A Chinese Robot Maker's Hardest Sale Is Now Its Own Board
Unitree, the Hangzhou-based robotics company that has become one of China’s most visible names in humanoid and quadruped machines, is discovering that going public changes the nature of the questions it faces. Its first earnings report since listing disappointed, yet its stock has rebounded roughly 70%, a split that tells you how much of the company’s value rests on expectations rather than current sales.
The immediate problem is not demand in the abstract. Chinese factories, warehouses and logistics operators are genuinely curious about legged and wheeled robots. The harder issue is that the buyers who matter most — large manufacturers — do not purchase novelty. They purchase uptime, service networks and a clear payback period. A robot that can walk across a showroom floor is not the same product as a robot that can work a night shift beside human staff without a technician on call.
That is why the earnings miss matters less than the composition of the company’s revenue. Publicly listed robotics firms in China are under pressure to show that their order books come from repeat industrial customers rather than one-off research purchases, media demonstrations or government-backed pilot programs. Unitree’s profile — strong brand recognition among technology enthusiasts, a broad product line spanning consumer and industrial machines — gives it options, but also makes it harder for investors to read where the durable money is coming from.
The broader Chinese robotics sector is entering a similar phase. Humanoid robot startups have raised heavily on the promise of embodied AI, where a machine learns tasks by observing and imitating rather than being programmed step by step. That promise is real, but deployment is slow. Factory managers interviewed about automation consistently describe a gap between demonstration videos and the reliability required for production lines, where a single failure can halt a shift.
Unitree’s advantage is that it sells actual hardware at scale and has built a recognizable brand. Its challenge is that the same enthusiasts who drive early sales do not necessarily become the industrial customers who drive stable margins. The company will need to show that its next set of orders comes from manufacturers who have integrated its machines into routine operations, not from buyers who want a robot on a promotional video.
For global readers, the interesting question is not whether Chinese robot makers can build impressive machines. They clearly can. The question is whether China’s manufacturing base, which is both the largest potential customer and the harshest testing ground, will adopt them fast enough to justify the valuations. That adoption curve will be decided by service contracts, spare-part logistics and training — the unglamorous parts of the business that rarely appear in launch events.
A stock that rises 70% after a weak report is a stock priced on belief. The next earnings season will begin to show whether the belief is converting into factory floors.