Optimization of bundling and pricing strategies in a supply chain network considering a stochastic consumer purchasing behavior
Date of Publication
Master of Science in Industrial Engineering
Gokongwei College of Engineering
Dennis E. Cruz
Defense Panel Chair
Charlle L. Sy
Defense Panel Member
Richard C. Li
Ronaldo V. Polancos
Alama Maria Jennifer A. Gutierrez
Jonathan R. Dungca
Bundling optimization studies have witnessed a spurt in literature because of the numerous opportunities it can provide both to the company and the customers. Optimization models developed in these studies have extended to a wide variety of considerations such as limited inventory, multi-period horizons, multi-product analyses, and varying consumer purchasing behavior, among others. However, there are two major lapses identified from these models.
First, the said bundling optimization studies are limited to a direct producer-buyer structure. The focus of the studies was on the purchasing side of products only and not on the supply chain operations involved in making the products flow to the end customers. Such consideration will facilitate cost reduction opportunities hence minimize cost while making the bundling initiative effective in maximizing revenue thus maximizing the total profit. With primary focus on the end customers demand through their purchasing behavior and the supply chain operations as the drivers of the decisions, the demand of the end customers will be met thus higher sales can be actualized and, at the same time, the cost of doing so will be minimized. The next lapse is about the absence of a bundling study for multiple product components that adopts an accurate representation of demand. Ineffective demand representation may lead to additional costs brought by shortage/surplus or to lower sales brought by the inability to cater correctly to the demand. Bundling studies represent demand through the consumers purchasing behavior which by its nature is stochastic. No study for multi-product bundling has modeled the end consumers purchasing behavior while preserving its probabilistic nature and this kind of demand representation will facilitate the effectivity needed to accurately capture what the end customers really want.
This study addressed these lapses by proposing a stochastic mixed integer non-linear programming model for a bundling-enabled supply chain system, optimizing both the bundling related decisions (e.g. bundle offering and pricing) and the supply chain-related decisions (e.g. production, inventory, and distribution), thus focusing on the end consumers demand and the supply chain operations involved in making the products available to the customers. The model was validated using the BONMIN solver of Python.
Sensitivity analysis was performed to address the sub-problems of the model and several insights were acquired. These insights include the following: (1) The stochastic reservation price parameters affect the profit and the decisions in different ways. High reservation price means lead to higher purchasing probabilities but this entails a decrease on the probabilities of other options. A high standard deviation and high correlation coefficient leads to a large variance of the reservation prices. Prices must be set at low levels in these cases to still attract customers to purchase. Bundling becomes more profitable when products are negatively correlated. And, bundling should be promoted among products that perform better together so that the firm can capitalize on the higher valuation of the customers for the bundle. (2) Bundling becomes more effective when the firm can reduce the inventory and the transportation cost of the bundle relative to its components. Products that have fixed costs or high setup costs must be sold as bundles to maximize the utilization of the expenses. On the other hand, products that have low variable production costs or high contribution margin ratios must be bundled so the firm can benefit from the high margin of the products. The bundle creation cost analysis also reveals that bundling should be performed in the factory only when the bundle creation cost is cheaper than that of the creation in the warehouse. Also, the possible sales that a bundle can generate must first justify the initial investments required for introduction and setup before it can actually be initiated. (3) Response Surface Methodology reveals that all costs inversely affect the profit but the costs associated with the warehouse (transportation to and from the warehouse, inventory, and bundle formation, altogether) has the greatest impact on the profit. These costs musts be reduced and controlled to further maximize the objective. (4) The analysis of Low Sales as a motivation for bundling reveals that bundling can be effective in boosting sales of a product which would no longer require its production to stop. Bundling can be an effective means of capturing the market as long as the right discount is available for the customers. Also, multi-period considerations allow firms to initially set a high price for products/bundles then eventually mark it down to attract more customers. (5) The hybrid push-pull model outperformed the stand-alone pull and the stand-alone push model in terms of maximizing the profit because of the equal focus on the end customer demand and the cost minimization aspect. (6) Lastly, selling individual products in bundles of 2 or bundles of 3 is more profitable than selling them individually. When these bundles will be offered, the pure bundling strategy becomes more effective as the customers would already prefer to purchase the bundles instead of the individual products given that the bundle discount is large enough
For future studies, it is recommended to extend the scope to consider delays in the supply chain operations, consider rework or reject on the outputs of production and bundling, consider the raw material ordering and replenishment decisions from the supplier, consider competition or the possibility of customers to switch suppliers, consider the decision power of the retailer for pricing and bundling, consider demand as a function of previous sales as well, and lastly, to consider a marketing context for determining which products to advertise to actually increase profit.
Archives, The Learning Commons, 12F Henry Sy Sr. Hall
1 computer disc ; 4 3/4 in.
Industrial engineering; Industrial procurement; Costs; Industrial; Cost effectiveness
Barrios, P. C. (2017). Optimization of bundling and pricing strategies in a supply chain network considering a stochastic consumer purchasing behavior. Retrieved from https://animorepository.dlsu.edu.ph/etd_masteral/5762