Before You Leap

The two look similar on paper and behave completely differently.

Export logic adds a country to the list, funds it from an existing budget, and judges it on this year's numbers. Growth-strategy logic means a named owner, a multi-year budget, and different expectations of when it pays back.

Brands that enter on export logic usually stall at the first point where China asks for more than planned — and withdraw concluding the market was wrong, when it was the resourcing.

This is the most common reason brands rule themselves out, and it is usually decided without a single number.

Size is not the qualifier. Margin structure, category fit and commitment are. Small brands with strong margins and a clear niche have entered through cross-border e-commerce without a local entity; large brands with thin margins have found the unit economics impossible.

The honest version of this question is not “are we big enough” but “does our product economics survive the cost of getting there.” That one is answerable with arithmetic.

“1.4 billion consumers” is not information. It cannot be planned against, budgeted for, or explained to a board.

China is not one market. It is many, separated by city tier, region, platform and consumer segment — and entry can begin with one province, one city, or one narrow consumer group.

The useful version of the question is smaller: who, specifically, would buy this product, at what price, and where do they already shop?

Partly — the operational discipline does. The playbook does not.

Discovery works differently: much less search, far more content and social recommendation. Traffic is bought rather than earned. The customer service and returns baseline is higher. Marketing tools you rely on are unavailable, with no direct equivalents.

Assuming transfer is what makes budgets wrong. The teams that adapt fastest are usually the ones who assumed least.

This is not the same question as which platform to join.

Discovery in China runs largely through content and social recommendation — short video, livestream, review-driven social platforms, group chat — before it reaches anything resembling a store. Many European brands have no mental model for this at all, which means it never appears in the plan or the budget.

You cannot plan for mechanics you have never seen. This is one of the cheapest gaps to close and one of the most expensive to leave open.

It means there is interest. It does not mean you have a market — and the difference matters commercially.

Grey-channel resale tells you demand exists, but you control none of it: not the price, not the story, not the customer relationship, not the quality of what arrives. It can also complicate your own launch later, because a price has already been set in the market without you.

This is worth checking before anything else, because it is free to check and it changes what you are deciding about.

Most brands are confident here, in one of two directions, and both guesses are expensive.

Nordic and European origin does carry meaning in some categories — clean, safe, natural — and almost none in others. Whether it supports a price premium is category-specific and testable, not a matter of opinion.

If the origin story is doing the heavy lifting in your plan, it is worth confirming that it lands before the plan depends on it.

Both are legitimate; they simply lead to different actions.

Geopolitical and regulatory concerns are real and belong in the analysis. But field conversations repeatedly surface something else: an unease that arrives before any commercial evaluation begins, and then quietly ends it.

It is worth separating the two. A decision not to enter, made on clear grounds, is a good decision. A decision made by mood is one you cannot examine, defend, or revisit.

A short visit updates a mental model faster than a year of reading, and it is one of the cheapest things on this list.

Walk a supermarket. Watch how people shop on their phones. Look at what your category costs on a shelf and online. Most brands report that a few days changed their picture more than any report.

Many brands never consider this option — not because it is expensive, but because it never occurred to them that it was available.

Most brands answer this with impressions rather than numbers.

Without them, every later conversation — with a distributor, a platform, your own board — happens with nothing to stand on. You cannot tell whether a partner's forecast is ambitious or fantasy, and you cannot tell whether a rejection was about your product or your price.

This is also where internal proposals usually fail. Boards rarely reject China; they reject a proposal with no numbers in it. Preparing that case is often what surfaces the real gaps, before any money is at stake.

It is the cheapest question here to answer properly, and the one most often skipped.

This is a different question from knowing the market. A brand can have solid category data and still never have modelled the cost of entering.

There is no published industry benchmark for what a China project costs, and anyone quoting one is quoting their own experience. What operators who have done it described, in a March 2023 EU SME Centre webinar: an agency running China projects for Italian brands treats roughly 1–2% of global annual revenue per year as the level below which a project cannot build enough awareness to compete. A Belgian operator running a multi-brand store on Tmall Global found first-year marketing consumed 25–35% of the revenue it generated, easing from the second year onwards, and described contribution breakeven arriving in year two.

