Young Chinese plan 13-day holiday trips with spreadsheets and AI drafts
In short
Young travellers in China planned their Mid-Autumn and National Day holiday trips with detailed spreadsheets, using AI for first-draft routes, in reports published by Huxiu, Jiemian and TMTPost. Qunar data cited in the reports showed travel on the Sept. 28–30 leave days rose 52% from a year earlier, and those days accounted for 15% of hotel bookings in the 13-day holiday window. Plans were easily disrupted: one traveller cancelled her first three stops before departure, and both said they would pay more for smoother, less queue-bound trips.
Read the full story 3 min read
Young travellers in China planned their Mid-Autumn and National Day holiday trips with detailed spreadsheets, a pattern the reports call “schedule people”, according to accounts published by Huxiu, Jiemian and TMTPost. Xiao Ge, who lives in Shanghai, began her holiday on Sept. 25 and used three days of annual leave on Sept. 28–30 to link the two holidays into a 13-day break, with a route running from Shangrao, Gexiancun and Nanchang to Shenzhen, Hong Kong, the Chaoshan area and Quanzhou. Huxiu described this as a route through six provinces and eight places. [ 1 , 2 , 3 ]
Qunar data cited in the reports showed the number of travellers taking the leave-days trip rose 52% from a year earlier, and hotel bookings on those three working days accounted for 15% of the 13-day window, about one booking in seven. Advance hotel bookings in popular holiday cities rose nearly 40% year on year, and flight bookings made a month ahead were 21% higher than a year earlier. [ 1 , 2 , 3 ]
Xiao Ge said she gave GPT a request covering her departure city, trip length and destinations to get a first draft, but did not follow it fully because the transport times were sometimes wrong or a stop did not warrant the time. Xiao Han, who also works in Shanghai, recorded wake-up times, transport, costs and buffer time for a weekend trip to Hong Kong Disneyland, comparing ticketing apps with posts on Xiaohongshu; she said official information lists schedules but not how much buffer to allow. The interviews were attributed to Jingzhe Research Institute in the Jiemian and TMTPost versions. [ 1 , 2 , 3 ]
Plans did not survive contact with events. Xiao Ge cancelled the first three stops — Shangrao, Gexiancun and Nanchang — before departure because of a family matter, and decided to fly directly to Shenzhen, discarding her route research and hotel bookings. She said sticking too closely to a checklist turns a trip into a task, and that she deliberately left blank time in this trip's schedule. Xiao Han said her body could not keep up with her spreadsheet, and that she checked the time constantly whenever a step ran a few minutes late. [ 1 , 2 , 3 ]
Both said they would pay more for smoother travel. Xiao Han said arriving an hour earlier can mean two or three more rides, so the same ticket buys less for a late arrival, and she was willing to plan ahead and spend more; Xiao Ge said she would choose a slightly more expensive hotel close to a station or the next day's destination and would consider paying to skip queues of two to three hours. [ 1 , 2 , 3 ]
Huxiu cited the National Advertising Research Institute's “Report on the Development of Cultural Tourism Brands in the New Era” as finding that 44.23% of Generation Z were willing to pay a moderate premium for unique experiences, while 41.35% sought value for money. Queue-Times figures cited by Huxiu put average waits at several popular Shanghai Disneyland attractions in 2025 close to or above an hour, including 85 minutes for Seven Dwarfs Mine Train, 72 minutes for TRON Lightcycle Power Run and 70 minutes for Soaring Over the Horizon; visitors buy Premier Access and Early Park Entry passes, and Universal Beijing Resort offers its Express pass. [ 1 ]
Huxiu also cited the “2026 First-Half AI Travel Application Trends Insight Report” as showing consumer use of travel AI tools at nearly 80%, and Fliggy data showing AI orders during the 2026 Spring Festival rose more than 800% from the pre-holiday period. The same report said 66.2% of consumers still returned to traditional platforms to double-check AI recommendations, and only 15.2% trusted AI enough to buy directly. Social media users cited outdated information and ignored queueing, stamina and transfer costs, and the Quzhou Daily said AI suits building a framework while dynamic information should be confirmed through official platforms or staff. The reports concluded that competition in cultural tourism services is shifting from providing information to improving experiences. [ 1 ]
Why it matters
The figures cited indicate demand concentrated in a short leave window, and paid time-saving products at theme parks are described as normal rather than exceptional. With consumer use of travel AI tools near 80% but only 15.2% ready to buy directly on an AI recommendation, the reports argue travel businesses will compete on experience rather than on providing information. The evidence rests on two travellers' accounts and platform data, so it points to a consumer trend rather than a measured market outcome.
