AI visibility: winning travel demand in AI assistants

A new first step in the customer journey

For two decades, direct-channel acquisition has rested on three pillars: search engines, metasearch and CRM. Generative assistants add a fourth that changes the rules: customers no longer compare ten links, they receive a synthesised answer with two or three recommendations.

The World Aviation Festival 2026 agenda reflects this concern with sessions such as "Winning travel demand in the age of AI", "Beyond discoverability: who owns the offer in the age of AI?" and "Your AI assistants are now booking your flights". The underlying question is the same: who controls the offer when the customer's first conversation is with a machine.

What makes an assistant recommend your brand

Content that answers real questions

Assistants favour sources that answer specific questions precisely: what a fare includes, which routes operate in winter, which baggage policy applies, how to reach the hotel. Websites that only show the booking engine and promotional messages provide little useful material.

Structured, consistent data

Schedules, indicative prices, services, conditions and locations should be published in a structured way and be consistent across the website, metasearch and third-party listings. Inconsistency erodes trust, just as it does when managing metasearch such as Google Flights, Skyscanner and Kayak.

Authority and external mentions

Models learn from what others say about the brand: press, reviews, guides and comparison sites. An AI visibility strategy is also a reputation strategy.

From discovery to booking

The next step is already under way: assistants that do not just recommend but book. The advantage there lies with businesses whose offers and actions are accessible via API, a topic we cover when analysing how AI agents may reshape airline distribution. For hotels, the logic mirrors a metasearch strategy: be present where the decision is made, with reliable price and availability. It also pays to tell those legitimate agents apart from abusive traffic, as covered in bots, scraping and look-to-book.

What the ecommerce team should do

1. Audit how assistants answer today. Ask about key routes, destinations and services and record who appears, with what data and what errors.

2. Publish answer content. Destination, route, service and policy pages that answer specific questions with verifiable information.

3. Structure the data. Schema markup, product feeds and consistency between owned and third-party sources.

4. Measure traffic from assistants. Separate AI assistant referrals in analytics and track their conversion.

5. Prepare the transactional layer. Offer and booking APIs with access rules, so the business does not rely solely on intermediaries.

How to measure progress

• Share of presence in answers: percentage of key questions where the brand appears.

• Accuracy of cited information: correct prices, conditions and services.

• Traffic and conversion from assistants: as a channel in its own right in the attribution model.

• Weight of intermediaries in answers: how often the assistant sends the customer to an OTA.

Measuring this channel has the same pitfalls as any other; it is worth reviewing the most common attribution mistakes before drawing conclusions.

Conclusion

Visibility in AI assistants does not replace SEO or metasearch, but it changes what gets rewarded: clarity, consistent data and the ability to transact. Brands that become the most reliable source about their own offer will keep control of demand; the rest will see it intermediated once again.

Consumer Services Hub - Strategic ecommerce consultancy for B2C service companies

consumerserviceshub.com