On pricing, an EU SME Centre session on the Chinese market described a stack that surprises most brands — freight, tariff, VAT (value-added tax), importer margin, retailer margin and promotional budget — multiplying a consumer-goods factory price by roughly 30–34x by the time it reaches a Chinese shelf. Categories vary enormously. The value is in doing the arithmetic for your own product, not in the multiple.

Treat all of these as other people's experience rather than as rules. Until the figures have met your own profit and loss (P&L) model, the decision is still an impression.

Largely, no — and there are no direct substitutes.

The familiar stack is unavailable: the global social platforms, the usual analytics, the advertising accounts, the email lists. What replaces it is a different set of platforms with different rules, different measurement logic and different content expectations.

This is not a translation problem. The whole framework of what can we see, who can we reach, and how do we measure it has to be rebuilt — and that rebuild belongs in the plan, not in the surprise column.

There is no single door, and the choice is not mainly about cost.

Cross-border e-commerce, bonded warehouse, general trade, distributor, representative office, local entity — they differ in tax treatment, compliance burden, speed, control, and what data comes back to you. One common and costly misunderstanding: a representative office cannot sell.

You do not need to choose yet. You do need to know the options exist before someone chooses for you.

Many brands rule China out because the only version they can picture is the expensive one.

Cross-border e-commerce exists partly as a testing mechanism: it can validate whether a product finds buyers before committing to registration costs, local entities or full-scale distribution. Several documented European brands used it exactly this way — to find out which of their products had real demand.

A test you can afford to have fail is worth more than a plan you cannot afford to start.

This is a question to settle before signing, not after.

Some structures return sales, pricing and consumer data. Others return a quarterly figure and nothing else — no idea how many units, to whom, at what price, or why. The difference is decided by the route and the contract, at a moment when you still have leverage.

Worth knowing: the data exists. Chinese e-commerce is unusually measurable, down to category, brand, store and product level. Brands who end up blind are rarely blind because of the market.

The first barrier is usually not that compliance is complex. It is not knowing which rules are yours.

Requirements vary sharply by category — food, cosmetics, supplements, medical devices, children's products and electronics each sit under different regimes, and the same product can be classified differently in China than in Europe. Classification determines everything downstream: registration, labelling, permitted claims, timelines.

This is answerable, free, and best answered early. The EU SME Centre takes specific questions at no cost from companies that qualify — see the resources page for who does.

Some categories are not slow to enter. They are effectively closed.

A few regulatory frameworks are structured in ways that make foreign participation impractical regardless of product quality or budget. Brands have invested substantially before discovering this — which is avoidable, because it is knowable in advance.

It is an unglamorous question to ask first. It is also the only one that can end the project cheaply.

This is one of the genuine dilemmas, and it has no clean answer.

Cross-border logistics — particularly cold chain — make some products structurally unviable to ship. Local production solves that, and requires giving a Chinese partner your recipe, process or know-how. If the partnership ends, that knowledge stays with them.

Both paths carry a real cost. What causes damage is choosing one without having priced the other.

Almost everyone offering information about China has something to sell.

That does not make them wrong. It does mean their picture is shaped by what they provide, and that a brand with no independent reference point cannot tell an informed view from a well-presented one.

Two free, publicly funded sources exist and are consistently under-used. The EU SME Centre, funded by the European Commission, gives free tailored advice, reports and webinars on market entry, compliance and pricing. The China IP SME Helpdesk, also EU-funded, gives free first-line advice on trademarks and intellectual property in China, Hong Kong, Macao and Taiwan. Neither sells anything. Both are funded to serve small and medium-sized companies established in EU Member States and in countries participating in the Single Market Programme. That is wider than the European Union, but it does not cover every European country. Checked September 2026 — the Centre confirms eligibility directly, so ask before assuming either way.

If you know another genuinely neutral source, please tell me and I will add it here.

Interest and capacity are two different things, and it is worth separating them honestly.

A brand can be genuinely curious, have a good product, and still be in a year where survival, a factory move, a succession or a domestic downturn takes everything it has. Entering China while distracted is how projects get half-resourced — which is the most expensive version.

Being well prepared does not guarantee success, and being unprepared does not guarantee failure. But knowing more means that when problems arrive — and they will — there is less frustration, less anxiety, and more composure. That is what preparation actually buys.

A clear “not this year, and here is what would change it” is a stronger position than a reluctant yes.

The honest answer changes what this whole journey will cost you — in attention, not only in money.

