Supermarkets divide into two types: selling products succeeds, selling shelf space dies
Ye Guofu argues that Sam's Club and Pang Donglai (胖东来) succeed by selling quality products, while other supermarkets profit from entry and shelf fees—this logic drives his acquisition of Yonghui (永辉) and his plan to replicate the Pang Donglai model.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
A good store reveals itself in one hour
Ye Guofu says there's no mystery to retail: you don't wait three months to judge a store—one hour is enough. Watch 100 customers walk in: how many buy? Miniso's store-entry conversion rate is now 30%, exceptionally high for retail. He uses this to counter the claim that offline loses to e-commerce. Offline store metrics are visible and can't be faked, unlike e-commerce sales figures, which can be inflated through bot traffic. Customers and competitors see the truth daily. This transparency is his confidence—growing through word-of-mouth alone, with minimal ad spend.
— Ye GuofuFranchisees only handle money and location; headquarters handles everything else
Miniso's franchise model borrows from five-star hotels: franchisees do exactly two things—find a location and fund it. Everything else—design, inventory, pricing, promotions, checkout systems, logistics—headquarters handles. Franchisees don't need to understand retail; they don't manage operations at all. This division of labor lets headquarters rapidly scale: store managers train in about three months and are ready. Franchisees earn 38% gross margin, with payback in one to one-and-a-half years on initial investments ranging from ¥1.5M to ¥10M. This specialization is how Miniso expanded thousands of stores in a single year.
— Ye GuofuThe precooked-meal debate accidentally became free advertising for dine-in restaurants
Luo Yonghao's earlier criticism of precooked meals drew fire from some who claimed he was harming offline dining, a draw for shopping centers. Ye Guofu directly pushes back: forcing chain restaurants to strip the fake "made fresh now" claims and actually cook hot only makes offline dining more appealing—a gift to offline retail traffic. Both men also note that high precooked-meal penetration in Western countries shows up in supermarkets and fast-casual chains, not fine dining restaurants. Using that data to defend precooked meals in China is misleading the public.
— Ye GuofuPop Mart's supply cut forced them to create original IP
Miniso had ten years of smooth sailing licensing international IP. Ye Guofu didn't feel pressure to build original IP until he saw what happened to Pop Mart (泡泡马特): the overseas IP licensor abruptly cut off supply. Pop Mart's sales plummeted from billions to hundreds of millions; the company wavered on whether to bet on homegrown IP. But that bet paid off. After watching the model stabilize, Miniso committed fully to original IP in early 2025, signing 17 exclusive artist IP partnerships in under six months. The first homegrown IP alone hit ¥40M in sales its first year.
— Ye GuofuEven Sanrio can't replicate a second Hello Kitty
Ye Guofu points out that super-IPs emerge by accident, not design. Sanrio has existed 52 years and launched hundreds of IP properties. To this day, Hello Kitty remains the company's dominant blockbuster, accounting for 30% of revenue—and Sanrio has never produced a second IP of equal magnitude in 52 years. His conclusion: reliably generating a portfolio of mid-tier profitable IPs is something methodology can achieve. But birthing a cultural phenomenon IP is luck. All a company can do is sign more IPs, test them quickly, kill slow ones fast, and multiply shots on goal—trying to engineer a blockbuster is futile.
— Ye GuofuTwo collectibles giants have raised barriers so high a third can't emerge
On the future shape of China's collectibles market, Ye Guofu predicts a two-giant structure like Disney and Universal Pictures in the US, not Japan's fragmented model with many players. Miniso and Pop Mart are far ahead now, and both have scaled into global infrastructure and distribution. The barriers they've built are too high for a third or fourth platform giant to rise. Still, the market is large enough that smaller creative studios can thrive in the long tail, coexisting with two titans and many small workshops.
— Ye GuofuHe decided Yonghui could be saved based on one store's 13× overnight surge
Ten years ago, touring Costco in the US, Ye Guofu planted a seed: China needs a quality supermarket. He never acted until witnessing one number: Pang Donglai (胖东来) overhauled a single Yonghui flagship store, and same-day sales jumped 13-fold. That metric convinced him the quality-supermarket model could work in China. He moved quickly, acquiring Yonghui as its largest shareholder. Notably, he didn't know Pang Donglai beforehand and never confirmed in advance whether the founder would keep helping. The decision was pure conviction.
