Why biztribution / Look-to-book
Future now
Look-to-book keeps climbing. AI is adding the next zeros.
Travel shopping has changed dramatically. Twenty years ago, a ticket followed a hundred or two searches. Metasearch multiplied every query across dozens of sellers, and today the industry answers around 20,000 searches for every ticket sold. AI agents are removing the last limit on search traffic: the human. Amadeus and Sabre project up to 200,000 searches per ticket as early as this year, and OAG sees around 2 million by the end of the decade. IATA itself warns the ratio could reach millions to one.
View the data
| Period | Searches per ticket | Era | Source |
|---|---|---|---|
| 2000–2005 | 100–200 : 1 | Airline.com | OAG, 2026 · PhocusWire (Gibergues), 2025 |
| 2005–2015 | ≈1,000 : 1 | Metasearch | OAG, 2026 · PhocusWire (Gibergues), 2025 · PhocusWire (Amadeus), 2015 |
| 2015–2025 | 20K : 1 | Modern OTA | OAG, 2026 · PhocusWire (Gibergues), 2025 |
| 2025–2030* | up to 200K : 1 | Agentic AI 1.0 | OAG, 2026 |
| 2030+* | ≈2M : 1 | Agentic AI 2.0 | OAG, 2026 · IATA white paper, 2025 |
* Projection.
2000–2005
100–200 : 1
Airline websites and the first OTAs: a ticket followed a hundred or two searches.
2005–2015
≈1,000 : 1
One shopper query fans out across many airlines, OTAs and metasearch engines: a tenfold jump.
2015–2025
20K : 1
Calendar search, virtual interlining, multi-market sourcing and packaging each add 10+ queries per search. Where travel sits today in mid-2026.
2025–2030 · projection
up to 200K : 1
Agentic AI 1.0[4]
An agent runs hundreds, even thousands, of searches for one request. Amadeus and Sabre both project up to 200,000 per ticket as early as 2026.
2030+ · projection
≈2M : 1
Another order of magnitude is plausible by the end of the decade, and 2 million is the midpoint of the range, not the ceiling. IATA also warns of millions to one.
Sources: OAG, 2026 · PhocusWire (Gibergues), 2025 · PhocusWire (Amadeus), 2015 · IATA white paper, 2025
The shopping shift
Airlines are taking shopping back from the GDS, and every search with it.
For decades, airlines delegated shopping: they filed fares and distributed schedules and availability, and GDSs answered the searches. That kept search volumes away from airline systems.[2] With NDC and Modern Airline Retailing, the industry is moving to 100% Offers and Orders[5], the airline creates every offer itself, so shopping traffic increasingly lands on airline systems.[2]
Yesterday
Shopping delegated to the GDS
- Airlines file fares and distribute schedules & availability
- GDSs build the offers and absorb the searches
- The airline mostly sees bookings
Who answers each search?GDS
Today
The transition
- NDC offers flow alongside traditional GDS content
- More and more searches reach airline systems
- Look-to-book becomes the airline’s own problem
Who answers each search?GDS + airline
Tomorrow
The airline in charge of shopping
- 100% Offers & Orders: the airline creates every offer
- Every search, human or AI agent, asks the airline
- Retail-scale shopping capacity becomes a core airline capability
Who answers each search?The airline
biztribution lets the airline take charge of shopping searches without its Offer & Order system carrying the load: every search answered on the airline’s behalf, from the airline’s own data, transparently for the traveller.
The agentic shift
One sentence. Thousands of searches. One ticket.
Tomorrow’s traveller won’t search, they’ll delegate. The agent re-runs the search around the clock across every date, connection and seller until the price appears.[4] IATA expects agentic AI to keep “continually modifying searches to seek the best deal.”[2]
- AI agents search continuously, erasing the natural peaks and troughs systems were sized for[3]
- 24/7
Sources: IATA white paper, 2025 · IATA Airlines, 2025 · PhocusWire (Gibergues), 2025
1
sentence typed
0
shopping requests
0
ticket booked
Illustrative simulation: a traveller asks an AI agent to find a trip from Singapore to Sydney for less than 450 Australian dollars. The agent re-runs searches across dates, connections and sellers for days, generating tens of thousands of shopping requests, until it finds a fare under the target and books one ticket.
Every one of those requests lands on someone's inventory, unless biztribution answers it on the provider’s behalf.
Sources
- [1]Look to Book: Managing search volumes while creating relevant offers for the customer , IATA white paper (14 airlines, 11 IT providers, 8 sellers), Oct 2025.
- [2]Toward a more efficient online shopping search , IATA Airlines magazine, Nov 2025.
- [3]The never-ending growth of flight search traffic: From travel agents to AI operators , PhocusWire: S. Gibergues, former VP digital search, Amadeus, Mar 2025.
- [4]Look-to-Book and the End of the Old Travel Tech Architecture , OAG: J. Burke, 2026.
- [5]Airline Retailing: the transition to 100% Offers and Orders (NDC, ONE Order) , IATA programme page, accessed Sep 2026.
- [6]Airline system look-to-book ratios soar, expected to go 10x higher , PhocusWire, citing Amadeus, 2015.
The look-to-book figures follow OAG’s 2026 timeline, which matches estimates published independently by a former Amadeus head of digital search (PhocusWire, 2025) and by Sabre (end of 2025). They are order-of-magnitude estimates, not measurements: the chart plots the midpoint of each published range, and the 2026 and 2030 values are projections (Amadeus and Sabre for the first AI agents, OAG for the end of the decade). IATA’s Oct-2025 Look-to-Book paper reaches the same order of magnitude, warning that the ratio could climb to hundreds of thousands, even millions, to one. Ratios vary widely by airline and channel. The Singapore → Sydney agent is an illustrative simulation.