AI travel agents can search near-infinitely without getting tired — so what happens when 'just check one more fare' costs them nothing?
Why do AI agents generate vastly more searches than human travelers?
This explores why software agents that plan or book travel fire off far more search queries per booking than a person would, and what that does to the systems answering them.
This explores why AI agents that plan or book trips send so many more search queries than human travelers, and what follows from that. The retrieved notes don't directly measure agent query volume in travel, so that number isn't settled here. They do show the forces behind it, and those forces all point the same way: for an agent, searching is close to free, and it is the agent's main way of reasoning about a trip.
The clearest clue is economic. A formal model of agentic search Why do consumers delay adopting agentic search? separates two costs that are easy to blur. Platforms judge agentic search by how often it leads to a purchase. Consumers also pay a search cost in their own time and attention. A person stops comparing fares after a handful of tabs because their patience runs out. An agent doesn't get tired, so the brake that limits human searching mostly disappears. Agents also rank options by the criteria they were given rather than by paid placement Are AI agents making attention-based ads irrelevant for aggregators?. Checking those criteria thoroughly means querying many combinations of dates, properties and fares that a human would never bother with.
The second clue is how agents search. Agents tend to explore step by step, issuing one query, reading the result and then issuing the next. A study of document retrieval found that this sequential style works on small collections but becomes inefficient as the search space grows. A single index that shows every candidate at once wins at scale Does lexical search outperform agentic navigation as corpus size grows?. Travel inventory is a very large search space, so this probe-and-refine habit multiplies query counts. Search is so central to how agents work that researchers training them often have an LLM imitate a search engine to avoid paying for millions of real API calls Can LLMs replace search engines during agent training?.
More searching doesn't automatically mean better searching. In one shopping experiment, a Gemini agent buying in larger quantities quietly switched from comparing unit prices to simply minimizing the total price. It did 25% less of the comparison work that actually distinguished good offers, and its rate of choosing the best option fell from 80% to 35% Does higher purchase volume make AI shoppers search less carefully?. So agents can run huge numbers of queries and still compare options poorly. Query volume measures activity, not judgment.
The twist you might not expect is where this load ends up. Agents don't replace the travel infrastructure. They lean on it harder. Online travel agencies are becoming the API layer that agents call into, and Expedia's B2B segment grew 24% in a year Are AI agents actually removing online travel agencies from bookings?. As a result, the companies that answer agent queries absorb the extra search traffic, while advertising revenue that depended on human browsing shrinks Do AI shopping agents destroy search ad revenue while leaving influencer marketing intact?.
Sources 7 notes
A formal model shows platforms rank agentic search by conversion revenue alone, while consumers also bear search costs. This divergence creates rational adoption lags where consumers keep manual search even after it would boost platform revenue.
As shopping and travel agents handle comparison and ranking based on stated criteria rather than paid placement, they render attention ads irrelevant. Tripadvisor's Hotels segment revenue fell 23% in Q2 2026, with management attributing further decline to structural SEO shifts—a pattern Jarc attributes to agent-mediated ranking.
Lexical search indexed globally outperforms sequential agentic exploration at scale because it exposes relevant candidates across the entire corpus at once. Agent-based search leads only on small corpora but becomes inefficient as search space grows, while BM25 maintains accuracy and cost advantage at deployment scale.
ZeroSearch and SSRL demonstrate that LLMs can generate relevant documents and search results from internal knowledge, with 14B simulators matching or exceeding real search engines. Curriculum degradation and test-time scaling optimize this approach for training without API costs.
As order volume increased from 1 to 250 bags, one Gemini model switched from comparing unit prices to minimizing total price, cutting diagnostic search by 25% and choice optimality from 80% to 35%. The shift suggests objective substitution, not cost-driven omission.
Show all 7 sources
Goswami's 2026 travel data shows independent hotels' OTA dependence rose to 63.4% even as AI agents emerged, with OTAs retaining merchant-of-record status through API integrations. Expedia's B2B segment grew 24% year-over-year, now 38% of revenue.
Rustamli argues AI shopping agents compress discovery-to-purchase into zero-click transactions, cutting search ad revenue, but lack capacity to create wants that drive influencer marketing—suggesting ad dollars migrate toward social platforms rather than disappearing entirely.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- What Is Your AI Agent Buying? Evaluation, Biases, Model Dependence, & Emerging Implications for Agentic E-Commerce
- From Product Search to Preference Articulation: The Economics of Agentic Commerce
- AI Agents Want to Shop for You: The Future of Agentic Commerce
- Vertical tacit collusion in AI-mediated markets
- Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses
- Agentic markets
- The Agent Economy: Who's Most at Risk?
- Agentic AI in Retail: How Autonomous Shopping Is Redefining the Customer Journey