A pattern that comes up repeatedly in Nordic internationalisation: neighbouring markets feel like an extension of home, but Germany is usually the first market that forces genuine organisational upgrades — a real export function, distributor management, multilingual operations, proper B2B contract structures. Many founders build their first real international infrastructure there.

China asks for that same organisational maturity, and then adds an entirely different ecosystem on top: different platform logic, consumer behaviour, regulation, digital infrastructure, communication norms.

A brand that has not yet built those muscles will meet China with fewer of them to draw on. This is not a reason not to go. It is a reason to know which challenge you are taking on — and to resource it accordingly.

In China the trademark belongs to whoever files first, not to whoever used it first.

The official fee itself is small: around RMB 270 per class for an electronic filing, roughly 40 euros, covering up to ten goods or services in that class. What a brand actually spends is higher once an agent, translation and notarisation are added — commonly several hundred to over a thousand euros per class. Buying your own name back from someone who filed it after meeting you at a trade fair costs many times that, and sometimes it is simply not for sale. Trade fairs — including European fairs with Chinese buyers present — are a documented high-risk moment.

Three details that catch brands out. China files by class, so one registration does not cover everything you sell and each class carries its own cost. Mainland China and Hong Kong are entirely separate systems — protection in one does not extend to the other. And registering the Latin-alphabet name is only half the job: if you have a Chinese name, or if consumers and distributors have already given you one, that version needs registering too, or someone else can hold it.

On timing: examination runs up to nine months, followed by a publication period in which oppositions can be filed — three months today, shortened to two from 1 January 2027 under the revised Trademark Law. With no opposition, that is roughly nine to twelve months from filing to certificate now, and nine to eleven from next year; an opposition extends it considerably. It is still the first clock to start.

Checked against the Trademark Law as revised on 26 June 2026 (Articles 32, 36 and 87) on 19 October 2026.

A brand without a Chinese name will be given one — by consumers, distributors or resellers — and you will not control it.

A phonetic transliteration is easy and often meaningless or unfortunate. A good Chinese name carries sound, meaning and character choice together, and it needs to be available to register in your classes.

This is worth doing deliberately and early, alongside the trademark filing rather than after it.

Products that are entirely legal in Europe can fail at the Chinese border on composition alone.

Ingredient limits, permitted additives and China's own national standards (GB, short for guobiao) differ, and a product may require reformulation, re-testing or a different specification for the Chinese market. For supplements and functional products the gap is widest.

This is checkable in advance against your specific formulation, and the cost of checking is a fraction of the cost of a rejected shipment.

For food and beverage this is one of the few things that stops a shipment at the border regardless of how good the product is. Overseas producers must be registered with Chinese customs, and the registration number has to be filed on the customs declaration when the goods arrive.

The framework changed on 1 June 2026, when Decree 280 replaced Decree 248. Registrations made under the old rules carry over and do not need reapplying, and most renew automatically — meat and bird's nest products are the exceptions. Overseas cold stores handling animal-origin and aquatic products are now in scope for the first time. Seventeen categories need an official recommendation from the competent authority in your own country before they can register; the rest can register directly. Which categories those are is managed dynamically, so check the current list rather than an article about it.

Cross-border e-commerce retail sits on a different track. Decree 280 itself says that producers of food imported through cross-border retail are handled under separate provisions, and the declaration rule in Announcement 27 applies to food imported as goods — which cross-border retail parcels are not. Neither text says in so many words that cross-border producers are exempt, and neither brings them within the requirement. The Finnish Food Authority’s China export guidance states plainly that Chinese authorities do not require CIFER registration for cross-border e-commerce trade, without naming the provision behind it. So the question to settle with your importer or cross-border service provider is which track your shipment is actually on. And if the plan is to move into ordinary import later, registration becomes a condition at that point — start it before you need it, not when you need it.

Registration itself is free of charge. Be wary of anyone charging for it.

Checked against Decree 280 (Article 30), Announcement 27 of 2026 and the Finnish Food Authority’s China export guidance on 19 October 2026.

Remaining shelf life on arrival stops more European food shipments than most brands expect, and there is no single national threshold written into law.

What exists is port-level practice: a minimum proportion of total shelf life that must remain when the goods arrive, commonly described as somewhere between a half and two-thirds, varying by product category and by port. At the stricter end, a 12-month product would need to arrive with more than 8 months left — after production, consolidation, shipping and customs.

