Introducing Make-to-stock to Make-to-order Platforms

Abstract

Many popular apparel print-on-demand (POD) services, such as Merch by Amazon, use the POD approach to meet scattered demand for a large variety of products on their platform. The POD approach is a form of the make-to-order (MTO) production mode; its flexibility in handling variety provides a quick response to uncertain customer demands, but its unit production cost is higher than the batch-producing make-to-stock (MTS) production mode. The business model for POD platforms is as follows. The platform licenses product designs from content creators and sells them on the website. These designs all use the same basic blanks (t-shirts, hoodies, tank tops, etc.). Once an order is received, the platform prints the requested design on site and ships the item(s) to the customer. The platform also handles all logistical issues, such as payment. The content creator receives a certain portion of each sale. The content creator is responsible only for making product decisions: managing portfolio size, marketing, pricing and so on. Compared to MTO, the traditional MTS production mode has advantages from a cost perspective. MTS production is cheaper per unit but is less flexible, and there is a fixed fee associated with each order. Its effective use requires demand to be sufficiently high and less variable. In this dissertation, we study whether MTS can be integrated into POD platforms and, if so, how to do so most effectively. In our analysis, we find that in the real-life decentralized system where the platform does not control what the content creators do, content creators will take actions that are optimal for themselves, but the result is that product demands are not “good” enough for the platform to use MTS. To address this issue, we take actions in two areas: product assortment and effort allocation. They are presented in Chapter 3 and Chapter 4, respectively. In Chapter 3, we study how the platform can limit the number of designs by a content creator so that demand can be more concentrated in each product, making MTS effective and improving profit at the same time. We use a Salop's circular model to reflect the underlying competition among products and to derive the relationship between the number of products and individual product demand. We propose a hybrid production mode and find conditions under which the hybrid mode can offer a lower overall cost than the MTO production mode. To implement this approach in a decentralized setting where the platform can induce but not control the actions of the content creators, however, we find that the platform must overcome an incentive misalignment with independent content creators. This misalignment occurs because the content creator's profit is determined only by sales, as is the prevalent practice on such platforms nowadays; consequently, content creators will offer as many designs as possible on the platform to maximize sales. However, the platform prefers to have a limited number of designs so that each design will have enough demand to employ hybrid production. To overcome this misalignment, we first analyze a centralized system to set a target for improvement and propose that the platform incorporate a listing fee for each item. We show that this simple addition can effectively align the content creator with the platform and improve profitability for both. Numerical studies demonstrate the potential profit improvement and identify conditions under which the approach using hybrid production and a listing fee can yield significant profit improvement. In particular, we find when the profit margin on each product is lower or the unit production cost saving is significant, the hybrid production mode can have the most positive impact on system profit. Inspired by the success in profit improvement achieved by the hybrid production mode, we study in Chapter 4 another operational aspect that can impact the addition of MTS to MTO; namely, how content creators allocate effort when adopting the hybrid production mode. Content creators can exert effort in various ways, including polishing their designs and marketing. Content creators promise to invest a fixed amount of effort for their products to be listed on the platform. With diminishing returns of the effort's effect on demand, we find that content creators prefer to spread their efforts across all products in their portfolio. When a content creator's amount of effort is low, this can cause the demand for an individual product to be too low for the platform to use the hybrid production mode. To address this inefficiency, we propose that the platform adopt a coordinating strategy for the hybrid mode that rewards the products with higher demand with a higher portion of the profits. We show that this strategy can effectively encourage content creators to concentrate their effort on fewer products so that individual demands for these products are high enough to justify hybrid production. Compared to the approach of limiting the number of designs in Chapter 3, here in Chapter 4 we argue that the coordinating strategy is less restrictive on total demand because non-promoted products remain in the portfolio, generating demand that can be satisfied using MTO production mode. In Chapter 4, we explore another strategy to enhance the implementation of the hybrid production mode by proposing that the platform exert additional effort to increase demand. These efforts from the platform can take the form of enhancing customer service, promotions for the platform, sharing data with content creators on trending designs, and so on. We first find the optimal amount of platform's effort; we then show that, with a small enough unit effort cost for the platform, a cooperative effort strategy can enable the hybrid production mode to achieve higher system profit improvement than the coordinated hybrid mode and can eliminate the system's incentive misalignment regarding allocation of effort. Numerical studies further quantify the magnitude of potential profit improvement achieved by both the coordinated hybrid production mode and the cooperative effort strategy and demonstrate that even when the platform exerts less than the optimal amount of effort, system profit can still benefit from the additional effort that the platform does exert.