Model to Determine Bank Teller Requirements and Predict Transactions Case for: Banking Industry
Customer Satisfaction is paramount to any business or industry; this is particularly felt in the Banking industry especially in Kenya where there is fierce competition among the players. The banks have conventionally been associated with queues that anytime one has to visit the banking hall, the thought of long wait in the queue deter them away. This has had a major impact on Customer Service. Simulation was applied to model the current scenario and estimate performance metrics. Some scenarios were considered to find out how the existing system operated. Resource utility and customer waiting time were used to evaluate the performance of the queuing system. A mathematical model based on mathematical theory of queues, Little's result, theorem, lemma, law or formula, expressed algebraically as: L = λ W was developed in-line with the bank‟s standard on customer waiting times. The teller staffing model was tested using ARENA simulation software and the result was a reduction of 60% in customer waiting time. The model provided a dynamic schedule solution that allocated tellers based on the customer arrivals/demand. DTREG was used to model the transaction dataand build the decision trees that were used to predict workload/transactions for any given working day of the week. Performance of queuing systems is dictated by the Input Source, Service System and the Queuing Discipline. The Input Source is categorized as Static or Dynamic. Most queuing systems are founded on Static arrival process where the probability of arrivals is described as the number of customers arriving per unit of time. This study however proposed a Dynamic approach where the Service Rate was determined by both the service facility (Tellers) as well as the arrivals (Customers). The service facility adjusted its capacity to match the changes in demand intensity by varying the staffing levels at different timings of the day.