Because it is applied port by port rather than by one national rule, confirm the figure that applies to your category with your importer and the specific port before producing for the market. Products with short shelf lives can find the arithmetic simply does not work through some routes.

Organic, gluten-free, fair trade, national quality marks — most European certifications carry no automatic recognition with Chinese regulators or consumers.

That does not make them worthless, but it does mean they become an education cost rather than a shortcut. What signals quality to a Chinese consumer is often something else entirely.

If your positioning depends on a certification, plan for explaining it rather than displaying it.

This is one of the sharpest binds for supplement, functional food and skincare brands.

Registering as a health food unlocks the claims that carry your differentiation — and triggers strict, slow and expensive certification. Not registering keeps entry simple and leaves you unable to say the one thing that makes you worth choosing over a local competitor.

There is no clean answer. But deciding it deliberately, early, is very different from discovering it after the packaging is printed.

A pattern that repeats: the trademark, the domain, the store account and the social media accounts end up registered on a distributor’s or an agency’s business licence rather than on yours.

It is rarely malicious at the start. Platform onboarding requires a Chinese entity, so a partner registers things in their own name because it is faster and it unblocks the launch. The problem appears at the end rather than the beginning: if the relationship stops working, the assets stay with them, and so does the audience you paid to build.

The paperwork that prevents this is unglamorous — registration in your own name wherever the platform allows it, a written authorisation chain where it does not, and a clause stating what transfers back and how. Ask where each asset sits before you sign, and ask again before you renew.

For food, cosmetics, supplements and other categories that need registration or filing, beginning compliance work 6–12 months before launch is the common recommendation, and several items run in parallel rather than in sequence. Categories with no special registration requirement can move faster.

The trademark clock is the longest. Food registration, formulation review, label approval and platform onboarding each have their own. Brands routinely discover this late and lose a season — not because anything went wrong, but because the timing was never mapped.

The practical move is to write the clocks down in one place, backwards from a target launch date, before committing to that date publicly.

Knowing the options and choosing between them are different moments. This is the choosing one.

Cross-border with direct shipping is simplest to start, but delivery commonly quoted at 7–15 days is well below what Chinese consumers accept, and it shows in ratings and repeat purchase. A bonded warehouse brings that to a service target of around 2–5 days, at the cost of holding inventory in advance and forecasting demand you cannot yet predict. These are operating norms rather than rules, and vary by platform, carrier and destination. General trade opens offline and full domestic sales, and carries the heaviest registration burden.

Switching later is possible and expensive. The choice also determines what data you will see once you are operating.

Compliance, logistics, platform operations and marketing are four different specialisms. Very few providers genuinely cover all four, and a brand that hires one expecting four will discover the gaps late.

Before looking for a partner, it is worth writing down which of the four you can do in-house, which you must buy, and which you intend to learn.

One question worth asking first: have you contacted the business support organisation in your own country? Most European countries have one, the service is usually free, and there may already be programmes, funding or introductions available that you are entitled to and have never used.

The documented failures here are rarely dishonesty. They are unverifiable reporting.

Marketing spend that cannot be traced to a product. Settlement data that will not reconcile with the operator's own reports. Invoices that describe nothing. Experienced operators — including ones with decades in the market — have reported taking a year and new hires to build visibility they should have had from day one.

The fix is upstream and cheap: write into the contract what gets reported, how often, in what format, against which KPIs, and what happens if the numbers cannot be reconciled. Operators repeatedly report that European contract templates proved hard to enforce in China, and that jurisdiction and arbitration clauses mattered more than they had expected. That is a question for a lawyer rather than for this map.

Cold chain is physics, not psychology, and it is priced accordingly.

For temperature-sensitive products, cross-border cold chain costs can make the unit economics impossible on their own — before any marketing spend. Disruptions, seasonal heat and gaps in the chain have caused real losses for European food brands operating there.

Get this costed early. It is one of the few constraints that can decide the whole question by itself.

Onboarding is heavier than most brands expect: business credentials, trademark evidence, deposits, annual fees, technical integration, and a review process that is neither fast nor transparent.

Requirements differ by platform and by category, and they change. Budgets that cover the product and the marketing but not the platform's own costs get squeezed at exactly the wrong moment.

