Inside a barbecue restaurant in Changsha, a Dyson hairdryer sits tucked away in storage, the hair-washing station has been converted into a fried rice counter, and dust gathers on the DJ booth that once pulsed with energy at 8 PM—this location has cycled through five store managers in just over a year.

Several hundred kilometers away in Tianjin, internet celebrity Bai Xianying, known as "Zhou Bing Lun" for his resemblance to Jay Chou, sighs into his phone camera. On August 24th, his first store, which had operated for two years, closed after rent jumped from 130,000 RMB to 200,000 RMB annually. By September 1st, his Hongqiao location also shut down due to a fallout with his business partner. He warns, "You must be extremely careful when entering partnerships!" (Oriental Finance)

One was once the most promising "son" of Haidilao, the other a celebrity with millions of followers. They seem unrelated but died from the same underlying disease.

Part One: From Boom to Bust

Yan Qing was the first seed to sprout under Haidilao's "Pomegranate Plan." Its first store in Xi'an opened in late 2023 and was profitable immediately, offering services like hair washing, braiding, DJ sets, and live bar performances. The nightlife segment contributed roughly 15% of sales growth, and management boldly projected opening 400 to 500 stores within three years.

In the first half of 2025, it aggressively opened 46 new locations, bringing its total to 70—during the same period, Haidilao's other 13 sub-brands combined for only 56. (Food Circle/NetEase) Then came the sharp reversal: between November 2025 and the end of June 2026, Yan Qing closed at least 19 stores with zero new openings in the first half of 2026. All three Guangzhou locations shuttered, and it fully exited Shenyang, Zhengzhou, Tianjin, Jincheng, and Linhai. Only around 60 stores remain operational, and Haidilao's financial reports no longer disclose its performance separately. (True Story Research Room/Wumian Finance)

Haidilao's latest resource allocation priorities are now its hot pot food stall concept and Ru Yi sushi—Yan Qing is no longer on the list.

Part Two: Two Different Demises, One Shared Root Cause

The giant's failure stems from an inability to transplant its organizational strengths, as gimmicks fail to drive repeat business. Yan Qing completely copied Haidilao's playbook: opening next to Haidilao outlets to piggyback on their traffic, implementing full-service hair washing, braiding, and DJ entertainment, and mass-producing store managers through its mentorship system.

However, hot pot thrives on standardized broths and service, while barbecue's core lies in meat quality, marinades, and flavor identity. Yan Qing's menu was largely indistinguishable from typical barbecue joints, with an average spend of nearly 100 RMB offering little value. Customers described it as "paying for the service, not the barbecue." (Food Circle/NetEase)

More critically, each service—hair washing, braiding, DJ—required significant manpower. Assistants were needed for grilling, and a single hair wash took 20 minutes of dedicated staff time. The Changsha store burned through five managers in over a year; each new manager brought in their own team, pushing out veteran employees. Of the initial 20-plus staff sent to Xi'an for a month of training, only one remained. A piece-rate wage system was tried for three months before reverting to fixed salaries, leaving some employees earning just over 2,500 RMB monthly while covering the work of three or four roles. (True Story Research Room/Wumian Finance)

Consequently, services were gradually stripped away: braiding went from a dedicated position to part-time, and eventually stopped altogether; DJs and live performers left one by one. Negative reviews consistently cited "service that doesn't live up to Haidilao." The service advantage Haidilao is famous for became a manpower cost black hole in the barbecue sector.

The influencer's downfall illustrates that traffic cannot fix a flawed single-store model, and the partnership structure was doomed from day one. Zhou Bing Lun's 20-square-meter store carried a 200,000 RMB annual rent, translating to about 556 RMB per day. Factoring in utilities, labor, and ingredients, he calculated he needed to sell over 100 of his 8 RMB pancakes daily just to break even. (Chief Business Insight)

The Jay Chou lookalike's fame drove initial curiosity visits, but an 8 RMB pancake is a high-frequency, low-margin business that simply couldn't sustain prime location rent. His first store died from a rent mismatch; the second from partnership disputes. "He wanted more money, it wasn't working for us, he wouldn't work for it," and "If we're not doing it together, let's not do it at all"—without clearly defined roles, profit-sharing, or exit mechanisms agreed upon upfront, success quickly tore them apart. (Oriental Finance)

Both share a single root cause: Haidilao's brand name and millions of followers are essentially "traffic endorsements." They solve the problem of the first visit, but nothing about product repurchase, cost structure, organization, or partnership can be outsourced to endorsement.

