He Xiaopeng: The Xpeng founder who bets, breaks, and rebuilds

In September 2022, Xpeng launched the G9, the SUV meant to show the world that a Chinese startup could compete at the top of the market. The engineering was sound. But the price list was a disaster — a tangle of trims and options so confusing that buyers could not work out what they were getting. Within days, the company was reissuing its prices. The episode showed how quickly a complicated product proposition can undermine an otherwise strong launch. Most founders would respond by replacing the marketing team. He Xiaopeng concluded that the problem was him. He is not a minor figure. He built China’s most popular mobile browser, UCWeb, and sold it to Alibaba for $4.3 billion. Now, he runs an electric-car company that delivered around 430,000 vehicles in 2025 while building humanoid robots, robotaxis, and flying cars on the side. What he did after the G9 is why any of that still exists, and it is the most useful part of his story for anyone running a business in a fast-moving market. A career built on one idea To understand why He keeps surviving, start with a browser. The 48-year-old co-founded UCWeb in 2004 with his friend Liang Jie in a China of cheap phones, expensive data, and slow networks, where pages built for desktops were unusable on the handsets most people could afford. UCWeb’s answer was not glamorous. Its browser squeezed pages on the company’s own servers before sending them on, so each arrived smaller, faster, and cheaper. It was plumbing, not invention, and it handed the internet to hundreds of millions of people who could not have otherwise afforded it. By 2014, UC Browser had over 500 million users. That year, Alibaba completed its acquisition of UCWeb in a deal widely valued at about $4.3 billion. That is the pattern the Chinese entrepreneur has repeated ever since: take something that works only for rich customers and squeeze the cost out until ordinary people can have it. Walking out of the gilded cage The Alibaba years left him wealthy and, by his own description, restless. He has called that period a gilded cage: comfortable, secure, and not what he wanted to be doing. In 2017, he left Alibaba to join Xpeng, the electric-car startup he had co-founded in 2014, as its CEO and chairman. He was 39 with no traditional automotive background, entering a Chinese EV market crowded with startups. To stand out, he treated the car as a software product, building Xpeng’s own operating system and assisted-driving software when licensing both would have been cheaper. Two moments of reckoning The first, in 2019, was the ordinary kind: Xpeng was running short of capital as China’s EV funding market tightened. The second, in 2022, was the G9 launch described at the start of this piece: a sound car undermined by a price list nobody could decipher. It was a public embarrassment for a company whose reputation rested on product judgment. He did not defend the decision or wait for the next model to fix it quietly. In October 2022, he began a broad organizational restructuring, and in January 2023, he hired Wang Fengying as president. Wang, who brought more than 30 years of experience at Great Wall Motor, became responsible for product planning, product portfolio management, and sales operations. He has since called that stretch the darkest moment of his working life. The striking part is not the failure; it is that a founder, in a culture that treats the founder’s instinct as close to sacred, looked at the wreckage, decided his own instincts were the weak point, and bought better ones. The comeback: A cheap car with expensive software The recovery arrived in August 2024 with a car called the MONA M03. It started at RMB 119,800 ($16,800), roughly half the price of a Tesla Model 3 in China, and carried the assisted-driving software that until then had justified premium prices. Ten thousand firm orders landed in 52 minutes. From January to September 2025, the M03 had delivered nearly 132,000 units, accounting for 42% of Xpeng’s deliveries. Something similar was happening elsewhere. In 2023, Volkswagen invested about $700 million in Xpeng and struck a technology partnership that has since expanded to the electrical architecture underpinning a wider range of its China-market vehicles. One of the world’s largest carmakers was buying software from a startup that had nearly died twice. Three impossible things at once The instinct that built UC Browser had been pointed at a car. Now, it is being aimed at three industries at once. He has told staff that 2026 is the year Xpeng pushes into mass production of humanoid robots while advancing robotaxis and flying cars. Its Land Aircraft Carrier began trial production in November 2025, while the company announced plans to launch three robotaxi models in 2026. Iron, the humanoid, is aimed at malls rather than factories; on its first outing at a Shenzhen mall in February 2026, it fell over. He compared it to a child learning to walk. Though the ambition is clear, the core business is still being tested. Xpeng’s sales fell year on year for five straight months in early 2026 before turning positive in June. The CEO believes only a handful of major car companies will exist in the next decade. What his career actually teaches Cheap is a technology strategy. UC Browser (the web browser developed by UCWeb) and the M03 are the same move 20 years apart: engineer the cost down until the market widens. In China, price is often the point of innovation, not the opposite of it. A good product can be killed by a price list. The G9 failed on presentation, not engineering. Founders who see themselves as product people are often weakest where the product meets the customer’s ability to understand it. Know which part of yourself to replace. Hiring Wang Fengying was not delegation; it was a founder identifying his own weakest function and handing it to an outsider while he still had the choice. Very few founders manage this. How China innovates His career is also a map of how Chinese innovation works, which is often misread from outside. The first misreading is about invention. Browsers, electric cars, humanoid robots, and small aircraft were all invented elsewhere. What He Xiaopeng has done three times is compress them — in cost, time, and scale — until they reach a mass market. That drives Chinese competitiveness more than invention does, because it attacks your margin rather than your patents. The second is about speed. He rebuilt his entire executive bench in about a year. At most large Western companies, that surgery takes three years and a consultancy. Tearing up the organization publicly is a real advantage, at a cost in stability that Chinese founders seem willing to accept. The third is about weather. He did not create China’s EV wave; state subsidies and industrial policy did. He read it early, rode it, and survived the price war that followed when that same policy environment turned unforgiving. Chinese founders operate inside a system that can build their market and then crush it, making timing and endurance as important as vision. For executives outside China, the lesson is uncomfortable. The competitor worth worrying about is rarely the one with the most polished launch. It is the one that has already failed in public, rebuilt itself in the open, and returned at a price you cannot match. The bottom line Put the pieces together, and the pattern has little to do with vision. Plenty of people saw electric cars coming, and plenty saw humanoid robots. What separates He Xiaopeng from the founders who did not survive the same decade is a willingness to be wrong in public, to say so quickly, and then to change the thing that was wrong, including, when the evidence pointed there, himself. The lessons above are all versions of that. The browser worked because he engineered around what customers could not afford. The G9 taught him that a good product dies on a bad price list, and his fix was not a better presentation but a different person running the commercial side. The M03 worked because he stopped trying to prove Xpeng belonged among premium brands and built the car people could actually buy. Each correction cost him status, control, or the company he had imagined. So the question his career puts to any executive is not whether the strategy is right. It is how fast the organization can admit the current one has stopped working and how much of yourself you are prepared to replace when it has. He is now betting that robots, flying cars, and robotaxis will enter millions of households. That bet may not land. On his record, the thing worth watching is not whether he is right but how quickly he will admit it if he is wrong.

He Xiaopeng: The Xpeng founder who bets, breaks, and rebuilds

How Bottega Veneta weaved China’s forever it bag

Upgrade to Pro

Sectors

Insights

Company

He Xiaopeng: The Xpeng founder who bets, breaks, and rebuilds

How Bottega Veneta weaved China’s forever it bag