Patent Application
20060023863
The present disclosure relates generally to mapping information to specific call center transactions.
Tens of millions of calls are routed to call centers annually. A large percentage of these calls use agents to identify and map the customer's opening statement to one or more agent transaction screens that are accessed to fulfill the customer's request. For example, a customer's opening statement may ask for the creation of an order for DSL provisioning and an agent may access one or more screens associated with provisioning DSL. Currently, agents simultaneously control multiple support systems by accessing these transaction screens. This accessing and mapping process accounts for a significant portion of the average call length primarily because of the considerable mental processing demands associated with existing agent support systems. It is estimated that approximately 10-20% of the agent time for handling an average call is associated with the mapping process. This mapping process reduces the volume of calls an agent can handle and increases the average per call expense. As such, there is a need for an improved system and method for mapping caller information to call center agent transactions.
In a particular embodiment, the disclosure is directed to a system including a call center routing and control module, a transaction database including a plurality of transactions that are suitable for execution by the call center, transaction selection logic responsive to the transaction database and including logic to select a set of transactions for execution in response to the transaction identifier, and display screen sequence logic responsive to the transaction selection logic. The call center routing and control module has an input to receive incoming calls and to receive a transaction identifier indicative of a caller transaction to be executed. The display screen sequence logic is configured to generate agent display screen commands corresponding to a sequence of display screens associated with the set of selected transactions.
In another particular embodiment, the disclosure is directed to a method of processing a call at a call center. The method includes receiving the call at the call center, receiving an indication element associated with a call center transaction, retrieving call center transaction data based on the indication element, and generating a sequence of pre-populated call center agent terminal transaction processing screens based on at least a portion of the call center transaction data.
In another particular embodiment, a method of call handling is disclosed. The method includes collecting data at an interactive voice response unit, forwarding the collected data to a remote analysis module, comparing the collected data to a plurality of call center transactions, populating display screens, displaying the populated display screens at a call center agent terminal in response to determining a match between a particular call center transaction and the collected data, and processing the particular call center transaction.
Referring to
During operation, a customer call 130 from a representative caller 120 is received at the telephone exchange 102. The telephone exchange 102 forwards the call and a customer opening statement 132 originated by the caller 120 to the interactive voice response unit 104. The interactive voice response unit 104 maps the customer opening statement to a transaction identification element, such as an action object data pair 140 that is fed through network 106 to be received at the call center routing and control module 108. The action object pair 140 is an identifier of a particular call center transaction that may be executed by the call center supporting the customer call. The transaction identifier, such as an action object, functions as a prediction of a requested call center transaction service that may be provided to the caller 120. In effect, the system, based on an initial evaluation and analysis of customer opening statement 132, predicts a particular call center transaction or set of transactions that are likely being requested by the caller 120. An example of an illustrative action-object pair matrix that maps various opening statements to call center transactions is provided below:
At call center 108, the transaction identifier, such as the action object pair 140, is received. The action object pair 140 is routed to the transaction selection logic 116. The transaction selection logic 116 receives transaction data entries from the agent transaction database 114. Out of a plurality of transactions that are suitable for execution by the call center, a subset of such transactions and data representative thereof is fed to the transaction selection logic 116 for further processing. Transaction selection logic 116 thereby provides a transaction subset 160 in response to the action object pair 140 and selected transaction data 150. The transaction subset 160 is fed to custom screen sequence logic 112. The custom screen sequence logic 112 transforms the particular transaction subset 160 and produces populated screen sequence commands 170. An example of such commands 170 includes terminal commands forwarded to agent terminal 110 to provide a programmed sequence of pre-populated agent terminal screens. The sequence of agent terminal screens corresponds to the selected transaction or subset of transactions identified by the transaction selection logic 116. The sequence of agent terminal commands also corresponds to the particular action object pair 140 communicated to the call center in response to the customer opening statement.
The agent terminal 110 includes a display screen that provides a sequence of pre-populated screen views to an agent in response to receiving the agent display screen commands from the custom screen sequence logic 112. An example of agent terminal commands is presented below:
Command Sequence for Change Due Date (CDUE) Transaction:
Referring to
In the event that there is no match between the collected data and the action object matrix, then processing proceeds to request input from a live agent at an agent terminal, as shown at 216. In this case, a live agent would request further information from the caller and would determine the requested transaction without the assistance of using the action-object pair data. The agent would then enter the requested transaction as well as associated information from the caller to fulfill the customer requested transaction, at 218. Thus, where a successful match is determined between the action object and transactions in the matrix 210, certain manual processing steps typically performed by a live agent at a call center may be automated, thereby reducing agent call center transaction time.
Given the high call volume of call centers and the benefits of enhanced customer care, the disclosed method and system of automating and pre-populating certain requested transactions may beneficially reduce call center costs and provide more efficient and responsive processing of requested customer actions.
Referring to
The disclosed method and system provides an automatic mapping of information from customer opening statements at IVR units to agent transaction screens to be accessed in order to fulfill customer call center requests. The mapping process may account for a significant portion of an average call due to the mental processing demands typically associated with agent support systems. Thus, the disclosed method and system may reduce demands on agent mental processing and thereby decrease the amount of time a particular agent takes to complete the mapping and call service process. Reductions in call length would be beneficial and provide cost savings for the support of inbound customer calls.
In addition, the disclosed mapping process utilizes an action-object pair that provides a likelihood estimator to provide an ordered set of possible transactions where the transactions are prioritized from most likely to least likely. This estimation and prioritization may be based on historical, marketing, customer, or other data. In addition, low likelihood transactions may be removed from consideration based on selected thresholds or criteria that are included in an analysis module. The analysis and determination of transaction likelihood may be adjusted as data is collected to improve prediction accuracy.
In a particular embodiment, upon completion of the action object pair analysis, the system may present an agent with an action object pair and a set of transactions ranging from those with the highest likelihood to the lowest likelihood of matching the perceived customer request, including ranking scores of such likelihoods. An example output of a set of transactions with likelihood scores is shown below:
Example Output:
The agent can either accept such transactions or can request further information from the customer to reduce the set of transactions that may be used to satisfy the customer's call request. The agent selections that are made during the process of caller interaction may be captured to provide further input for the estimator to increase the accuracy of future mappings. Accordingly, the disclosed system and method shifts significant processing of filtering and mapping customer opening statements from a human agent to an automated computer analysis module and thereby reduces agent demands and decreases calling time.
The above-disclosed subject matter is to be considered illustrative, and not restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other embodiments, which fall within the true scope of the present invention. Thus, to the maximum extent allowed by law, the scope of the present invention is to be determined by the broadest permissible interpretation of the following claims and their equivalents, and shall not be restricted or limited by the foregoing detailed description.
