Franchising is sold to foreign brands as a shortcut into China. Someone else pays for the fit-out, hires the staff and carries the lease, and you collect a fee. I have sat in enough of these meetings since 2012 to say it plainly: it is not a shortcut. Franchising is the hardest expansion model to run well in China and the one that fails most often for foreign brands. The regulation is stricter than people expect, contract enforcement is weaker than they expect, and the unit economics in 2026 are much tighter than the pitch deck says.
I am Olivier Verot, founder of Gentlemen Marketing Agency in Shanghai. I have spent thirteen years advising European and American brands on China entry, and I have watched clients sign master franchise deals, lose control of their trademark, and buy their own concept back. This article is what I tell them before they sign, not after.
How China got here
Franchising in China is younger than most people assume. In the 1990s there was no clean Mandarin translation for the word. The closest term was “chain stores”. In 1997 the government set up the China Chain Store and Franchise Association (CCFA) to bring some order to the sector, and it now counts hundreds of members operating well over a hundred thousand outlets.
The model does offer real rewards. A local operator starts their own business while buying an established name and a system that already works. The cost is not small: a single unit license for a recognised foreign brand runs upward of $50,000, and a regional development license in China can approach a million dollars.
The 2026 numbers, and what they actually say
Two datasets from this year tell you everything, if you read them together.
- The China Franchise Business Development White Paper 2026, published in January 2026 by 红餐产业研究院 with 鹏友群, puts the Chinese chain and franchise sector at roughly 2.77 trillion yuan for 2025, up 4.8% year on year.
- The same white paper found that nearly 90% of Chinese franchisees now run two or more brands at once, and that 77.1% treat franchising as their main occupation and main source of income.
- CCFA’s China Chain Top 100 ranking published in June 2026 reports 2.07 trillion yuan of sales for the top 100 chains in 2025, down 2.7%, across 289,000 stores, up 12.4% and 32,000 outlets more than the year before.
Read that last line again. Store count up 12.4%. Sales down 2.7%. Even on the adjusted same sample it is 5.8% sales growth against 6.3% store growth. Sales per store are flat at best. The market is not growing fast enough to absorb the outlets being opened, and the people opening them are professionals running several brands in parallel, not devoted brand ambassadors. That is the environment your first Chinese franchisee works in.
The regulation is not a formality
Franchising in China sits under one specific text, the Regulation on the Administration of Commercial Franchises (商业特许经营管理条例), in force since May 2007, plus two implementing measures on filing and on information disclosure, the filing one revised again at the end of 2023. Most foreign brands I meet have never read any of them. Here is what they impose.
You must own a business resource registered in China
You need at least one trademark, patent, design patent, copyright or equivalent registered in the PRC before you can franchise anything. Your European trademark does not count. And the trademark license itself has to be recorded with CNIPA. If you have not filed your marks in China yet, you are not ready to franchise, you are ready to be squatted.
The “2+1” rule kills the shortcut
Before you can grant a single franchise, you must have owned and operated at least two outlets for at least one year. Two units, twelve months, on your own money. That is the 2+1 rule, and it exists precisely to stop what the shortcut crowd wants to do: land, sell licenses, cash in, leave. The outlets can sit with a subsidiary or an affiliate, and foreign franchisors can sometimes evidence compliance with units operated outside China through a trade association statement, but the intent of the rule is clear and Chinese courts read it that way.
This rule arrived with China’s WTO commitments, when foreign companies were finally allowed to develop without a joint venture partner. Brands complained about the cost. They were right that it is expensive, wrong that it is pointless. Look at who did it properly. KFC and Pizza Hut entered in 1987 and 1990, McDonald’s in 1990, all company-operated or in joint ventures with a Chinese partner. KFC started franchising in 1992. McDonald’s waited until 2004. Fourteen years of running their own stores before letting anyone else touch the brand. That is the actual playbook, and nobody wants to copy the boring part of it.
MOFCOM filing, and the annual report people forget
You file with MOFCOM within 15 days of signing your first franchise contract, through the commercial franchise information system. Foreign franchisors file centrally in Beijing rather than with a local bureau. Any change to the filed information goes in within 30 days. Every year in the first quarter you file an annual report listing contracts signed, terminated and renewed, plus the number and turnover of franchised units and of your own units.
