Once hailed as China’s 'magic website' for everything from apartment hunting to finding a plumber, 58.com is now haunted by the very technology it hoped would save it. Recent reports of 'ghost orders'—where the platform automatically generates service contracts and leaks user data to unverified third parties—have triggered police investigations and exposed a systemic rot within the company’s business model. This scandal marks a new low for a firm that once dominated the country’s classified information market.
The mechanics of these ghost orders reveal a platform prioritizing lead-generation revenue over consumer safety. In one high-profile case, a user merely requesting a quote for appliance repair was suddenly swarmed by multiple uncoordinated technicians, some of whom lacked basic qualifications. Investigation found that 58.com’s backend automatically creates valid orders the moment a user clicks for a quote, selling these 'leads' to any business willing to pay for 'grab-order points,' regardless of their actual industry credentials.
This predatory practice is a symptom of a broader financial collapse. Between 2019 and 2024, 58.com saw nearly 10 billion RMB (approximately $1.4 billion) in annual revenue evaporate, falling from a peak of 15.58 billion RMB to around 6 billion RMB. Once a high-margin cash cow with 86% gross profits, the company has spiraled into consistent losses, forcing it to slash its workforce from over 40,000 employees to roughly 20,000 in a series of brutal layoffs.
The decline of 58.com is fundamentally a story of failing to adapt to the 'transactional' era of the Chinese internet. While 58.com remained a shallow directory of unverified listings, specialized rivals like Beike in real estate and Boss Zhipin in recruitment built deep, trust-based ecosystems that handle the entire transaction. By the time 58.com attempted to pivot, it had already lost the war for user trust, becoming a platform synonymous with scam listings and low-quality service.
Founder Yao Jinbo is now doubling down on a pivot to become an 'AI-native enterprise,' claiming the platform processes hundreds of billions of tokens daily. However, high computational costs and internal pressure to use AI have yet to yield a 'killer app' that can restore the company's reputation. For a platform currently under police scrutiny for administrative violations, the transition to AI may be a case of high-tech window dressing for a fundamentally broken middleman model.