Ask for the full fee schedule before committing to a launch date.

On Chinese e-commerce platforms, a brand flagship store, a distributor-operated store and a marketplace listing look similar to a consumer and are completely different businesses for you.

They differ in who owns the customer relationship, who controls pricing and content, who holds the data, and what happens if the relationship ends. Some brands have discovered they were building someone else's asset.

Decide which one you are opening, and confirm that the contract says the same thing.

Platform rules, category requirements and promotional mechanics change frequently and often without notice, and an operating approach that worked last quarter can quietly stop working.

This is a standing job, not a launch task. Brands that leave it entirely to an external operator tend to learn about changes after they have cost something.

On the major Chinese platforms, a good product is not usually found on its own. Discovery is generally bought, built through content, or both — and for an unknown foreign brand with no search volume behind its name, the organic share of that is close to nothing.

Paid promotion is therefore a baseline operating cost, not a growth option to add once sales start. A brand that budgets for the store but not for being seen has funded a shop with no street outside it — and the first months are exactly when visibility costs most and returns least.

This is the line item most often missing from first-year plans.

The sequence that worked for documented European brands: content seeding first, then genuine reviews, then paid traffic, then livestream.

Reversed — livestream first — is among the most expensive mistakes documented in cross-border launches, and one that experienced operators warn about repeatedly. Viewers who have never heard of you arrive in discovery mode, not buying mode, and the production quality cannot compensate. Livestream also needs a specialist team, real-time inventory and high-frequency content to work at all.

Content started at launch is content started late. The brands that convert began being visible before there was anything to buy.

Influencer and creator marketing is close to unavoidable in China and unusually hard to measure.

Rates for mid-tier creators are opaque, and on some platforms promotional links are restricted — meaning impact has to be inferred from search volume and traffic rather than tracked directly. Budgets can be consumed with only anecdotes to show for it.

Agree in advance what evidence of effect will look like. A first campaign sized as a lesson costs less to learn from than one sized as the budget.

The major festivals dominate the calendar. Not participating means very little traffic during the periods when consumers are actually buying; participating means discounting, often below contribution margin.

Both are normal. Pre-sale and warm-up typically open 3–6 weeks ahead of the peak date, though each platform announces its own calendar every year. The platform weight and review volume gained are usually the point, rather than the margin on those specific sales.

What causes damage is meeting a festival unprepared: no inventory plan, no content, no customer service capacity, no budget for the discount.

The rule is not that a claim must be true. It is that it must be evidenced, in a form a regulator will accept, before it is published.

European marketing language often works by positioning rather than proof — clinically inspired, the natural choice, trusted since 1920. In China the burden sits with the brand to produce the substantiation on request, and comparative or superlative wording carries the highest exposure.

Write the claims down before the campaign is built, and next to each one, write what you would show if you were asked.

A brand paid or gifted a creator to post. In Chinese law that post is an advertisement, and the brand is the advertiser.

This surprises European teams because the post reads like an ordinary user review — informal, personal, in the creator’s own words. The informality is the point of the format, and it is also where unreviewed claims enter. Content the brand never wrote is still content the brand answers for.

The same applies to livestreams, where a host improvises around an agreed script because improvising is the skill being paid for. A claim gets stronger, a comparison gets made, a benefit gets promised — and the brand may only learn of it when someone complains. Contracts with agencies and hosts can allocate this, but only if it was thought about before signing.

Two things worth settling before the first campaign rather than after: who on our side reads this content before it goes live, and what happens if something is said that we cannot support.

Approval usually happens in the language, and the file, that head office reads.

The English deck is signed off in Europe. The Chinese copy that actually runs is written locally, and it is the Chinese version that is regulated. Wording that was careful in English can become a definite claim in translation — often because the local team was trying to make it clearer.

The same gap applies to images. Festival imagery, zodiac illustrations and traditional motifs are frequently sourced or adapted locally, and much of that material has a copyright owner. Using it feels like cultural respect; the brand is nonetheless the one publishing it.

The version that needs reviewing is the one consumers will see — in the language they will read it in, with the visuals that will appear beside it.

Chinese-language customer service is a baseline expectation, not a premium feature, and response speed feeds directly into ratings and visibility.

In practice this is mostly written Chinese — platform chat, reviews, after-sales messages — rather than spoken, so it is a hiring and process question more than a language-school one.