The direct penalty is small, up to 100,000 yuan, and MOFCOM rarely chases it. That is why people ignore it, and why it hurts them later. An unfiled franchisor is a weak plaintiff. When your franchisee stops paying royalties, the first thing their lawyer raises is your own non-compliance. I have seen a brand lose a straightforward royalty claim on that alone.
Disclosure and the termination window
You must hand the candidate a written disclosure at least 30 days before signing, covering the items listed in the Information Disclosure Measures: registered business resources, existing network, unit investment figures, litigation history. If you hide something or overstate it, the franchisee can walk away from the contract, and no clean time limit protects you.
On the cooling-off period, be precise. Chinese law does not fix a statutory number of days. Article 12 of the Regulation requires the parties to write into the contract a period after signing during which the franchisee can unilaterally terminate. The length is negotiated. If your Chinese lawyer drafted a generous window because it looked friendly, you have handed your counterpart a free option on your brand. Set it short, set it clearly, and start it from signature, not from opening. The legal detail is laid out well in the ICLG franchise report on China.
When your franchisee opens across the street
This is the scenario nobody plans for and everybody should. Your franchisee runs your concept for eighteen months. They now know your supplier list, your recipes or your product specs, your staff training, your store layout, your unit economics. Then they drop the sign, keep the fit-out, rename the business, and reopen. Sometimes in the same street.
IP protection has improved a lot in China, and the courts are now genuinely usable for trademark cases. But a trademark protects your name and your logo. It does not protect a business format. Recipes, layouts and operating manuals are trade secrets, which means they are only protected if you treated them as secrets: numbered documents, signed confidentiality annexes, restricted access, and a non-compete with a defined radius, a defined duration and compensation, because an uncompensated non-compete is often unenforceable here. Starbucks had to litigate its own name in China in 2006 and won. Most brands do not have that legal budget, so they have to win at the drafting stage instead.
Practical defence, in order of usefulness: register your marks in every relevant class before any negotiation, keep one critical input under your control (an ingredient, a component, software, a certification), and never hand over your full supplier chain in year one. If they can buy everything without you, they will eventually try.
The F&B closure rates nobody puts in the pitch deck
Most foreign franchise projects in China are food and beverage. So the honest question is how long a Chinese F&B unit survives. The research house NCBD (餐宝典) measured a closure rate of 22.66% across the sector for 2024-2025, on a study covering 81 sub-categories. More than one outlet in five closed within the year.
The spread between categories matters more than the average. Crayfish tops the list at 37.2%. Bullfrog, braised chicken and malatang all exceed 31%. Seven categories sit between 29.4% and 37.2%, which are the trend-driven ones. At the other end, coffee closes at 10.3% and Anhui cuisine at 8.9%. The lesson is not subtle. Concept-led, hype-led formats die. Habit-led formats with daily repeat purchase survive.
Meanwhile the chain penetration rate in Chinese food service climbed from about 19% in 2021 to 23% in 2024. More chains, more units, same customers. If your unit-level model in China assumes European footfall and European average tickets, rebuild it. And if your franchise fee only pays back over five years, look at that 22.66% again and ask who is really carrying the risk.
Why master franchise usually ends badly
The master franchise looks perfect on paper. One partner takes a whole country or region, pays a large upfront fee, and builds the network for you. Here is why it goes wrong so consistently.
- The upfront fee becomes the deal. The master pays a large sum, so their priority is recovering it. They recover it by selling sub-franchises fast, not by making units profitable. Bad locations get approved. Bad operators get approved.
- You lose the customer. The master owns the WeChat account, the Xiaohongshu account, the member database, the Dianping listings. When the relationship ends, you keep the trademark and they keep every customer you paid to acquire.
- Territories are too big. China is not one market. Licenses here are granted city by city or province by province for good reason. Wealth, taste and channel behaviour differ enormously between Shanghai, Chengdu and a tier-3 city in Henan. A single master rarely operates well in all of them.
- Exit is expensive. Terminating a master with sixty sub-franchisees underneath means sixty contracts you did not sign and cannot easily assume. Brands end up buying the network back at a price set by the person who failed to build it.
