Inside Skyscanner's Explore with AI: 100 Billion Daily Prices Power the Discovery Layer of AI Commerce
Skyscanner's summer release beta tests Explore with AI and an AI road trip planner built on the roughly 100 billion prices it scans daily. We unpack the metasearch's two-front strategy across its own channel and ChatGPT, and what the booking handoff means for operators downstream.
Key Takeaways
- Skyscanner's June 30, 2026 summer release beta-launches Explore with AI, a natural-language destination discovery tool, and an AI road trip planner, alongside upgrades to DROPS price alerts, Flight Tracker and the rebranded Stays platform
- A comparison service that holds no inventory is wiring the roughly 100 billion prices it scans daily into conversational AI, while running its own AI channel and a ChatGPT app in parallel — a preview of how the discovery layer of commerce is being rebuilt around AI
- Appearing in the comparisons and itineraries an AI assembles requires supplying machine-readable inventory and price data, and operators downstream of those recommendations need a reason to be chosen plus a way to measure AI-driven traffic
A Summer Release That Hands Discovery and Planning to AI
Skyscanner is testing personalized destination recommendations and road trip itineraries while expanding flight tracking, price alerts and accommodations.
www.phocuswire.comOn June 30, 2026, Skyscanner, which operates one of the world's largest flight metasearch platforms, announced a bundle of updates ahead of the summer travel season. A metasearch holds no airline seats or hotel rooms of its own; it aggregates and compares prices across many sellers. At the center of this release are two new AI features entering beta.
The first is Explore with AI. Instead of typing a destination and dates into a form, travelers can search in plain language, asking for something like "cheap flights to Japan in December." The screen lays out flight prices, duration, local weather and what Skyscanner calls destination vibes, with AI-generated commentary such as "September is typically 24% cheaper than December" adding context for the decision. The tool is available on Skyscanner's website, starting in English-speaking markets with expansion planned. According to the company, 60% of users in early testing clicked through from Explore with AI's suggestions to view flight options.
The second is the road trip planner. Given a starting point, dates and a journey style, the AI assembles a driving itinerary automatically. Five styles are on offer: Scenic Route, Fastest Route, Cultural Exploration, Adventure Trip and Relaxing Getaway. The tool recommends stops and attractions along the way, together with car hire options suited to the route. It is beta testing in all markets on desktop.
Existing features were lifted at the same time. Laid side by side, the release reads as an effort to cover discovery, planning, pre-booking judgment and in-trip support as one continuous surface.
| Feature | What Changed |
|---|---|
| Explore with AI | Natural-language discovery comparing price, duration, weather and vibes with AI commentary. Beta on the website in English-speaking markets |
| Road trip planner | Auto-generates driving itineraries from five journey styles, surfacing stops and car hire options. Beta in all markets on desktop |
| DROPS | App-exclusive price-drop alerts. Scans 100 billion prices daily and surfaces up to 822% more flights that fell 20%+ within seven days |
| Flight Tracker | Adds real-time gate, terminal and baggage belt information alongside departure and arrival status |
| Stays | Rebrands the accommodation platform, growing inventory from 3.5 million to more than 5 million properties, from hostels to farm stays |
CEO Bryan Batista framed the release as building tools that help travelers plan and book with confidence. Rather than a grab bag of improvements, it is groundwork for gathering the entire arc of a trip behind an AI-powered front door that Skyscanner owns.
Behind the Conversation: 100 Billion Machine-Readable Prices a Day
What separates Explore with AI's suggestions from a generic chatbot's travel advice is what backs the answers. Skyscanner serves 160 million travelers a month across 52 countries and 37 languages, aggregating fares from more than 1,400 travel partners. The prices it scans run to roughly 100 billion a day. Explore with AI is best understood as that live dataset repackaged into conversation.
Ask a general-purpose AI for travel advice and you may get stale prices or options that do not exist, artifacts of its training data. A metasearch's AI, by contrast, is connected to real inventory and real prices from the moment it makes a suggestion. Even the line "September is cheaper than December" is a summary derived from prices already scanned. Grounding answers in verified, live data is becoming the standard playbook for trustworthy AI-driven product discovery, a design choice this release shares with Marriott's Ask Bonvoy.
For retail and e-commerce, this maps directly onto the question of product data readiness. The foundation of any AI-assembled comparison is accurate, machine-readable data on inventory, prices and attributes. Merchants that can supply it see their products surface inside the conversation; merchants that cannot never appear in the comparison at all. It is another confirmation that organizing product data is the prerequisite for shelf space in the AI era.
