01. Losses narrowed, but still very short of money
Before starting an online ride-hailing business, Zhou Hui was an entrepreneur. He and several friends worked on Douyin e-commerce in Hangzhou. They worked for more than a year but failed to find a way, so they returned to their hometown in Hebei in frustration. Before determining his next life plan, Zhou Hui needed an income, so he came to Beijing and registered as an online ride-hailing driver on the Caocao travel platform.
Although Cao Cao’s travel volume is not as high as Didi’s, fortunately the threshold is low. He only needs to pay a monthly rent of 5,200 yuan to rent a “Cao Cao 60”.
However, after driving the online ride-hailing service for more than two months, Zhou Hui gave up. “I paid three months’ rent and stopped working after the last month,” Zhou Hui told Caijingtianxia.
“The online ride-hailing industry is too complicated.” Zhou Hui said bluntly that the main reason why he didn’t want to continue doing it was that it didn’t make money. Zhou Hui made a rough calculation and found that he drives for 10 hours a day, and his daily turnover is 400 to 500 yuan. Excluding 20% to 30% of the platform commission, vehicle rental, charging and other expenses, the daily net income is only more than 200 yuan. With such an income level, it is difficult to save money in Beijing.
It’s not just Zhou Hui who can’t keep money. The Cao Cao travel platform where he works is also facing financial difficulties.
Caocao Travel was founded in 2015 and is affiliated to the Geely system. Li Shufu, founder and chairman of Geely Holding Group, holds 73.41% of the shares of Caocao Travel through his wholly-owned Ugo Investment Limited and is the company’s controlling shareholder. Among the shareholders of Cao Cao Travel is Gong Xin, who holds 2.58% of the shares and is the company’s CEO.
Gong Xin is a veteran in the travel field. She once served as the head of Didi’s main business unit and joined Caocao Travel as CEO in 2021. Gong Xin has participated in many important nodes of Caocao Travel. The first major event he took office was to promote Caocao Travel to complete a 3.8 billion yuan Series B financing and lead the company to transform into an independent travel service provider; then, he proposed the “N³ Strategy” and launched exclusive customized models such as Caocao 60. In June 2025, Caocao Travel was officially listed on the main board of the Hong Kong Stock Exchange, becoming “the first domestic shared travel stock.”
However, Gong Xin still has to face extremely severe financial tests.
Cao Cao’s heavy historical burden, combined with the radical strategic choice of “exchanging debt for scale and using capital to secure the future” in recent years, has kept the company’s operations and capital chain in a tight state.
On August 27, Cao Cao Travel handed over a mixed semi-annual report. Fortunately, in the first half of 2026, the company’s revenue exceeded 10 billion yuan, and the loss during the period narrowed 17.5% year-on-year to 380 million yuan. The gross profit margin reached 9%, an increase of 0.3 percentage points from the same period last year. The scale of platform operation has grown steadily. During the same period, the GTV (total transaction volume) of Caocao’s travel platform increased by 13.6% year-on-year to 12.445 billion yuan. The average monthly active users were 44.6 million, a year-on-year increase of 17.1%; the average monthly active drivers were 758,000, a year-on-year increase of 36.8%.
Worryingly, Caocao Travel’s net cash flow from operating activities in the first half of 2026 was only 65.99 million yuan, a drop of more than 80% from 340 million yuan in the same period last year. The company’s explanation for this is: This is a change in working capital caused by the time difference between receipts and payments. At the same time, Caocao Travel’s cash reserves have also shrunk significantly, from 2.33 billion yuan in the first half of 2025 to 1.56 billion yuan in the first half of 2026.
As of the end of June 2026, Caocao Travel’s total liabilities were 9.381 billion yuan, total assets were 5.221 billion yuan, and the asset-liability ratio was as high as 179.69%. It is highly dependent on Geely Holdings and external financing channels to continue its life.
This can be seen from Cao Cao’s fundraising behavior at the beginning of this year. On January 28, Cao Cao Travel issued an announcement on the placement of new shares, intending to raise HK$383 million in additional shares at a discount. This fundraising comes just 7 months before its IPO.
Returning to Caocao Travel’s revenue structure, in the first half of 2026, the company’s core travel service segment revenue increased by 13.9% year-on-year to 9.79 billion yuan, but vehicle sales revenue plummeted by about 60% to 300 million yuan.
Caocao Travel’s travel services are highly dependent on aggregation platforms such as AutoNavi and Baidu. From 2022 to 2024, the GTV of orders received by Cao Cao Travel from the aggregation platform accounted for 49.9%, 73.2% and 85.4% of the company’s total GTV in the same period respectively. Cao Cao Travel told Caijingtianxia that since 2023, the company has significantly increased its cooperation with aggregation platforms, and it has devoted more management energy to improving service quality and brand awareness.
Although relying on aggregation platforms can quickly leverage massive orders, this is by no means a free gift, and there are extremely high channel costs behind it. In the first half of 2026, the commission paid by Cao Cao Travel to the aggregation platform was 830 million yuan, accounting for 6.7% of its GTV. This is a gross profit margin bottleneck that Cao Cao Travel cannot break through. The company can only save other expenses to reduce losses.
The financial report shows that in the first half of 2026, Caocao Travel’s general and administrative expenses and R&D expenses were 290 million yuan and 100 million yuan respectively, a year-on-year decrease of 37.0% and 14.0% respectively. The two expense ratios dropped by 2.1 and 0.3 percentage points respectively, saving a total of approximately 190 million yuan.
02. Is it too late to join RoboX now?
China’s online ride-hailing industry has experienced more than 10 years of development. The growth in user scale has basically peaked, and the industry has entered the era of stock. In such an almost saturated market, the focus of competition has shifted to Robotaxi. Platforms such as Didi and T3 have launched their deployments, and Caocao Travel is no exception.
