Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems

2026-06-15

AI Summary

This study investigates brand competition dynamics in large language model (LLM)-based recommendation systems, revealing a “conditional monopoly” wherein well-known brands receive 100% recommendation shares (IAI = 10.0) under otherwise equal conditions. Through controlled experiments on GPT-4o-mini, Claude Sonnet, and Gemini 3 Flash—validated with search goods—the research demonstrates that a mere +0.1-star rating advantage suffices to disrupt this monopoly. Authoritative marketing language induces a bias equivalent to a +0.17-star boost. Furthermore, strategic interactions among multiple brands employing generative engine optimization (GEO) lead to a dramatic decline in individual gains—from +0.802 to +0.007—highlighting a social dilemma. This work is the first to position GEO as an emerging marketing practice shaping market competition and quantifies structural biases inherent in LLM-driven recommendations.

Abstract

Large language models (LLMs) are becoming a major way for consumers to find products, but we do not yet understand how brands compete in this new channel. We study brand dynamics in LLM recommendations using skincare products -- a category where consumers cannot easily judge quality before buying and must rely on brand reputation -- across three commercial LLMs (GPT-4o-mini, Claude Sonnet, Gemini 3 Flash), with a robustness check on search goods. In three experiments, we find: (1) a Conditional Monopoly where well-known brands get recommended 100% of the time (IAI = 10.0) when all products have the same specifications, but this dominance disappears with less than a +0.1-star rating advantage for a competitor; (2) authority-style marketing language, including fabricated clinical-evidence claims, breaks this monopoly at a Bias Surplus Value equal to +0.17 rating points, with each model responding differently; and (3) a social dilemma in multi-brand GEO competition: when all brands adopt the same optimization strategy, individual payoff falls from +0.802 to +0.007 in our payoff proxy, and non-participating brands receive zero recommendations in our tests. Our results suggest that generative engine optimization (GEO) should be studied not only as a security risk, but also as an emerging marketing practice that shapes market competition.

Problem

• brand bias
• cognitive manipulation
• LLM recommendation systems
• generative engine optimization
• market competition

Innovation

• Generative Engine Optimization
• Brand Bias
• Conditional Monopoly
• Cognitive Manipulation
• LLM Recommendation Systems