The one time master franchise works is when the master is an operator with existing units in your category, not an investor with capital and ambition. Ask what they run today. If the answer is a portfolio rather than a business, walk.
A French case, and what changed the outcome
A French bakery and café group came to us with a signed master franchise for eastern China. Bertrand, the founder, had taken a 1.8 million yuan upfront fee from a Shanghai investment company and a commitment to 40 stores in three years. Eighteen months in, 11 stores were open and 4 had already closed. Average monthly revenue per surviving store was about 210,000 yuan against a 340,000 yuan break-even the partner had modelled. Nobody in the network could explain the gap.
The first fix attempted was marketing spend. The partner pushed roughly 600,000 yuan into Dianping promotions and group-buy vouchers over two quarters. Traffic rose, average ticket fell by 22%, and margin got worse. Discount buyers do not come back at full price. That is not a China problem, it is a discount problem, and it happens everywhere the operator has no other lever.
What worked was structural. The group bought back the three best-located stores and ran them directly. Those three became the reference: real product, real staffing ratio, real numbers. Then they rewrote the sub-franchise contract around operational control instead of fee collection. Central supply for the frozen dough and the coffee beans, POS integration so head office saw daily sales per SKU, and a store manager who spent four weeks in the company-owned units before opening. On the demand side we moved the customer relationship to brand level: one official WeChat account holding the member base and the coupons, and Xiaohongshu content produced centrally with local KOC seeding around each opening, so a new store opened with search results already in place instead of an empty feed.
Fourteen months later the network was at 19 stores, average monthly revenue per store had moved to roughly 305,000 yuan, and two of the three company-owned units were the top performers, which is what makes the model credible to new franchisees. Not a spectacular turnaround. A survivable one. The reason it worked is that the brand stopped selling licenses and started running an operation that licenses could be attached to.
The demand side is real, and that is what fools people
None of this means the opportunity is fake. There is a large urban customer base of young dual-income households with real discretionary spend, and foreign brands still carry a quality premium in several categories. Western franchises also arrive with what Chinese operators actually want: a finished system, training material, product design, a supply chain.
Adaptation works too. McDonald’s added rice, chicken and local spice profiles and saw returns, and the reverse works as well, putting ice cream or specialty coffee onto a Chinese format. Beyond food, the sectors that took to franchising here are the ones with an obvious local operator profile: business services from the 1990s onward, hotels from Hilton to Super 8 that became default stops for domestic business travellers in tier-2 and tier-3 cities, real estate agencies like Century 21 and RE/MAX during the housing boom, children’s fitness, and management training.
The demand being real is exactly what makes the shortcut tempting. Good market, easy money, someone else’s capital. That is how brands end up signing the wrong partner.
The two problems the model never solves for you
Staff. Turnover in Chinese retail and food service is high, training costs are real, and trained employees leave quickly because someone down the road pays 400 yuan more. Experienced local store managers are scarce and expensive. Franchising does not solve this, it moves the problem to someone with less money to fix it than you. Budget for a proper training pipeline or accept inconsistent execution. The corporate training market in China exists for a reason.
Partner selection. Vetting franchise candidates is the hardest part of the job, and most foreign brands do it badly because they are flattered by whoever shows up first with money. Use your national chamber of commerce, the commercial section of your embassy, law firms with a real China practice, and CCFA itself. Then check what the candidate operates today, visit their stores unannounced, and talk to their suppliers. Anyone who resists that is telling you something.
What actually works
Three models hold up in China in 2026. Pick one deliberately.
Wholly foreign-owned units first
Set up your own entity, open two or three units in one city, run them for a year or two, and get the unit economics right before anyone else touches the brand. Expensive and slow, and it satisfies the 2+1 rule as a side effect. It is also the only way to know what a franchisee can realistically earn, which is the number your whole offer rests on. Practical questions are covered in our guide to setting up a company in mainland China.
Joint venture with an operator
Share the capital and the risk with a Chinese partner who already runs stores in your category. You keep board control or at least veto rights on brand, product and site selection. The partner brings landlord relationships, which in Chinese malls is worth more than money. This is what the early fast food entrants did, and it is still the fastest safe route. The whole thing lives or dies on the negotiation, so read our guide on negotiating with Chinese partners before your first meeting, and if you are raising local capital alongside it, how startups attract Chinese investors applies to franchise projects too.
