Chat Your Way to Better Shelves: InteraSSort Revolutionizes Retail Assortment Planning

Table of Links

Abstract and 1 Introduction

2. Related Work

3. Preliminaries

4. The InteraSSort Framework

5. Illustration

6. Conclusion and References

Abstract

Assortment planning, integral to multiple commercial offerings, is a key problem studied in e-commerce and retail settings. Numerous variants of the problem along with their integration into business solutions have been thoroughly investigated in the existing literature. However, the nuanced complexities of in-store planning and a lack of optimization proficiency among store planners with strong domain expertise remain largely overlooked. These challenges frequently necessitate collaborative efforts with multiple stakeholders which often lead to prolonged decision-making processes and significant delays. To mitigate these challenges and capitalize on the advancements of Large Language Models (LLMs), we propose an interactive assortment planning framework, InteraSSort that augments LLMs with optimization tools to assist store planners in making decisions through interactive conversations. Specifically, we develop a solution featuring a user-friendly interface that enables users to express their optimization objectives as input text prompts to InteraSSort and receive tailored optimized solutions as output. Our framework extends beyond basic functionality by enabling the inclusion of additional constraints through interactive conversation, facilitating precise and highly customized decision-making. Extensive experiments demonstrate the effectiveness of our framework and potential extensions to a broad range of operations management challenges.

1. Introduction

Assortment planning (K¨ok et al., 2015; Rossi & Allenby, 2003) is a pivotal marketing strategy employed by managers/store planners in the retail industry. These planners are responsible for designing the layout and product assortments for physical retail stores to maximize sales, customer satisfaction, and profitability. These seasoned planners, with their deep domain knowledge, often need to generate insights for questions that involve variations of the assortment optimization problem. However, due to the complexity inherent in store planning, as well as the absence of optimization expertise among store planners, significant challenges often arise. The process of insight generation as a result requires collaboration with multiple professionals, resulting in prolonged decision-making processes and significant delays. Consequently, there is a clear demand for a framework designed to assist store planners as shown in Figure 1. This framework should be able to provide dynamic solutions to various assortment planning problems, all while eliminating the necessity for a detailed understanding of technical optimization framework, thereby assisting in the decision-making process.

The advent of artificial intelligence (AI) has brought about a revolutionary transformation in the way businesses operate. Among these cutting-edge innovations, large language models (LLMs) such as GPT-4 OpenAI (2023) and LLaMA Touvron et al. (2023) have emerged as pioneers of generative AI, leading the forefront of the latest technological disruptions. However, it is only in recent years that the intersection of AI and marketing has captured the attention of researchers. This has prompted further investigations into AI-related topics and their roles in marketing Jain et al. (2023). In light of this, LLMs with their advanced capabilities can serve as a fundamental component in creating an interactive framework tailored for solving marketing challenges, such as assortment optimization, thus assisting the store planners in making informed decisions.

In addition to integrating LLMs, a significant challenge lies in selecting suitable assortment optimization algorithms capable of delivering swift and scalable solutions, keeping the aspect of interactivity at the forefront. Addressing these challenges, we propose a collaborative framework InteraSSort that effectively integrates LLMs with optimization tools to tackle the assortment planning problem in an interactive manner. InteraSSort enables planners to present their optimization objectives using natural language through input prompts, and the framework will respond by making appropriate calls to optimization tools and solvers. Our approach goes beyond basic functionality by incorporating the ability to include additional constraints through text prompts and generate solutions interactively. We summarize the list of our contributions below.

• We design InteraSSort to feature a user-centric chat interface via Streamlit [1], with LLMs and optimization algorithms seamlessly integrated into the backend to carry out the tasks based on the input prompts provided by the user.

• Our framework leverages the conversational history and function-calling capability of LLMs to accurately invoke the requisite functions in response to input prompts, facilitating the execution of optimization scripts to deliver solutions to the user.

[1] https://streamlit.io/