KR20050045568A - Method for assigning agents in call center system - Google Patents

Abstract

In the call center system, there is provided a method of estimating the number of counselors required for a specific date from the counseling call statistics data for each date and hour, and assigning the number of persons who can achieve a given average service level by using the data. Calculate the number of people needed each time period to meet the target service level on a specific date. It is assumed that the maximum number of counselors to be inputted among the calculated time slots is calculated. When the average service level is higher than the target service level, the average service level is calculated by assuming that the number of agents inputted from the assumed number of agents is reduced by a predetermined number. Repeat from step to do, otherwise assign the assumed number of counselors to the number of counselors on the day. When calculating the average value, it is desirable to calculate the average by weighting the time period when the calculated service level is lower than the target service level.

Description

How to assign agent staff in a call center system {Method for assigning agents in call center system}

The present invention relates to a method for assigning an agent to a call center system. More specifically, the call center system calculates the required number of counselors per hour of a specific date from the statistics data for each date and time, and uses the same to calculate a given average service level. It is about how to allocate the number of people that can be achieved.

The call center system is a system for efficiently responding to a large number of calls from customers such as a telephone service company, a department store, an after-sales service center, a courier company, and the like, and the configuration is generally as shown in FIG. 1.

The call center system 100 uses a private exchange (PBX) 110, which is an on-site telephone exchange equipment, an interactive voice response (IVR) server 120 that provides a predetermined voice guidance message to a called customer, and a computer. CTI server 130 that provides telephone exchange service, a plurality of agent computers (140: 140 ₁ ~ 140 _n ), information related to the customer (for example, customer ID, password, personal information of the customer), call center operation and DB server 160 storing various related information, call center management server 170 which is in charge of call center management tasks such as call volume prediction, agent placement, service level control of call center system is connected to each other through Ethernet (E1), for example. It is composed.

In addition, the private exchange 110 of FIG. 1 is connected to a plurality of counseling telephones 150: 150 ₁ to 150 _n and the IVR server 120 through a plurality of telephone lines L1, respectively, so that the IVR is connected to a customer. The voice guidance message output from the server 120 is transmitted to the corresponding customer, and is connected to the Ethernet E1 through the well-known CTI link 111, and is controlled by the plurality of agent phones 150 under the control of the CTI server 130. It will selectively connect the channels between the customers.

According to the above configuration, when there is a call connection request of a certain customer from the public telephone network, the private exchange 110 connects the corresponding customer and notifies the IVR server 120 of the call connection state of the customer. The IVR server 120 transmits a predetermined voice announcement message to the private exchange 110, and the private exchange 110 sends the voice announcement message to the corresponding customer. In addition, the IVR server 120 collects predetermined customer information such as a customer ID and a password input by the customer according to the voice guidance message, transmits the predetermined customer information to the CTI server 130, and communicates with an agent available to the CTI server 130. You will be asked.

Thereafter, the CTI server 130 of FIG. 1 checks the call connection state of each agent phone 150 currently connected to the private exchange 110 through the CTI link 111, and the agent phone 150 of the agent currently waiting for the call. It connects the channel between the call and the customer call. In addition, the CTI server 130 extracts the cumulative counseling history and personal information of the customer from the DB server 160 based on the input customer information and transmits it to the agent computer 140 of the connected agent, the agent is the customer Consultation history and personal information of the will proceed smoothly consultation. On the other hand, the statistical data related to such a telephone consultation is collected in the call center management server 170 is used for the prediction of the call volume on a specific day, a specific month and the arrangement of counselors accordingly.

However, in the conventional call center system, as shown in FIG. 2, the number of counselors required for each working period of a specific date is estimated and the counselors are arranged accordingly. However, as can be seen in Figure 2, since the number of counselors required for each time zone shows a big difference, it is difficult to arrange the counselors. In other words, if an agent is placed in a time zone where a lot of agents are needed, the waste of labor costs is incurred because the agent remains in other time zones. There is a problem that it cannot be satisfied.