Franchise with tight operational control
If you do franchise, franchise like an operator, not like a licensor. Small territories, city by city. Central supply of at least one critical input. POS integration with daily data flowing to head office. Mandatory training in your own units. Site approval by you. Ownership of the WeChat official account, the member database and the Xiaohongshu presence at brand level, with franchisees given local content rights but never the account. A single franchisee cannot build search visibility on Xiaohongshu or a private domain community on WeChat, and if twenty of them try you get twenty brand voices. Driving footfall is a brand-level job, same logic as attracting Chinese shoppers to your store.
One thing changed recently. Chinese buyers researching a franchise, and consumers researching your brand, increasingly ask an AI assistant first. DeepSeek, Doubao and Kimi answer from indexed Chinese-language sources: Baidu Baike, Zhihu threads, industry media, Xiaohongshu posts. If nothing credible exists about your brand in Chinese, the assistant says nothing or repeats whatever a competitor published. Documenting your brand and your standards in Chinese, on the sources these models read, is now part of franchise development.
My position, stated plainly
Franchising in China is a scaling tool, not an entry tool. Used to scale a model you have already proven locally it works, and a 2.77 trillion yuan sector proves Chinese operators want good systems. Used to enter, it hands your brand, your know-how and your customer relationship to someone whose incentives diverge from yours in month one. The brands that win here spend two boring years running their own stores. The ones that lose sign a master franchise in month six and spend three years in arbitration.
FAQ
Can I franchise in China without setting up a Chinese entity?
Technically yes. A foreign franchisor can grant franchises directly and file with MOFCOM in Beijing. In practice it is a bad idea. Without a local entity you have no clean way to collect royalties, no way to run the two company-owned units the 2+1 rule expects, no employer for a quality control team, and a weak position in any dispute. Most brands that go this route incorporate within two years anyway, after the expensive mistakes.
What does the 2+1 rule really require of a foreign brand?
Two outlets owned and operated for at least one year before you grant a franchise. The units can be held by your subsidiary or an affiliate. Foreign franchisors have in some cases evidenced compliance using outlets operated outside China, supported by a statement from a recognised trade association. Do not build your strategy on that exception. Chinese franchisees and Chinese judges both take the rule at face value, and units you actually run in China are what makes your offer credible commercially, not just legally.
How much does it cost to launch a franchise network in China?
Budget in three blocks. Legal and IP first: trademark filings across relevant classes, CNIPA licence recordal, contract drafting, MOFCOM filing, typically a five-figure euro amount. Then the company-owned units, which for a food concept in a tier-1 city means several hundred thousand euros for two stores plus a year of operating losses. Then brand building in Chinese, because franchise candidates check whether the brand exists on Xiaohongshu and Dianping before they call you. The legal block is the smallest and the one people over-worry about.
How do I stop a franchisee from copying my concept?
You cannot stop it completely, so make it unattractive. Register your marks in China before negotiations start. Keep one input they cannot source without you. Structure your manuals as documented trade secrets with restricted, traceable access. Write a non-compete with a defined radius, a defined duration and actual compensation, because uncompensated non-competes are hard to enforce here. And keep the customer database at brand level. A copycat who has to rebuild the supply chain and the customer base from zero usually decides it is not worth it.
Is F&B still the right category to franchise in China?
It depends entirely on the sub-category. NCBD measured 22.66% closures across Chinese food service in 2024-2025, but the range runs from 8.9% for Anhui cuisine and 10.3% for coffee up to 37.2% for crayfish. Daily-habit formats with repeat purchase survive. Trend formats do not. If your concept depends on being novel, you have a 24-month window and a franchise contract lasting five years. That mismatch is where most foreign F&B franchise projects in China actually die.
Gentlemen Marketing Agency
We are a Shanghai-based digital marketing agency, and we work with foreign brands building retail and food networks in China.
We handle the part a franchise contract never covers: brand visibility in Chinese, Xiaohongshu and Dianping presence, WeChat member databases owned at brand level, and franchise candidate lead generation.
If you are about to sign a master franchise or you already regret one, talk to us first.
Read Also:
