Operation Management in the Context of a Bank Call Centre - Essay Example

A look into the perspective of Operation Management in the context of a Bank Call Centre The intent of thispaper is to clearly convey the actual situation of operational management in the field of Banking Call centers. For this purpose the chosen bank in this case is the UTI Bank or Unit Trust of India Bank. The reason of choosing India lies in the perception that the BPO industry of India is regarded as a highly booming market where telemarketing or call center business occupies a bulk, about 70% of the outsourcing industry. This is the reason this country has turned the subject of Operation Management into a dynamic science. Now the cause of choosing UTI as the target bank is somewhat different. UTI is not the leading bank of India and neither is it the most elite. Then why UTI This is because UTI can boast to possess one of the most sophisticated tele calling organization of the country. One major characteristic of rationalist research is the belief that the phenomenon being studied exists 'out there', independent of the research context or beliefs and assumptions of the researcher. Thus, the relationships and observations are considered to be independent of the theories used to explain them and can hence be studied, manipulated at will, and controlled as needed by the researcher. Another major characteristic of rationalist research is the goal of determining the distributions of a set of pre-specified variables in the population or verifying a set of pre-specified relationships. The main focal pint is to primarily considering the rationalist methods of modeling by equations, laboratory experiments, and statistical survey analysis in making comparisons to case research. Note that the division here does not parallel a frequently-used one of dividing operations management research between so-called 'theoretical' a misnomer, since all these paradigms can be theoretical and 'empirical' methods. Typical equation-modeling research would include variants of the economic order quantity formulation as well as production-inventory system simulations. Statistical survey research is currently being used in quality management, supply chain, and technology management investigations, among many others. Rationalist research methods also have their draw-backs. First, obtaining valid empirical generalizations depends to a large extent on the use of sampling procedures that are rigorous, representative of a well-specified population, and provides a source of information concerning the constructs to be measured. Yet, despite the importance of employing rigorous sampling criteria for generalizability of the findings, the great majority of quantitative studies based on sampling appear to use samples of convenience or opportunity. A call center is a service network in which agents provide telephone-based services. Customers that seek these services are delayed in tele-queues. The attempt summarizes an analysis of a unique operation management record of call center operations. The data comprise a complete operational history of a small banking call center such as UTI, call by call, over a full year. Telephone call centers allow groups of agents to serve customers remotely, via the telephone. They have become a primary contact point between customers and their service providers and, as such, play an increasingly significant role in more developed economies. For example, it is estimated that call centers handle more than 78% of all business interactions and that they employ more than 2.5 million people. While call centers are technology-intensive operations, often 70% or more of their operating costs are devoted to human resources, and to minimize costs their managers carefully track and seek to maximize agent utilization. Well-run call centers adhere to a sharply-defined balance between agent efficiency and service quality, and to do so, they use queuing-theoretic models. The UTI center provides several types of services: information for current and prospective customers, transactions for checking and savings accounts, stock trading, and technical support for users of the bank's internet site. On weekdays the center is open from 7am to midnight; it closes at 2pm on Friday, for the weekend, and reopens around 8pm on Saturday. During working hours, at most 13 regular agents, 5 internet agents, and one shift supervisor may be working. A simplified description of the path each call follows through the center is as follows. A customer calls one of several telephone numbers associated with the call center, the number depending on the type of service sought. Except for rare busy signals, the customer is then connected to a VRU and identifies herself. While using the VRU, the customer receives recorded information, general and customized (e.g. an account balance). It is also possible for the customer to perform some self-service transactions here, and 65% of the bank's customers actually complete their service via the VRU. The other 35% indicate the need to speak with an agent. If there is an agent free who is capable of performing the desired service, the customer and the agent are matched to start service immediately. Otherwise the customer joins the tele-queue. Customers in the tele-queue are nominally served on a first-come first-served (FCFS) basis, and customers' places in queue are distinguished by the time at which they arrive to the queue. In practice, the call center operates a priority system with two priorities - high and low - and moves high priority customers up in queue by subtracting 1.5 minutes from their actual arrival times. Mandelbaum et al. (2000) compares differences between the behaviors of the two groups of customers. While waiting, each customer periodically receives information on her progress in the queue. More specifically, she is told the amount of time that the first in queue has been waiting, as well as her approximate location in the queue. The announcement is replayed every 60 seconds or so, with music, news, or commercials intertwined. Each call that crosses the dashed line can be thought of as passing through up to three stages, each of which generates distinct data. The first is the arrival stage, which is triggered by the call's exit from the VRU and generates a record of an arrival time. If no appropriate server is available, then the call enters the queuing stage. Three pieces of data are recorded for each call that queues: the time it entered the queue; the time it exited the queue; and the manner in which it exited the queue, by being served or abandoning. The last stage is service, and data that are recorded are the starting and ending times of the service. Note that calls that are served immediately skip the queuing stage, and calls that abandon never enter the service stage. In addition to these time stamps, each call record in our database includes a categorical description of the type of service requested. The main call types are Regular (PS in the database), Stock Transaction (NE), New Customer (NW), and Internet Assistance (IN). (Two other types of call -Service in English (PE) and Outgoing Call (TT) - exist. Together, they accounted for less than 2% of the calls in the database.) The UTI call center, even given its small size, the original data set included more than 1,200,000 calls, roughly 450,000 of which were from customers who wished to speak with an agent. The call-by-call data could be well used to characterize queuing primitives, such as the arrival process (as inhomogeneous Poisson with additional randomness in its arrival rate), the service-time distribution (as lognormal), and the distribution of customer impatience. These building blocks could be used to develop additional tools that are useful in call-center management: theoretical and empirical patience indices, as well as prediction intervals for the offered load in the context of operation management. Even the could test the robustness of several queuing-theoretic results. It could be summarized that a simple multi-server generalization of the classical Khintchine-Pollaczek formula produced biased waiting-time predictions. In contrast, queuing results concerning the (linear) relationship between average waiting times and abandonment rates, as well as predictions derived from the Erlang A (M/M/N+M) model, could prove to be surprisingly robust. Specifically, the characterization of the service-time distribution gave rise to new, nonparametric methods for estimating regression models with lognormal errors. The large volume of highly censored abandonment data motivated us to develop nonparametric methods of estimating and graphing associated hazard-rate distributions. There are too many hazards in the field of tele calling and UTI experience quite a few of them. But studying the field of operation and analyzing the outputs would reduce the problems subsistent rate. The main point is to use the modes of operational management properly. Read More