In mid-June this year, Caocao Travel announced at the 2026 International Automobile and Supply Chain Expo (Hong Kong) that it would officially launch a comprehensive AI transformation and release a new RoboX strategy. In the new strategy, Caocao Mobility is no longer positioned as an online ride-hailing platform, but as a physical AI mobile technology platform, serving not only people, but also “things.”
According to the plan, by 2030, Caocao Travel will deploy a total of 100,000 Robotaxi and 100,000 Robovan (cars for unmanned cities). In April this year, Cao Cao’s Robotaxi prototype Eva Cab was unveiled at the Beijing Auto Show and is expected to roll off the production line in June 2027.
Before the mass production of Eva Cab, Caocao Travel has been conducting verification work around vehicle dispatching, energy replenishment operation and maintenance, and remote security. In February last year, Cao Cao Travel deployed the first-generation Robotaxi in Hangzhou and Suzhou. This year it has iterated to the second-generation, with about 140 vehicles. Caocao Travel said that the current overall business progress is in line with expectations.
Compared with friends and businessmen, Cao Cao’s travel progress is a bit slow. As of the second quarter of 2026, Luobo Kuaipao’s Robotaxi business has covered 28 cities around the world, with cumulative public travel service orders exceeding 23 million. It has achieved a bicycle economy (UE) break-even in Wuhan. WeRide has covered more than 60 cities in 13 countries around the world, and holds autonomous driving licenses in 8 countries. The company’s business in some overseas regions has achieved staged operating profits. As of the first half of 2026, Pony.ai’s global fleet will be close to 2,000 vehicles, and the App’s registered users in China will exceed 1.5 million. Didi, which is also an online ride-hailing platform, is currently continuing its investment and aims to expand its fleet to more than 1,000 vehicles by the end of this year.
In the Robotaxi competition, the gap between players is not only reflected in explicit numbers such as fleet size, but also in their respective cost reduction paths. Wang Haojun, co-founder and CFO of Pony.ai, once explained that the cost of Robotaxi is roughly divided into two parts: hardware depreciation cost and operating cost, each accounting for about 50% of the cost. Domestic technology companies that entered the Robotaxi track earlier, some models have been iterated to the sixth and seventh generations, and many optimizations have been made in terms of bicycle costs.
The BOM (bill of materials) cost of the ADK (autonomous driving kit) of Pony.ai’s seventh-generation Robotaxi has been reduced by 70% compared with the previous generation. By 2027, the total cost of its fully unmanned Robotaxi vehicle can be controlled within 230,000 yuan. The cost of a single car of the sixth-generation mass-produced unmanned vehicle RT6 of Luobo Kuaipao has dropped to 200,000 yuan, a drop of more than 60% compared with the previous generation. Caocao Mobility is backed by Geely Holdings, and the cost of its Eva Cab bicycles is also very competitive.
In addition to hardware costs, there are operating costs. Some people may think that one of the most important capabilities of online ride-hailing platforms such as Didi, Caocao Chuxing, and Ruqi Chuxing is operations. In the field of Robotaxi, their operational capabilities are simply a dimensionality reduction blow to players in other tracks.
However, Lu Liang, a senior practitioner in the autonomous driving industry, told Caijingtianxia that Robotaxi’s operating system is different from the traditional operating system, and the online car-hailing platform does not have inherent advantages in this regard. “The remote takeover and ground handling systems must be rebuilt.”
An executive from an L4 autonomous driving technology company also expressed a similar view. He believes that both online ride-hailing and Robotaxi need to clean and maintain their vehicles, face traffic police, and deal with passenger emergencies. The human driver of an online ride-hailing service can handle all situations. The latter does not have a driver, so it is more difficult to handle emergencies. “For car manufacturers and travel platforms, Robotaxi is actually a new species and cannot be said to be automatically grafted,” the executive said.
Industry insider Liu Xiao told Caijingtianxia that Robotaxi is a very challenging project for new players. First of all, the security and technical thresholds of this industry are high, and first-mover companies have deeper accumulation and know what “good data” is, and “know-how is very important.” Secondly, cold-starting a fleet of 1,000 Robotaxi vehicles from scratch will require huge costs in terms of vehicles, R&D, and operations teams. However, Liu Xiao believes that there is still a chance for Cao Cao Travel to enter the game now. “New players can learn from mature experience and shorten the self-driving demo time from stage 0 to stage 1.”
Liu Yu, a former employee of an L4 autonomous driving company, also believes that it is not too late for Cao Cao Travel to enter the Robotaxi field. “The entire industry is far from saturated.” Moreover, Cao Cao took an “advantage” when traveling to do Robotaxi.
Caocao Travel said that in the Geely system, the company is first and foremost an independently operated technology travel platform. It is also an important business platform for Geely Holding Group in the field of smart travel and future mobility services. It plays a role in connecting vehicle manufacturing, smart driving technology and real travel operation scenarios. It is also the most important commercial carrier of the group’s RoboX strategy. In other words, Cao Cao’s RoboX strategy is backed by the Geely system, with Geely Automobile as its manufacturing partner and Qianli Technology as its technology.
“Cao Cao’s travel relies on the Geely system, and its original customized models combined with large-scale fleet procurement can narrow the cost gap with its friends.” Lu Liang analyzed, “But the premise is that the scale of the vehicle installation is sufficient, otherwise the scale effect cannot be released.”
On the one hand, there is a very oppressive financial status quo, and on the other hand, there is the grand transformation vision of RoboX. What lies in front of Cao Cao Travel and Gong Xin is a test question full of temptations and challenges.
(Zhou Hui, Lu Liang, Liu Xiao and Liu Yu are pseudonyms in the article)
