Arcor AG

The customer satisfaction get in the telephony and DSL business increasingly important and represents an increasingly important decision criterion for or against a company for the customer. Therefore, any company operating in this segment must ensure that both telephone as well as written questions are answered as soon as possible. At Arcor each day server 10,000 clients requests arrive that are answered at peak times by more than 100 agents. If you consider that a call is possibly an order, an unanswered call may be a measurable financial loss. Because not to the satisfaction of the customers answered support requests may result in the loss of customers, thus answering all calls in any case constitutes a financial advantage.

The optimal planning and control of a call center therefore is for every major company for the entire financial success of great importance. However, the operation of a call center is subject to a number of uncertainties that complicate the planning and control process:

  • The clients arrive at random times in the system. This makes it difficult to determine the necessary number of agents. In particular, it can be shown that a permanent full utilization of the agents is incompatible with short waiting times for the clients.
  • The clients only have a limited waiting time tolerance and will leave the system if they have to wait too long. In this case they usually start another try after a certain time and in this case raise the load in the already overloaded system. This effect makes it difficult to determine the optimum number of additional agents.
  • The actual call centers are usually operated by sub contractors, so that only a limited influence on the specific planned manpower is possible. In addition, the optimization goals of sub constractors differ from the optimization goals of the actual company.
  • Workforce management is subject to certain shift plan specifications, so that the number of agent can not be increased at short notice during peak loads, but only can be raised for only a longer time period.

The complete call center system at Arcor including retrying clients, forwarding, different clien types and multiple external call centers was modeled as a queuing system in form of a simulation and optimization tool to check all possible configurations without needing to change the real call center network.

As a first step various alternatives of the call routing to the various external call centers were investigated using the simulator. As next issues the impact of retrying cleints, the skill distribution of the agent and the combination of telephone conversations and correspondence processing were examined.


Sitemap  Contact  Data Privacy  Imprint
© TU Clausthal 2023