Chinese travellers no longer follow a linear path from search engine to online travel agency and booking. A traveller may discover a destination on Xiaohongshu, ask DeepSeek for an itinerary, verify hotels on Ctrip, check maps, review payment methods and finish inside a WeChat Mini Program. An organisation can appear several times in that journey and still lose the sale if information is inaccurate or the final action is difficult.

In brief
  • AI is becoming a decision layer that summarises, compares and organises choices before booking.
  • Visibility needs to be measured inside China’s own digital ecosystem, not only on Google or an international website.
  • For travel retail, the goal is to connect recommendation, location, assortment, payment and purchase before the passenger reaches the airport.
The key point The new competition is not simply about appearing in an answer. The AI needs to describe the offer correctly, connect it with a specific need and direct the traveller towards a trusted booking or purchase path.

What is happening

China Trading Desk has launched China Traveller GEO Concierge, a service that audits how organisations and destinations appear across AI answers, social search, travel platforms, maps, official websites and WeChat journeys. It uses Simplified Chinese queries based on real behaviours, including destination comparison, store location, tax refunds, payment methods, appointments, authenticity and after-sales support.

The launch cases tested a high-intent query across DeepSeek, Doubao, Yuanbao, Qwen and ERNIE. Changi Airport scored 89 out of 100. Dubai Duty Free reached 90 and Bulgari at Marina Bay Sands scored 79. These are proprietary snapshots of a specific query, not a universal benchmark. Their value is to show that a highly visible brand may still share recommendation space with competitors or present incomplete details on terminals, opening hours, payment or availability.

The conclusion matters. AI does not create a new and perfect database. It summarises signals from websites, maps, OTAs, social content, reviews and partners. When those signals conflict, the result can be an ambiguous, outdated or commercially unusable recommendation.

Why the model is changing now

AI has reached consumer scale in China

China ended 2025 with 602 million generative AI users and a 42.8% adoption rate, according to CNNIC. Usage is no longer confined to experimenters. It is becoming part of search, translation, planning and customer service.

International travel creates a need to reduce uncertainty

Visas, flights, transport, language, payments, tax refunds, reservations and opening hours make an overseas trip complex. AI conversation allows several constraints to sit in one question. “Where can I buy authentic cosmetics near my gate, pay with Alipay and collect before departure?” carries more intent than a generic duty-free search.

Social search and peer recommendations still build trust

AI does not replace Xiaohongshu, Douyin, WeChat or OTAs. It connects them. Dragon Trail found that Xiaohongshu was the leading source of overseas destination information for 57% of respondents in September 2024. A traveller can receive an AI recommendation and then look for real experiences, prices and reviews.

The commercial value of the market supports better competition

UN Tourism estimated Chinese outbound tourism expenditure at US$251 billion in 2024. Recovery remains uneven by destination and segment, but the opportunity is too large to depend on partially translated content or campaigns disconnected from purchase.

From answer to commercial action
1

Query

A traveller asks with need, route, language, payment and time constraints.

2

Recommendation

AI selects brands, airports, stores or destinations from available signals.

3

Verification

Social search, OTAs, maps and reviews confirm whether the answer feels true.

4

Conversion

Booking, navigation, reservation, payment or purchase must be easy to complete.

5

Measurement

Visibility only matters when it links to visits, bookings, sales and value.

Every transition requires consistent information and a trusted action path.

What it means for tourism and travel retail

For an airport, being recognised as a shopping destination is insufficient. AI needs to understand what exists in each terminal, what can be reserved, which payment methods are accepted and which duty-free conditions apply. An answer that sends the passenger to the wrong terminal can destroy trust.

For a retailer, the physical store enters a decision that began before travel. Seventy-eight per cent of travellers surveyed by China Trading Desk researched products in advance. Assortment, price, authenticity, exclusives and availability therefore become pre-trip content. The aim is to enter the shopping list before the passenger passes security.

For destinations and hotels, recommendation becomes contextual. AI may select a hotel for metro access, Chinese breakfast, family rooms or proximity to shopping. Generic luxury descriptions are less useful than concrete and verifiable attributes.

Retail media changes too. A campaign may influence Xiaohongshu, appear later in an AI answer and finish as a brand search inside an OTA or Mini Program. Last-click attribution cannot capture that sequence. Companies need to measure presence, accuracy, consideration and conversion.

Where the commercial opportunities appear

Airports

Make the terminal readable

Publish locations, opening hours, payments, reservations, tax refunds and internal routes in Simplified Chinese.

Retailers

Enter the pre-trip list

Connect categories, brands, exclusives, stock and collection with high-intent questions.

FMCG, beauty and luxury

Describe purchase occasions

Explain gifting, authenticity, price differences, travel sizes and after-sales support.