Therefore, in the conventional system, this problem is solved by hiring a part-time counselor during a time when a large number of counselors are needed. In other words, during peak hours of high consultation, part-time counselors are added to reduce labor costs while matching service levels.

However, under the employment situation in Korea, it is not only easy to hire part-time counselors, but also, part-time agents often fall short of regular workers in terms of skill, education, and qualifications, which can cause problems in their relationship with customers. There is plenty of room.

The present invention has been made in view of this point, and it is an object of the present invention to provide a method for assigning a counselor who does not need to allocate a person for each time zone while maintaining an average service level.

In the present invention, as in the prior art, the number of people required for each time slot is calculated to meet the target service level on a specific date. In addition, it is assumed that the maximum number of counselors to input the calculated number of personnel per time zone, and calculate the average service level of the date when the input. When the average service level is higher than the target service level, the average service level is calculated by assuming that the number of agents inputted from the assumed number of agents is reduced by a predetermined number. Repeat from that step. Otherwise, the assumed number of counselors is assigned to the number of counselors on the day. When calculating the average value, it is desirable to calculate the average by weighting the time period when the calculated service level is lower than the target service level. On the other hand, when calculating the average service level for the first time, the number of people to be substituted may be the minimum number instead of the maximum number. In addition, if the calculated number of counselors exceeds the maximum available number, it is also possible to repeat these steps after modifying the target service level.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

In order to calculate the number of counselors needed, it is necessary to estimate the service level according to the number of counselors. Service level refers to what percentage of calls are processed within a defined number of seconds. Prediction of service level predicts that a certain level of service will be possible if there is a certain number of agents from the statistical data and predicted events, or vice versa, to meet a certain level of service. It is predicting.

As a method of estimating the service level or a method of estimating the number of people to satisfy a given service level, many methods are proposed as in US Pat. No. 5,185,780, US Pat. No. 6,226,377, US Pat. No. 6,556,659 and the like. It is not limited to the method.

First, the required number of hours per hour for the target service level is calculated using the conventional technology (S310). That is, the estimated number of persons as shown in FIG. 2 is calculated as in the prior art. At this time, it is also possible to calculate the required number of people by varying the target service level for each time zone.

Next, it is assumed that the number of counselors to input the maximum number of personnel from the calculated time slots (S320). For example, in the case of FIG. 2, since 70 people are the most people, it is assumed that the number of counselors to input 70 people.

Next, the average service level is calculated according to the number of counselors to be put (S330). That is, after calculating the service level when the number of counselors assumed for each time slot is input, the average value of the counselors is calculated.

On the other hand, when the calculated service level is lower than the target service level, there are more customers waiting for the call with the counselor and giving up the call. By the way, some of these calls are made again, so the lower the service level, the more calls are expected than the number of incoming calls expected. Therefore, there is a high probability that the service level will be lower than the simple prediction.

In view of this, it is desirable to give a weight to the time period when the calculated service level is lower than the target service level when calculating the average service level. For example, the estimated timephased service level

90% 80% 70% 60% 60% 70% 75% 75% 80% 90%

, Then the average simple service level is

(90 + 80 + 70 + 60 + 60 + 70 + 75 + 75 + 80 + 90) / 10 = 75%

Becomes

However, assuming that the target service level is 75%, a weight of 1.5 times for a time zone with a lower service level is given.

(90 + 80 + 75 + 75 + 80 + 90 + (70 + 60 + 60 + 70) x1.5) / 12 = 73.3%

Becomes

Therefore, it can be seen that the average service level is lower in the case of giving a weight. That is, by weighting, the lower the service level is, the lower the service level, the more incoming calls than the expected number of incoming calls and the actual service level lower than the expected service level is well reflected.