The 60% clickthrough figure deserves attention too. Six in ten users who narrowed down a destination in conversation went on to view specific flights. When AI takes over the discovery stage, the demand flowing downstream is no longer a vague inquiry but purchase intent with defined parameters. Conversational discovery is starting to function not as a marketing flourish but as a replumbing of the top of the purchase funnel.
AI on the Owned Channel, an App Inside ChatGPT
This release needs to be read alongside the moves Skyscanner has been making on external AI channels. On February 27, 2026, the company launched a flight search app inside ChatGPT, initially in the U.K. and U.S. A conversational request such as "the cheapest flight to New York in December" returns candidates with live prices. Going back further, in January 2025 Skyscanner joined as a launch partner for Operator, OpenAI's agent that autonomously drives a browser, accumulating experiments in AI-led flight search.
In other words, Skyscanner is making its own site conversational while simultaneously carrying its data into external AI assistants. The ChatGPT app store is a venue where competitors such as Expedia and Booking.com also operate, so setting up shop there means accepting the risk of being compared. Skyscanner chooses the two-front posture anyway, because it cannot ignore the shift of travel conversations toward general-purpose AI.
There is a familiar shape to this. Metasearch itself rose as an aggregator sitting above airlines and hotels, holding the customer relationship. Now general-purpose AI is building one more aggregation layer above it, and the aggregator faces the prospect of being re-aggregated. Polishing the owned channel while feeding data to external AI is exactly the question of balancing the open garden and the walled garden that defines channel design in the AI era.
Retail and e-commerce operators face the same call. Investing in an AI entrance on your own site and preparing product information for ChatGPT or Google's AI are not mutually exclusive. Both routes rest on the same machine-readable product data, and Skyscanner's moves show how an early investment in that foundation pays off across both channels.
AI Runs Discovery, but the Transaction Is Still Handed Off
The other thing to note is that none of these features steps into the transaction itself. Metasearch is a referral business, and the Explore with AI conversation ends its job at presenting flight options. Inside the ChatGPT app as well, booking still completes with the linked travel partner, as it always has. This is a different stage from an autonomous purchasing agent that executes payment on the user's behalf.
Agentic commerce refers to transactions in which AI carries out comparison through purchase for a person. Skyscanner's current position covers the earlier stages: delegated discovery and planning. Delegating payment would additionally require identity verification and consent management, and there is no announcement that the referral-model company intends to go there. Put without exaggeration, this release attacks the industry-wide split between discovery and checkout from the discovery side.
That split matters greatly to the operators receiving the transaction downstream. Even when AI assembles the itinerary and the comparison, booking, payment and post-purchase service remain in the hands of airlines, hotels and car rental companies. Knowing the exact moment the customer relationship is handed to you, and having a reason to be chosen at that moment, is the condition for capturing AI-originated demand.
What Retail and E-commerce Operators Should Take Away
First, the AI transformation of the comparison layer is not a travel-only phenomenon. Physical retail has its own venues of comparison, from shopping search to price comparison sites, and their entry points are shifting to conversation in the same way. The condition for appearing in an AI-assembled comparison table is supplying accurate, machine-readable data on products, inventory and prices. A product whose data is not exposed can be competitively priced and still never make the shortlist.
Second, contextual insertion of related products becomes a new form of exposure. The road trip planner recommended car hire along the thread of an itinerary. Translated to retail, this corresponds to surfacing complementary products in the context of a use case or a season. Mapping the contexts in which your products could be called up, and preparing the attribute data and connections for it, is how you earn a place on the AI-mediated shelf.
Third, you need a yardstick for AI-driven traffic. Skyscanner explained the traction of its new feature with a 60% clickthrough figure. For a merchant, the equivalent is designing how to identify inflows from AI channels and attribute outcomes to them. The more discovery moves into AI, the further measurement built around conventional search traffic drifts from reality. In a structure where the transaction is handed off, that measurement becomes nearly the only basis for investment decisions.
Conclusion
Skyscanner's summer update reads as a comparison service with no inventory of its own moving to claim the discovery layer of the AI era, armed with its price data asset. Running a conversational owned channel and a ChatGPT app at the same time is the pragmatic answer in an environment where the aggregator itself risks being re-aggregated. Meanwhile, booking and payment are still handed off to travel partners, leaving the industry-wide split between discovery and transaction in place. For retail and e-commerce operators, the concrete preparations this points to are supplying inventory and price data AI can read, organizing attributes so products get called up in context, and measuring AI-driven traffic. However far the entry point of discovery moves into conversation, the transaction at the end still lands with the operator.
