When 80 million foreigners flood into China, what does Ctrip rely on to keep foreigners staying?

Whether it’s Liangma River in Beijing or Anfu Road in Shanghai, visibly foreign faces seem to herald the return of inbound tourism. Data indicates that by 2025, the total number of inbound visitors is expected to exceed 80 million, setting a new record.

The outside world often attributes the growth of inbound tourism to policy advantages, such as the continuous expansion of visa-free “friend circles.” However, policies are only the first step in attracting overseas travelers. The more difficult challenge is solving the “how to play” — how to help foreign tourists understand Chinese hotel cancellation and modification policies? How to showcase hidden intangible cultural heritage in counties to the world?

This is the key to encouraging foreign visitors to stay longer and spend more, and it is also a critical question for the entire tourism industry chain.

Ctrip’s solution is AI. By 2025, product R&D expenses will reach 15.1 billion yuan, a 15% year-over-year increase. They are leveraging AI to translate China’s complex tourism resources into products that the entire world can understand and directly book.

Rebuilding Connections: First, Let China Be “Seen”

For a foreign tourist who has never been to China, the biggest obstacle is often not visa procedures but feeling lost in the Chinese internet landscape.

Many foreign tourists want to visit Dali, Zibo, or Sichuan to see pandas, but on mainstream overseas travel websites, information is extremely limited. Most of China’s high-quality tourism resources—from特色民宿 in small towns, hidden eateries in hutongs, to intangible cultural heritage experiences in counties—are often only available on Chinese-language websites.

Ctrip’s AI translation engine, “Smart Translate Future,” produces 6 billion words of content annually, covering 25 core languages. The value of this data lies not just in volume and speed but in accuracy.

Previously, translating a county-level scenic spot into authentic English, French, or Korean required professional human translators, which was costly and inefficient. Many small scenic spots simply gave up on overseas markets. Language barriers often blocked foreign tourists at this step.

Now, Ctrip uses AI to help with multilingual coverage of attraction tickets and entertainment lines, allowing foreign tourists to see options that originally didn’t exist in China. Small attractions in second- and third-tier cities, previously only described in Chinese, can now be searched and booked by global tourists.

Data shows that by 2025, the total inbound tourists are expected to surpass 80 million, reaching a new high.

Ctrip’s financial reports also confirm the booming “China Travel” trend. Last year’s Q4 data shows that bookings on Ctrip’s international OTA platform for hotels and flights increased by 60% year-over-year.

Another feature is that foreign tourists’ footprints are gradually moving from first-tier cities to third, fourth, and even lower-tier cities. For example, according to Ctrip data, Zhangjiajie in Hunan, with its奇山峻岭, has attracted a large number of Korean tourists, with inbound hotel bookings more than doubling. Destinations like Yili in Xinjiang, Aba in Sichuan, and Qingyuan in Guangdong are also beginning to enter the international tourist spotlight.

Rebuilding Decision-Making: Translating “Chinese-Style Hotel Rules”

Seeing is just the beginning; being willing to pay is the key to completing the business cycle.

A significant industry pain point is that general AI can translate literal meanings but struggles to explain complex business rules. Especially when it comes to China’s特色退改签政策 and hotel highlights, any semantic ambiguity can cause foreign tourists to hesitate because they “don’t understand the product rules.”

This is where vertical industry-specific large models create a moat.

Unlike general models trained on publicly available internet texts, Ctrip’s AI is built on the company’s own vast tourism data assets, meaning it not only understands language but also understands tourism and business.

A deeper reconstruction occurs in the itinerary decision-making process. When faced with the massive information of “How to play in such a big China,” traditional search methods are highly inefficient. Ctrip’s AI itinerary recommendation product acts as a tireless “local guide.”

After experiencing it, users can let Ctrip’s AI itinerary assistant plan a trip to Dali. The assistant will understand their needs, find destinations, check popular attractions, select suitable hotels, plan transportation routes, and then organize a complete itinerary.

It doesn’t just mechanically list attractions but provides a closed-loop plan including hotels, tickets, and transportation. The data sources for Ctrip’s itinerary learning mainly come from platform orders and other real transaction data.

During the planning process, tourists can click on specific attractions to see reviews, ticketing options, and more, greatly reducing the decision-making difficulty for inbound travel.

When a complex half-month itinerary can be planned in minutes, technology directly translates into increased orders.

The “2025 Ctrip Travel Decision Double-Engine Insight Report” shows that by 2025, the proportion of users using AI for travel planning in searches will increase by 270%, and the number of users deeply engaging with the AI Q&A product “Wendao” will surge by 367%.

Service Rebuilding: Bridging the “Last Mile” of Inbound Tourism

If connection and decision-making solve the “whether to come” problem, then service experience determines whether they will “come again.”

In actual inbound tourism scenarios, the most awkward moments often occur in the “last mile”: long waits at hotel front desks due to communication issues, or struggling with complex ticket machines at scenic spots.

Data shows that Ctrip has deployed smart ticket machines at 241 scenic spots nationwide. These machines support over 16 languages, achieving full native-language UI interaction and ticket printing.

In online services, AI customer support also eliminates issues of time zone differences and language barriers.

In instant messaging (IM) consultations, AI supports foreign tourists to communicate with customer service in their native language in real-time, with bidirectional translation within 5 seconds, breaking the service bottleneck of multilingual support. This means a Spanish-speaking tourist can ask questions to a Chinese-speaking hotel front desk in real-time, both feeling like they are speaking their native language.

In text communication, Ctrip enhances international inquiry handling through AI-powered customer service assistants, knowledge base CopilotAI, providing real-time answer recommendations, smart scripts, order guidance, and multi-scenario Q&A support, greatly improving efficiency and professionalism.

In the past, inbound tourism relied heavily on human resources—hiring foreign-language guides and customer service staff—making scale difficult and costly. As AI matures, the marginal cost of inbound tourism will further decrease, and revenue from improved user experience will continue to grow.

When a French traveler can easily book a homestay in Guizhou, chat happily with the owner, and handle all tickets just like in Paris, the policy dividends truly turn into lasting industry benefits. In this process, Ctrip is building its competitive moat with AI.