ACD (Automatic Call Assignment) provides users with a wide variety of call center statistics; use ACD statistics to use ACD statistics, which are great beneficial for predictions and arrangements for call center labor management.
ACD provides many details and related statistics, understanding and explanation of this information like trying to drink water from the fireproof faucet. Therefore, when establishing scientific labor management, statistics using the ACD should pay attention to the correct usage.
Four steps
The establishment of the call center labor management mode can be divided into four steps, as shown in the flowchart in the figure.
The prediction call is the first step, which is a method of creating a call center call quantity prediction. This prediction is based on the analysis of historical calls and current call quantities. According to the conclusions of the first step, the calculation of the business representative demand, and the required amount of the agent representative at different time sessions per day. Then, a labor-running table can be made, create a labor working schedule that matches the demand. The final step is to compare the actual results with prediction.
Labor management cycle, ACD statistics work in each step.
ACD statistics play an important role in performing the above four steps. As we will elaborate, use incorrect statistical data, or use statistics in a wrong way, will have a serious negative impact on the quality of the call center management mode.
Predicting call quantity
To accurately predict the future traffic volume of the call center, the call center manager must consider all aspects of information. This information comes from market forecasts, commodity catalogs or number of advertising playbacks. However, some of the most important information comes from historical calls. Usually, the current call amount and call change trend can be predicted from history call. Historical calls must be accurate. Common errors are the amount of call that has been processed as the amount of calls. Please note: The processed call is only a call that the agent represents successful response!
The following example can help deepen understanding. Telegift is ready to use last year's statistics to predict a week of traffic during Christmas this year. Last year, they received 3,000 calls every day, with 8% of them were abandoned. This year they expect the business growth rate of 10%. If they use the number of calls already processed last year: (3000 × 92% = 2760), plus 10% growth rate, they plan to receive 3036 phones. This figure is significantly lower than the correct prediction: 3000 × 110% = 3300 calls. Due to the incorrect statistical number, the number of predicted calls is 264 × 7 = 1848 calls for the Christmas holiday. Telegift's call center manager ignores the abandoned call, because the call is increasing, making this error higher. The prediction call should use the correct statistical data, ie all calls to the call center.
There is also an error that will lead to excessive estimation of dialogue and use multiple queues and call overflows. For example, suppose Telegift has two queues: one is a sales queue, one is a customer service queue. If you overflow from the sales queue to the customer service queue, you will be repeatedly counted by the ACD when the same call arrives at different queues. Thus, a call will be counted twice, resulting in an overestimation of the total call quantity.
Calculate agent representative
First of all, I hope that a misunderstanding will often be corrected here, and there will often be a call center manager to think that "Erlan C is a method of predicting the amount of call", in fact, this concept is wrong. In fact, the correct prediction call is the prerequisite for Erland C.
Here, I hope to remind the call center manager to pay attention to several important ACD statistics. The accuracy of these data will directly affect the accuracy of the agent represents the accuracy of the required amount. The first is to process the time, that is, the agent representative from the beginning to process a call to the average time required to process the next call, this processing time includes all times of the traffic representative to process the traffic, including the time and customer putting the client phone After the end of the post-telephone post-handed time. Pay special attention, don't mistake the conversation time as processing time.
Even if the ACD supports the post-processing time, it should be ensured that the agent represents this feature correctly. If your traffic representative does not operate correctly, it may enable the system to enter the auxiliary working state (or idle state) instead of post-processing status, so that the statistics are distorted. This type of error makes the agent represent the calculation distortion of the required amount. Using the same example, assuming Telegift has received 600 calls per hour, and the company expects 90% of them to get a response within 30 seconds. If Telegift uses only 3 minutes of conversation time without being included in after 20 seconds, then the agent represents the agent representing the agent of Erland C calculated will be 36 people instead of 40 people should need. Arrange 36 people on the post. The company can only reach 63% in 30 seconds, which is much less than 90% target value, which is the result of ignoring the post-processing time. When calculating the agent represents the required amount, it is necessary to manage the management factors, that is, due to the absence of time caused by the agendo or the handle.
Create a tandem table
In this step, the ACD statistics does not work. The basic data used in the discontinuation is: the skills of the agent and its availability, the rules of labor, the time of work, and the intermediate rest time. Handmade interpolation is a very time consuming job. Using labor management tools can help effectively analyze relevant situations and quickly get relevant results.
Actual situation and expected
The most important step in labor management may be learned from the actual facts. In this step, it is necessary to assess the accuracy of each step, appropriately correct, and repeat each step. Use the correct ACD statistics to manage the call center management very helpful. When the operation performance of the call center does not meet the target, it can be identified and predictive is wrong. Even if the call center operates well, it is also possible to use ACD statistics to predict the relevant details so that the operation of the call center will be optimally in a certain resource limit.
