Does caching fares give infrastructure providers pricing power?

Mauricio Prieto, citing an OAG post and Amadeus's own statements, argues that AI agents searching on travelers' behalf are driving search volume per ticket sold to levels that make fresh per-query pricing uneconomical, pushing at least one major travel technology provider to precompute and cache fares rather than calculate them anew for every request. He cites a developer who "pointed an AI agent at Etihad's website and asked it to find one flight" and got back "881,076 fare options." OAG estimates the industry has gone from roughly "100 to 200 searches per ticket sold in the early online era" to "somewhere around 10,000 to 20,000 today," and "Amadeus and Sabre have both pointed to a possible future of up to 200,000 searches per ticket sold once agents are doing the shopping."

Prieto frames the cause as the loss of a natural limit on search volume: "A human search has a natural brake: people get tired, bored, or satisfied with 'good enough.' Agents don't." Someone has to pay for the resulting compute, and he reports Amadeus's answer, from its president of travel Decius Valmorbida, as no longer "calculating a fresh price for every single query." Instead, for common routes and dates, Amadeus "precomputes likely fares in advance" and serves those saved answers instantly, "refreshing them periodically rather than recalculating from scratch every time an agent asks."

Prieto draws out the market-structure implication of this caching layer directly: "the infrastructure that decides which fares an agent sees, how quickly it sees them, and whether the offer can still be honored" gives whoever runs "the cache, refresh logic, and revalidation layer real influence over the market," and an airline whose offers are "less current or less reliably bookable risks being repriced at checkout, or becoming a worse option for the agent to recommend." This is a different lever on agent-visible outcomes than the one in Can AI seller agents shift buyer preferences through description edits?, where a seller edits its own listing text — here the chokepoint is a third-party infrastructure layer that mediates what every seller's agent-facing offer looks like, regardless of how the seller describes its product. It also gives a concrete mechanism for the dynamic in Why do consumers delay adopting agentic search?: a caching layer's refresh cadence is itself a lever a platform can tune toward its own economics rather than query-time accuracy for the consumer. The companion claim in Are AI agents actually disintermediating online travel platforms?, from the same piece, is one account of where this infrastructure value shows up in the market.

The search-volume figures are Amadeus's, Sabre's and OAG's own estimates as reported by Prieto, not an independently benchmarked measurement, and the excerpt gives no detail on how caching affects actual fare accuracy or consumer outcomes, nor any evidence yet that this dynamic has conferred measurable pricing power on Amadeus or any single infrastructure provider. The implication, held no more strongly than the evidence supports, is that as agentic search scales, control over caching and revalidation becomes a plausible new site of competitive advantage in travel distribution — a claim about structural incentives, not yet a documented market outcome.