Destinations

Turn inspiration into an itinerary

Structure transport, districts, events, shopping, accessibility and time requirements.

Hotels

Compete on context

Show attributes for families, silver travellers, shopping, business and connections.

Airlines

Integrate recommendations

Connect route, terminal, baggage, services and destination offers before departure.

Payments and services

Reduce friction

Clarify Alipay, WeChat Pay, cards, eSIM, mobility, insurance and support in Chinese.

Technology and retail media

Link content with inventory

Create structured data, assistants, audiences and measurement of AI influence.

A question for leadership teams

Do you know what a Chinese AI says when a traveller asks about your airport, store or brand?

Audit visibility

Risks and practical barriers

  • Chasing a score. A proprietary index can diagnose issues but cannot replace sales, satisfaction or share.
  • Outdated information. Hours, terminals, brands, stock and policies may change faster than answers.
  • Fragmented ecosystem. Each model, OTA, map and network uses different sources and criteria.
  • Translation without localisation. Correct text may still fail to answer how Chinese travellers actually ask.
  • Hallucination and weak attribution. AI can combine facts incorrectly and its influence is absent from last-click analytics.
  • No route to action. Recommendation has little value without accessible booking, payment, location or support.
  • Privacy and consent. Personalisation and tracking must respect regulation and user expectations.
  • Unconvincing synthetic content. Artificial images or reviews can reduce trust rather than improve it.

The Changi, Dubai Duty Free and Bulgari scores come from China Trading Desk launch snapshots for one query across five platforms. They are not general market rankings and do not prove commercial performance.

How Marksyte can help

Marksyte can connect digital visibility, traveller behaviour and commercial outcomes so that AI work does not end with a content audit.

Visibility audit

Evaluate answers, accuracy, competitors and action paths for Chinese-language queries.

Decision-journey map

Identify how AI, social search, OTAs, maps, partners and owned channels interact.

Structured data layer

Organise locations, products, services, hours, payments and policies for reuse.

Segmentation and demand

Prioritise routes, cities, profiles, travel occasions and categories with higher opportunity.

Assortment and personalisation

Adapt products, bundles and messages by route, segment and shopping mission.

Retail media

Design measurable pre-trip, airport and post-trip sequences.

AI assistants

Answer in Chinese about terminals, stock, payments, tax refunds, bookings and services.

Commercial measurement

Link visibility with branded search, traffic, bookings, conversion and sales.

A practical 90-day agenda

  1. Create a query library. Capture real questions by destination, route, category and profile.
  2. Audit five ecosystems. Compare AI models, social search, OTAs, maps and owned channels.
  3. Correct critical information. Prioritise terminal, hours, payments, availability, authenticity and after-sales.
  4. Connect action. Add booking, Mini Program, navigation, contact or purchase where relevant.
  5. Define measurement. Track presence, accuracy, branded search, visits and incremental outcomes.

AI visibility is not simply a new version of SEO. To reach Chinese travellers, companies need coherent information infrastructure, localised content, commercial data and a path that converts a recommendation into confidence and purchase.

Frequently asked questions

What does GEO mean in tourism and travel retail?

Generative Engine Optimisation aims to make an organisation accurately understood and represented in AI-generated answers. In Chinese tourism it should also cover social search, OTAs, maps, official sites, WeChat and the other channels that shape a decision.

Does appearing in an AI answer guarantee sales?

No. Visibility only creates an opportunity. The information needs to be correct, relevant to the query and connected to inventory, booking, payment, location, support or purchase. Without a clear conversion path, the recommendation may benefit another operator.

What data should airports and retailers publish?

Terminal and location, opening hours, brands and categories, availability where possible, payment methods, tax refund, reservations, languages, duty-free rules, services, accessibility and after-sales support. The information should be consistent in Simplified Chinese and kept current.

Sources

  1. TRBusiness, CTD launches GEO Concierge service for Chinese traveller engagement, 24 July 2026
  2. China Trading Desk, China Traveller GEO Concierge launch, 22 July 2026
  3. Chinese Government and CNNIC, internet and generative AI users, 5 February 2026
  4. UN Tourism, Chinese outbound tourism expenditure in 2024, 27 May 2025
  5. China Trading Desk, survey on the self-guided digital-first Chinese traveller, 25 June 2025
  6. Dragon Trail Research, Chinese Traveler Sentiment Report, September 2024
  7. China Trading Desk, AI Is Becoming the New Decision Layer in Chinese Travel, 11 June 2026
  8. Reuters, Chinese international travel recovery and airline capacity, 12 February 2026

Adoption, expenditure and behaviour figures come from the cited sources. GEO scores are proprietary China Trading Desk diagnostics and are presented with that scope.