Most European brands do not have this in-house. Outsourcing works, but the tone of voice becomes someone else's decision, and product knowledge has to be transferred and maintained. It is considerably easier to arrange before launch than once the first complaint has arrived.

In domestic Chinese e-commerce, free no-reason returns with doorstep collection have become a widely held expectation rather than a differentiator. Cross-border arrangements are often different, and consumers do not always know which one they are buying from — which is where the disappointment comes from.

Cross-border returns are genuinely expensive, and some brands choose not to accept them — which is a legitimate decision that directly reduces conversion and invites negative reviews. What causes avoidable damage is applying European returns logic by default and reading the resulting experience gap as bad luck.

Offering returns and declining them are both defensible; what differs is whether the cost was in the budget before it arrived.

Rating systems carry disproportionate weight for new and small brands, and the cause is often outside your control — a delayed delivery, a damaged carton, a misunderstanding about the product.

A few early negative reviews can suppress visibility before you have enough volume to absorb them. Which makes the unglamorous operational things — packaging, delivery speed, responsive service — a marketing investment rather than a cost.

A review base built early costs far less than one rebuilt after it has become load-bearing.

This is one of the moments where European entry conversations most often stall, and it is usually answered with a discount or a silence.

A brand that has mapped its own price stack — freight, tariff, VAT (value-added tax), importer margin, retailer margin, promotional budget — can respond with a specific and credible explanation of what the shelf price is made of, and where there is room. A brand without it simply loses the negotiation.

Do this arithmetic before the meeting, not after it.

Gifting is a large and distinct part of Chinese consumption, and it follows different rules.

In gifting contexts, recognisability often matters more than product quality, because the buyer is signalling to someone else. An unknown foreign brand, however good, can lose to a familiar one before the product is ever tried.

If gifting is part of your opportunity, brand awareness is not a marketing nicety — it is the product requirement.

Finding a distributor solves shelf access. It does not create consumer demand, and the two are easy to confuse.

In China, business-to-business (B2B) and business-to-consumer (B2C) work have to run at the same time: a distributor cannot sell what no consumer has heard of, and will lose interest in a brand that generates no pull. Brands that hand over both roles usually find that neither is being done.

Whatever the arrangement, brand building stays with the brand.

The answer was set by how you entered, not by how well you are doing.

Selling through your own store on a platform, you can generally see orders, traffic and consumer behaviour. Selling through a distributor or a third-party operator, what reaches you may be a periodic figure with no breakdown — no units, no buyers, no price realisation, no reason why. Brands have operated for years in that condition without ever being told it was avoidable.

If the numbers you receive today cannot answer “how much of what, to whom, at what price,” that is a contract and structure question, not an analytics one.

Money is going out and something is happening, but without attribution the connection stays a guess — and next year's budget gets set by whoever argues most confidently.

The underlying issue is usually structural rather than analytical: the data needed to attribute results sits with a partner or platform and is not being passed back in usable form.

It is worth asking for it explicitly, in writing, with a format specified. It is much harder to arrange after the fact than before.

Without a category benchmark, a rising number and a falling share look identical.

Brands have celebrated growth while quietly losing position, and panicked about a soft quarter that the whole category shared. Neither leads to a good decision.

One comparison — your movement against your category — changes the quality of every meeting about China.

Once several channels are selling, price positioning can drift away from you without anyone deciding it should.

Distributors clearing stock, resellers undercutting, platform promotions and grey-channel imports can produce a spread wide enough that consumers no longer know what the product is worth — and that reprices the brand, not just the transaction.

Price monitoring is unglamorous and is one of the few things that protects positioning after entry.

Official channels are rarely the only ones.

Grey-market and daigou channels operate independently of your plans — at prices you did not set, in packaging you did not approve, sometimes in condition you would not accept. Most brands discover this by accident.

You cannot manage a channel you cannot see, and the checking is straightforward once you know to do it.

Chinese consumer conversation happens on platforms most European brands never look at, in a language most of their teams do not read.

That means sentiment, complaints, misunderstandings about how the product is used, and early signs of counterfeits are all forming without you. By the time something reaches you through a partner, it is usually well established.

Listening is cheaper than repairing.

The answer depends almost entirely on what was registered before the copying started.