Key facts
- Young travellers in China planned holiday trips with detailed tables of times, transport, costs and buffer time, a pattern the reports call “schedule people”. [ 1 , 2 , 3 ]
- Xiao Ge, who lives in Shanghai, took three days of annual leave on Sept. 28–30 to link the Mid-Autumn and National Day holidays into a 13-day break. [ 1 , 2 , 3 ]
- She gave an AI tool a first-draft route but checked it, saying the transport times were sometimes wrong or a stop did not warrant the time. [ 1 , 2 , 3 ]
- Xiao Ge cancelled her first three stops — Shangrao, Gexiancun and Nanchang — before departure because of a family matter, and decided to fly directly to Shenzhen. [ 1 , 2 , 3 ]
- Qunar data cited by the outlets showed travel on Sept. 28–30 rose 52% year on year, with hotel bookings on those days accounting for 15% of the 13-day holiday window. [ 1 , 2 , 3 ]
- Advance hotel bookings in popular holiday cities rose nearly 40% year on year, and flight bookings made a month ahead were up 21%. [ 1 , 2 , 3 ]
- Both travellers said they were willing to spend more for a smoother trip, including paying to skip queues of two to three hours. [ 1 , 2 , 3 ]
- Huxiu cited figures putting consumer use of travel AI tools at nearly 80%, but only 15.2% of consumers trusted AI enough to buy directly. [ 1 ]
Confirmed by several sources
- Two Shanghai residents, identified as Xiao Ge and Xiao Han, planned their trips with detailed tables covering departure times, transport and costs. [ 1 , 2 , 3 ]
- Xiao Ge used a “three days off, 13 days holiday” strategy, taking leave on Sept. 28–30. [ 1 , 2 , 3 ]
- Xiao Ge used an AI tool for a first-draft itinerary but did not follow it completely. [ 1 , 2 , 3 ]
- Xiao Ge cancelled her first three stops before departure because of a family matter and decided to fly directly to Shenzhen. [ 1 , 2 , 3 ]
- Qunar data cited by all three outlets: travel on Sept. 28–30 was up 52% year on year, and hotel bookings on those days made up 15% of the 13-day holiday window. [ 1 , 2 , 3 ]
Still unclear
- How Xiao Ge's revised trip went after she changed her plans. The documents say only that she decided to fly directly to Shenzhen after cancelling the first three stops.
- The number of destinations on Xiao Ge's route. Huxiu describes a six-province, eight-place route while naming seven places; Jiemian and TMTPost name the same seven places without giving a count.
- Theme-park queue times, customized-tour booking growth and AI adoption figures. These appear only in the Huxiu article; the Jiemian and TMTPost versions end before that section.
- The year of the holiday described. Huxiu refers to the 2025 Mid-Autumn and National Day holiday while citing 2026 reports, and all three documents were published in October 2026; the documents do not reconcile this.
What local media are saying
Timeline, local time
- Sept. 25: Xiao Ge's planned departure time at the start of her 13-day trip. [ 2 , 3 ]
- Planned departure of the high-speed train. [ 2 , 3 ]
- Planned arrival in Shangrao. [ 2 , 3 ]
- Huxiu publishes its report on the trend. [ 1 ]
- Jiemian publishes its version of the report. [ 2 ]
- TMTPost publishes its version of the report. [ 3 ]