MaRGen: Multi-Agent LLM Approach for Self-Directed Market Research and Analysis

2025-08-02

AI Summary

This study addresses the high cost and lengthy turnaround time of traditional manual market analysis by proposing an LLM-driven multi-agent collaborative framework for end-to-end automated generation of market research reports. The system comprises four specialized agents—Researcher, Reviewer, Writer, and Retriever—integrated with in-context learning, structured data querying, visualization generation, and an LLM-based automatic evaluation mechanism, augmented by an iterative refinement process leveraging unstructured expert knowledge. Experiments demonstrate that the system generates six-page, high-quality reports in approximately seven minutes at a cost of ~$1 per report, achieving strong agreement with human expert assessments (Spearman’s ρ = 0.92). Key contributions include: (i) the first closed-loop, business-analytic–oriented multi-agent collaboration paradigm; and (ii) a scalable LLM-based automatic evaluation and feedback enhancement mechanism that significantly improves report credibility and practical utility.

Technology Category

• Multiagent Systems: Teamwork
• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search

Abstract

We present an autonomous framework that leverages Large Language Models (LLMs) to automate end-to-end business analysis and market report generation. At its core, the system employs specialized agents - Researcher, Reviewer, Writer, and Retriever - that collaborate to analyze data and produce comprehensive reports. These agents learn from real professional consultants' presentation materials at Amazon through in-context learning to replicate professional analytical methodologies. The framework executes a multi-step process: querying databases, analyzing data, generating insights, creating visualizations, and composing market reports. We also introduce a novel LLM-based evaluation system for assessing report quality, which shows alignment with expert human evaluations. Building on these evaluations, we implement an iterative improvement mechanism that optimizes report quality through automated review cycles. Experimental results show that report quality can be improved by both automated review cycles and consultants' unstructured knowledge. In experimental validation, our framework generates detailed 6-page reports in 7 minutes at a cost of approximately $1. Our work could be an important step to automatically create affordable market insights.

Problem

• Automating end-to-end business analysis and market report generation
• Replicating professional analytical methodologies using multi-agent LLMs
• Evaluating and iteratively improving report quality autonomously

Innovation

• Multi-agent LLM system automates market analysis
• Agents learn from professional consultants' materials
• LLM-based evaluation improves report quality iteratively