T97 turned one viral chant into 500 signed coffee stores — and lost ¥100M in five years

T97 Coffee was founded in Hangzhou in April 2021 by Li Xiao, a social-e-commerce entrepreneur whose membership platform had once reported tens of billions of yuan in gross merchandise value. In autumn 2022 the brand went viral on Douyin, propelled by an addictive livestream chant — "咖啡你冲不冲,冲冲冲冲冲" (coffee, do you rush it, rush rush rush). A single livestream drew 10.82 million views and monthly exposure passed 500 million. With only 70-odd stores at the time, daily revenue per store broke ¥3,000 and 90% of stores were profitable.

Li Xiao set out to outrun the market leaders. He promised to open 1,001 stores within a year and to surpass Luckin, China's largest coffee chain, within 35 months — "T97 first, Luckin second, Starbucks third." At its peak the brand had signed more than 500 stores and opened over 200, operated by 75 franchisees. The attention was real; the store economics were not.

By the founder's own account, T97 made four mistakes. It opened million-yuan-rent flagship stores in premium malls — two lost at least ¥8 million — before it had a standard model worth copying. It hired 150 people, nearly 20 in research, while the chain had only dozens of stores. It ran four site-selection teams for free, where rivals charged, spending over ¥800,000 a month on travel; 50 trips to Guizhou opened two stores. And it let franchisees who closed within a year keep their fees and sell back equipment at half price, so even profitable stores quit. Each closure cost about ¥200,000.

Five years and more than ¥100 million in cumulative losses later, T97 was down to 11 stores, according to the restaurant-tracking service Jianmen Canyan. Li Xiao's verdict on himself: he should have polished the standard store model before chasing hype.

T97 burned ¥100M in five years on million-yuan flagship rents, a 150-person team, and free site-selection, then paid franchisees to leave, collapsing 200+ stores to 11.

Viral attention was mistaken for demand. Livestream views converted into franchise signings, not proven unit economics, so the brand scaled its name before it had a store worth scaling.

Overhead ran ahead of revenue. Million-yuan flagship rents and a 150-person team were funded while the chain had dozens of stores — costs only hundreds of profitable stores could carry.

Free site-selection removed the discipline a fee imposes. It attracted signings but not viable locations; 50 trips to open two stores is the bill for that filter.

The exit policy paid stores to leave. Full fee refunds and half-price equipment buybacks made closing cheaper than staying, so even profitable franchisees quit early.

Viral attention signs franchises, not customers. Prove one store makes money before you sell a hundred — and never subsidise the exit, or your best partners will be the first to take it.

Li Xiao said slower hiring alone would have saved ¥50 million, and that his first mistake was chasing flagship-store hype before the standard store worked. T97's own website once claimed ¥1 billion in sales within a year of launch — a figure built on franchise signings, not cups of coffee.