The average service level calculated in this way and the target service level is compared (S340). In step S310, when the target service level is set differently for each time zone and the required number of people is calculated, the average service level may be compared with the average value of the target service level for each time zone.

If the calculated average service level is higher than the target service level, it is assumed that the number of counselors reduced by a predetermined number from the assumed number of counselors is the number of counselors to input (S350) and the process is repeated from step S330. If you reduce the number of agents at once, the calculation will be faster but the accuracy will be lower. If you reduce the number of agents at the same time, the calculation will be slower but the accuracy will be higher, so the number of agents should be appropriately adjusted in consideration of factors such as the total number of agents and the accuracy required. Select.

On the other hand, if the calculated average service level is not higher than the target service level, the number of counselors used when calculating the average service level is assigned as the number of counselors on the day (S360). However, when the number of agents at this time is input, since the predicted average service level is always equal to or lower than the target service level, when the average service level higher than the target service level is desired, the predetermined number is determined from the number of agents when the step S360 is reached. It is also possible to assign an increased number.

Meanwhile, in the above description, when calculating the average service level, the number of agents is calculated by counting the number of agents by the maximum number of agents per time zone, and the number of agents corresponding to the target service level is reduced. It is also possible to calculate the number of counselors to meet the target service level while increasing the number of counselors by calculating the minimum number of counselors.

Although the present invention has been described in detail with reference to some examples, the present invention is not limited to the specific embodiments. Those skilled in the art will appreciate that many modifications and variations can be made without departing from the spirit of the invention.

According to the present invention, since the number of people who can match the target average service level is calculated from the predicted number of people at each time slot, there is no need to allocate a person at each time slot. In addition, since the average service level is calculated by weighting the time period when the service level lower than the target service level is predicted, it is possible to prevent the expected service level from the difference between the actual service level due to the decrease in the service level.

1 is a block diagram showing the configuration of a general call center system.

2 is a view showing an example of assigning the number of counselors in a conventional call center system.

3 is a flowchart illustrating a procedure of a method for allocating counselors according to the present invention.

Claims (6)

The first step of calculating the number of people required for each time zone on a specific date to meet the target service level, A second step of assuming that the number of counselors to input the maximum number of persons from the calculated time slots, A third step of calculating an average service level for the day when the assumed number of agents is input; A fourth step of comparing the calculated average service level with a target service level, If the average service level is higher than the target service level, the number of agents reduced from the assumed number of agents is assumed to be the number of agents to be input, and the procedure is repeated from step 3. Otherwise, the assumed number of agents is the number of agents on the day. The fifth step to assign Agent staffing method in a call center system having a. The first step of calculating the number of people required for each time zone on a specific date to meet the target service level, A second step of assuming that the number of counselors to input the minimum number of persons from the calculated time slots, A third step of calculating an average service level for the day when the assumed number of agents is input; A fourth step of comparing the calculated average service level with a target service level, If the average service level is lower than the target service level, the number of agents increased from the assumed number of agents is assumed to be the number of agents to be input, and the procedure is repeated from step 3. Otherwise, the number of agents that are assumed is the number of agents on the day. The fifth step to assign Agent staffing method in a call center system having a. The method according to claim 1 or 2, A sixth step of requesting to modify the target service level if the number of counselors calculated in the fifth step exceeds the maximum available number; Step 7 of repeating step 1 for the modified target service level Agent number assignment method in a call center system further comprising. The method according to claim 1 or 2, The target service level in the first stage can be specified for each time unit, and the target service level in the fourth and fifth stages is an average value of the target service level designated for each unit of time. How are people assigned? The method of claim 1 or 2, wherein the third step Steps 3-1 to calculate the service level when inputting the number of agents assumed for each time zone, Step 3-2 to find the average value of the calculated service level Agent personnel assignment method in a call center system comprising a. The method of claim 5, When the average value is calculated in step 3-2, the number of counselors assigned to the call center system is calculated by weighting the average for the time period when the calculated service level is lower than the target service level.