By the end of 2025, the Pomegranate Plan had incubated 20 sub-brands, yet "other restaurant operating income" accounted for only 3.5% of Haidilao's total revenue—most had not achieved sustainable success. (Food Circle/NetEase)

Similarly, T97 Coffee, which once vowed to "surpass Luckin in 35 months," peaked at over 200 stores but now has just 12 remaining, having lost 100 million RMB over five years. Its founder admitted the biggest losses came from franchise guarantee policies, and it has pivoted to a convenience store model embedded inside supermarkets, bookstores, and gyms—a clear retreat toward "lower rent, borrowed scenarios." (Daily Economic News)

Part Three: 2026, The Dining Industry Deflates Its Bubble

This isn't just bad luck for two brands; it's an industry-wide debt repayment. Bakery stores saw a net decline of 81,700 outlets, with an average survival span of 32 months and 57% failing within two years. (China Business Journal) Small taverns face a closure rate exceeding 30%, with Helens' same-store daily sales down 21%. (Professional Catering Network) The clay pot rice category saw a 40.4% closure rate—the steepest drop among 80 monitored categories. (36Kr/Canbaodian)

According to the China Hotel Association, rent costs now average 11.1% of revenue in the food and beverage sector, rising 2.6% annually. (Chief Business Insight) The industry's logic has shifted from "who expands fastest" to "whose single-store model is genuinely profitable."

Here are five cold truths to consider:

First, big brands' sub-brands are not blueprints for an average person to copy. Yan Qing's closures represent part of the tuition paid from Haidilao's approximately 522 million RMB loan facility. (Food Circle/NetEase) Giants can afford to lose money, close stores, and pivot to new ventures; a closure is just "efficiency improvement" in their reports. But an ordinary person betting their life savings and borrowed money faces only a landlord's rent demands and a "for lease" sign on the window when a store fails.

Second, online buzz is rented, not owned. Zhou Bing Lun's viral fame came from a celebrity face; once the hype fades, the daily rent won't wait for the next trending topic. With small taverns seeing over 30% closures (Professional Catering Network) and clay pot rice at 40.4% (36Kr/Canbaodian), industry trends can vanish overnight—a celebrity face has an even shorter shelf life. When the wind stops, those without swimwear feel the cold first.

Third, partnerships with friends and family often die from an unwillingness to discuss money. How many brotherly ventures start with clinking glasses and end without anyone daring to open the books? Investment proportions, management roles, loss accountability, share buybacks, and decision-making authority must be settled in writing before opening. The better business gets, the uglier the breakup without these terms.

Fourth, beware of franchise guarantee schemes and influencer recruitment pitches. T97 once promised "full refund of franchise fees for closures within one year and 50% equipment buyback"—it sounded foolproof, but the 100 million RMB loss was mostly from these guarantees. (Daily Economic News) Such guarantees protect the brand's expansion metrics, not your principal.

Fifth, even a successful 0-to-1 model can collapse in the 1-to-10 phase. Yan Qing's 70 stores crumbled due to management limits: store manager replication lagged, veteran employees left, and service quality deteriorated location by location. (True Story Research Room/Wumian Finance) A single store's profitability relies on your craftsmanship and hands-on effort; scaling to ten stores demands organization, systems, and people. Many who earn 30,000 RMB monthly from one store rush to open three more and lose money across all of them—these are fundamentally two different businesses.

Four Pieces of Advice for Aspiring Entrepreneurs

First, calculate your break-even point before dreaming big: fixed costs (rent plus labor) divided by per-unit gross profit equals your daily break-even volume. Zhou Bing Lun's 100 pancakes were derived this way—after calculating, spend three days at the location counting real foot traffic. If the numbers don't work, walk away.

Second, low-ticket businesses should avoid high-rent locations. If you're selling 8 RMB pancakes or 15 RMB milk tea and rent exceeds 15% of revenue, you're effectively working for the landlord.

Third, be a "gentleman on paper, not just in spirit" in partnerships. Define capital contributions, role responsibilities, profit distribution, share buyback pricing, and deadlock resolution (who has the final say) in writing before opening. Ideally, pre-agree on an exit price for an amicable parting.

Fourth, don't idolize endorsements or traffic. Big franchise brands and personal follower counts only guarantee a queue on opening day. When evaluating a project, demand six consecutive months of actual financials from a currently operating store in the same city and trade area—never trust a single data point from a model store.

Endorsements and traffic bring people through your door once. The third visit depends on the product. Whether the store survives to year three depends on the numbers and the people.