With a registered trademark, platform takedown mechanisms exist and work reasonably well. Without one, enforcement is slow, expensive and uncertain — and in the meantime the copies can travel: counterfeits of European brands have flowed back into European marketplaces, damaging the home market of brands that had no idea they had a China problem.

This is the point where the trademark decision made much earlier turns out to have been the whole answer.

A brand can operate in China for years and extract nothing transferable — no consumer insight, no product learning, no view of where the category is going.

That is usually a structural outcome: the learning happens at the partner, not at the brand. Which means when the partnership ends, the knowledge leaves with it and the next attempt starts from zero.

A practical question behind this one: is anyone on our side actually being developed to understand this market — the platform mechanics, the consumer, the operating rhythm? Not necessarily a large team. But if no named person is learning, the company is not learning either.

It is worth defining, deliberately, what you intend to learn from being there — separately from what you intend to sell.

Without agreed criteria, continuation decisions get made by fatigue.

The stronger version is decided in advance: a review point at 12 months, with specific measures — repeat purchase rate, contribution margin, cost of acquiring a customer, quality of consumer engagement — and a plan for what each outcome means. Documented operators are consistent on one point: scale on evidence of repeat purchase and sustainable margin, never on gross sales alone.

A well-founded exit is a legitimate result. A blind one is just an expensive way to stop.

This is worth testing on paper before it is needed.

Without your own visibility into performance, you cannot evaluate the current partner, cannot set targets for a new one, and cannot tell whether a problem is the market or the execution. Where the store, the accounts, the trademark or the customer data sit on someone else's licence, switching may not be practical at all.

The time to check what you own is before you need to move it.

China projects are unusually dependent on one motivated individual — often the person with a personal connection to Asia.

When that person leaves, projects stall or die, and the learning leaves with them. The next attempt starts from zero, pays the same tuition again, and often concludes that the market is difficult.

The questions worth answering early: who owns this by name, who else understands it, and where does what we learn actually get written down?

No study measures this, but it is the most consistent thing brand owners say afterwards: building recognition and a consumer base in China took several times longer than planned. Three to five times is the figure people reach for.

That gap is what kills otherwise sound projects: budgets sized for a shorter horizon run out during the awareness-building phase, precisely when spending stops and visibility disappears. Operators who have run these projects commonly plan on 6–12 months of test-and-learn inside a three to five year horizon — a habit rather than a rule, but a habit formed by people who paid for the lesson.

A slower plan that survives is worth more than a fast one that stops halfway.

European cost and return logic applied to a Chinese entry will almost always produce the conclusion that it is not worth it.

Customer acquisition costs more. Content requirements are heavier. Payback is slower. Judged against a familiar European market, the numbers look wrong — but they are being compared to the wrong thing, and the comparison, rather than the market, is making the decision.

Compare it to what it is: a different system with different economics, over a longer horizon.

Mental models of this market age faster than almost any other, and they age invisibly. Nothing announces that a picture has expired — you simply keep answering a market that has moved.

Before a project starts, two sources of distortion are common: impressions formed 5 or more years ago, and content seen on TikTok — which is not Douyin. They are different platforms with different audiences and different commerce behaviour. What you saw was a filtered version of somewhere else.

The less expected half is that this keeps happening afterwards. A picture that was accurate on the day it was formed goes quietly out of date while the project is running: rules are replaced, scope lists change, and port-level practice can move ahead of the written text. The section of this site on food producer registration carries the date it was checked for exactly that reason — the framework it was first written against has since been replaced.

So the question is not only when the picture was formed. It is how you would find out if it had expired.

These are three different problems and they are often treated as one.

Knowledge is knowing how the system works — the cheapest to close, mostly through reading and asking. Judgement is knowing what to do when something unexpected happens — built through doing, or borrowed from someone who has. Trust is being credible to partners on both sides — the slowest, and it cannot be bought.

Naming which one is missing tells you whether you need information, experience or time. Most brands buy information for all three.

Business cultures use agreement language differently.

In many Asian contexts a warm, positive response can mean “this conversation is going well and trust is still being built” — while in much of Northern Europe the same words would read as a commitment. Neither side is being unclear on purpose; they are using different conventions.

So rather than reading tone, which only produces anxiety in both directions, test with specifics: what is the next concrete step, who owns it, and by when. A relationship that is genuinely moving will produce a date.