How a low-cost carrier lifted GenAI shopping 27% with AI alignment
AI was answering travelers' route questions from whatever it could crawl, and the carrier couldn't tell whether fixing that would change shopping behavior on its own site.
Lift in GenAI shopping sessions on test routes over untreated control.
GenAI shopping session growth on test routes. Control grew 2.1x.
AI vital signs moved in the test group's favor.
The question
Travelers now ask AI about route availability, bag allowances, fare rules and flight times. Models piece the answer together from whatever they can crawl: old pages, OTAs, forums, review sites. Before alignment, leading AI models get airline products, policies and programs wrong 32.7% of the time, based on Bonafide's airline benchmark.
A low-cost carrier wanted to know two things. Does fixing the context change what AI says about its routes? And does that show up in shopping behavior on the airline's own site?
The design
The carrier ran a live A/B test. Twenty origin-destination routes in one market were split into two matched groups.
Control
No changes. Scored only.
Test
Route content curated, verified and orchestrated to AI surfaces through Bonafide, then tuned every measurement cycle.
Anything else happening on the airline's site hit both groups equally, so the gap between them is the signal. Bonafide tracked five AI vital signs, from accuracy to competitive rank, and tied them to shopping data from the airline's analytics.
Curate
Verified answers to real traveler questions, including the topics the airline's site left blank.
Orchestrate
Deployed that knowledge in machine-readable form, where AI platforms ingest it.
Tune
Measured how AI answered every cycle, test against control, tied to shopping data.
The results
AI got it right more often, ranked the airline higher and sent more shoppers.
Lift in GenAI shopping sessions on test routes over untreated control.
MeasuredLow-cost carrier · 20 routes · test vs control · three-month average before vs after orchestration
Same routes and period. Every other channel combined was flat to down.
Add-to-cart rate overtook control after orchestration began.
Measuredclimb in competitive rank on test routes. Control slipped 6.7%.
MeasuredTest vs control · October 2025 baseline through mid-2026
AI accuracy gain on test routes. Control moved 1.4%.
MeasuredTest vs control · October 2025 baseline through mid-2026 · original right/wrong grading
Better context upstream. More shoppers downstream.
All five AI vital signs moved in the test group's favor, including recommendation frequency and citation share.
What it shows
When AI had verified route content to work with, it answered more accurately, recommended the airline more often, ranked it higher and cited its site more. Those gains showed up on the airline's own site as more GenAI shoppers, and more of them adding to cart.
GenAI shopping is growing fast while every other channel holds flat or slips. Carriers that shape what AI knows about their routes now keep that direct path as the channel scales.
New channel. Same data problem.
If this is relevant to your network, we'd love to show you what it looks like on your routes.
Schedule a DemoMethodology
VerifiedMeasured result. Live A/B test across 20 matched origin-destination routes in one carrier's market. Test routes received Bonafide curation and orchestration; control routes did not.
