Analysing service level agreements with multiple customers.

Alamri, O 2017, Analysing service level agreements with multiple customers., Doctor of Philosophy (PhD), Science, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Alamri.pdf Thesis Click to show the corresponding preview/stream application/pdf; Bytes
Title Analysing service level agreements with multiple customers.
Author(s) Alamri, O
Year 2017
Abstract Within numerous production and distribution environments, maintenance of effective customer service is central to securing competitive benefits. Globalised industries are becoming more commonplace as well, further increasing the competitive pressure. Companies, as a result, are forced to expand product availability and deliver to the demand on schedule. As part of a supply chain, service levels are an important measure of performance in operations management and are widely used to evaluate and manage supplier performance. This thesis examines the SLA for the supplier under two types of contracts to guarantee the agreed customer service level. Specifically, this dissertation will shed light on the two most important (SLA) measurements for inventory systems: fill rate and ready rate. Both SLA measurements are commonly used as performance measures in SLAs between customers and suppliers. Throughout this thesis, we examine performance-based contracts in which the supplier has either: a single customer with a large demand, or multiple customers with a smaller demand. Our experiments were designed so that the demand distribution for the single customer case was similar to the aggregated demand distribution in the multiple customer case.

The thesis primarily focused on four main questions, with each question being examined in its own chapter. The first research problem is addressed in Chapter 3. Earlier studies of finite horizon fill rate only consider the situation in which there is a single customer in the supply chain. In Chapter 3, we develop a model to analyse the fill rate distributions for a supplier that has multiple customers, each with its own SLA. In particular, we examine the impacts of performance review period length and the correlation between customer demands on the average fill rate and the probability of overreaching the target fill rate when a supplier has multiple customers. Under the multiple customer contracts, two service policies for demand fulfilment. In the first policy, First-Come-First-Served (FCFS), demand is filled with no prioritization (e.g., in the case of two customers, there is a 50% chance that the first customer is served first). In the second policy, Prioritized Lowest Fill Rate (PLFR), customers are prioritized so that the customer with the highest negative deviation from its target fill rate in the current performance review period is served first. The results and findings in Chapter 3 provide insights that can assist suppliers in the design and negotiation of SLAs. The second research problem is addressed in Chapter 4. Previous studies on the finite horizon fill rate are limited and assume a zero lead time for the supplier. We create a model to examine the impact of different supplier lead times on the finite horizon fill rate, considering either single customer or multiple customers. As lead time exists in reallife supply chains, we explore the effect of various lead times on the fill rate distribution and required base stock over finite horizons with a variety of review period lengths.

The results revealed that to fix the long-run fill rate, as the lead time increases, more stock is required; however, the probability of exceeding the target fill rate (the probability of success) increases as the lead time increases. The results indicate that the increase in the probability of success as the lead time increases is higher when the review period is shorter. For the third research problem Chapter 5 presents further results related to the fill rate, an important measure of supply chain performance, specifically ensuring that a customer’s service need is met with maximum reliability. These results mainly concentrate on variability, an aspect that is largely ignored in the literature on fill rate. Related results concerning consistency and asymptotic normality extend the range of application of the fill rate in evaluating reliability and determining the optimal stock level of a supply chain. Chapter 6 explores the fourth research problem which considers the ready rate, a widely used performance measure in SLAs. The ready rate considered in this study is defined as the long-run fraction of periods in which all customer demand is filled immediately from on-hand stock. Previous studies of SLAs have been solely concerned with one supplier serving one customer, whereas in practice, a supplier usually deals with more than one customer. In multiple customer cases, the supplier has an SLA with each customer, and a penalty is incurred whenever the agreement is violated. In this chapter, we create a model to examine the impacts of various factors such as the base-stock level, the type of penalty (lump-sum and linear penalty), and the review period duration on the supplier’s cost function when the supplier deals with multiple customers. The results show that dealing with more customers is preferable for a supplier (assuming the overall demand is the same) and that under a lump-sum penalty contract, a longer performance review is beneficial. Finally, Chapter 7 closes with a brief review, discussion on the models constructed and suggests areas for future studies.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Science
Subjects Applied Statistics
Applied Mathematics not elsewhere classified
Operations Research
Keyword(s) Service level agreement
Fill rate
Ready rate
Multiple customers
Versions
Version Filter Type
Access Statistics: 157 Abstract Views, 106 File Downloads  -  Detailed Statistics
Created: Fri, 27 Oct 2017, 09:51:41 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us