— Ye GuofuRestoring a struggling supermarket begins with an immediate leadership overhaul
After acquiring Yonghui, Ye Guofu's overhaul followed three steps. First: personnel. On day one, he replaced both the CEO and the chief product officer, the latter with a professional hired from Sam's Club—the prior leadership's vision couldn't lead the team forward. Second: supply chain. He cut suppliers from 8,000 to 4,000, targeting ≤2,000 in the future, and publicly promised suppliers: people can change, but suppliers can't. Third: store portfolio. He shut underperforming locations that didn't fit the target model. His benchmark is Costco: Kirkland (Costco's own brand) accounts for 40%+ of Costco's revenue, and Yonghui will replicate that model.
— Ye GuofuIn their own words · checked verbatim
Our store-entry conversion rate today is 30%—one hundred people walk in, thirty people buy something.
到明朝优品到今天 我们的进店转化率是30% 进来一百个人有30个人买单
Ye Guofu10:09
This is the heart of human life—what consumers truly want in hot-cooked dishes: not precooked meals, not central-kitchen stuff, not dishes made months or a year before.
这是人间烟火啊 是 热炒的消费者真正想要的 是 不是预制菜 你的中央厨房 或者是几个月一年前做好的
Luo Yonghao53:24
I say this year, in IP acquisition, if you haven't burned through a hundred million yuan, you've failed—only through wasteful spending do you get better outcomes, only waste leads to growth and better products.
我说今年 在收购IP上面 你不过浪费一个亿 就失败 只有浪费了 才能有更好的 才有成长 才有更好的东西
Ye Guofu1:20:37
Sanrio has existed 52 years, and to this day their biggest blockbuster is still Hello Kitty. There's no second one.
三立欧这个公司已经50多年了 今年52年历史了 52年历史了 到现在依然是 第一大爆火的 还是Hello Kitty 没有第二个
Ye Guofu1:39:47
The seed was planted ten years ago—I wanted to build a quality supermarket in China if I got the chance.
因为我是 这个种子是 十年前我都种下来了 我就想 有机会 能在中国干一个有品质的超市
Ye Guofu2:18:08
That's why Sam's Club and Pang Donglai are selling products—other supermarkets are selling shelf space. That's the fundamental difference.
所以 山姆和胖东南 是卖产品的 其他超市 是卖货架的 他们本质区别在这里
Ye Guofu2:21:08
From now on at Yonghui, you can change people, but you can't change suppliers.
不要因为过去换一个采购员 换一批供应商 过去所有 就是换一个采购员 换一批供应商 我说 以后 未来永辉 人可以换 供应商 不可以换
Ye Guofu2:48:22
Figures
| Miniso franchisee gross margin | 38% | 13:10 |
| Miniso franchise payback period | 1–1.5 years | 24:15 |
| Miniso store-entry conversion rate | 30% | 10:09 |
| Miniso Shanghai Nanjing Road flagship store, 9-month revenue | ~¥100M | 13:10 |
| Pang Donglai's overhaul of a Yonghui flagship store, same-day sales increase | 13× | 2:16:07 |
| Costco's own brand (Kirkland) annual sales | ~$80B (40%+ of total revenue) | 2:30:14 |
| Yonghui suppliers (pre-overhaul to post-overhaul) | 8,000 reduced to 4,000, target future: ≤2,000 | 2:48:22 |
| Miniso global stores | ~8,000, split roughly 50/50 domestic and international | 2:05:00 |
| Per-capita collectibles IP spending | China: ¥51, Japan: ¥580, US: ¥2,077 | 2:04:00 |
Glossary
- SKU / Stock Keeping Unit
- A metric for counting product variety; one SKU represents one distinct product type.
- Fine dining
- Upscale restaurants featuring fresh-prepared dishes rather than precooked meals.
How to listen
Founders and investors seeking to understand offline retail fundamentals, how to monetize IP and collectibles, and the specific mechanics of supermarket supply-chain transformation.
The opening (0:00–1:19) discussing personal shopping preferences has low information density and can be skipped.