Adventures in Demand Analysis Using AI

2024-12-31

AI Summary

Accurately estimating causal effects—particularly price elasticity—in product demand analysis remains challenging due to complex, nonlinear relationships among multimodal product attributes (text, images, structured data), price, and sales. Method: We propose the first framework integrating fine-grained multimodal large language model (MLLM) embeddings into a causal inference pipeline via Double Machine Learning (DML), enabling interpretable, heterogeneous price elasticity estimation across modalities. Leveraging Transformer-based architectures and cross-modal fusion, our approach jointly models product features, price, and sales dynamics. Contribution/Results: This work pioneers the systematic incorporation of fine-tuned multimodal embeddings into causal inference, revealing and quantifying micro-level price elasticity heterogeneity across products. Empirically, it significantly improves sales ranking accuracy and price prediction fidelity, yielding more robust, product-specific elasticity estimates. Our method establishes a new paradigm for data-driven pricing and product design grounded in interpretable, multimodal causal modeling.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: ML for personalized search and recommendations

Abstract

This paper advances empirical demand analysis by integrating multimodal product representations derived from artificial intelligence (AI). Using a detailed dataset of toy cars on extit{Amazon.com}, we combine text descriptions, images, and tabular covariates to represent each product using transformer-based embedding models. These embeddings capture nuanced attributes, such as quality, branding, and visual characteristics, that traditional methods often struggle to summarize. Moreover, we fine-tune these embeddings for causal inference tasks. We show that the resulting embeddings substantially improve the predictive accuracy of sales ranks and prices and that they lead to more credible causal estimates of price elasticity. Notably, we uncover strong heterogeneity in price elasticity driven by these product-specific features. Our findings illustrate that AI-driven representations can enrich and modernize empirical demand analysis. The insights generated may also prove valuable for applied causal inference more broadly.

Problem

• Demand Analysis
• Product Pricing
• Sales Prediction

Innovation

• Artificial Intelligence
• Demand Analysis
• Causal Inference