shown herein are by way of example and for purposes of illustrative discussion of the embodiments of the present invention only and are presented in the cause of providing a useful and readily understood description of the principles and conceptual aspects of the present invention. In this regard, no attempt is made to show structural details of the present invention in more detail than is necessary for the fundamental understanding of the present invention, the description taken with the drawings making apparent to those skilled in the art how the several forms of the present invention may be embodied in practice.

The present invention integrates customer and agent models into a single perspective. A customer desires to accomplish a particular task, e.g., requesting information about a particular product or service, ordering a product or service, or requesting additional information concerning a