How International Brands Can Test China Demand Before Setting Up a Local Entity
International brands do not need to make a mainland-China entity their first market test. A better sequence is to prove three things first: that a relevant audience exists, that a localized proposition can produce meaningful behaviour, and that the business can serve that demand through a workable route.
Only then should leadership ask whether a local entity would remove a real constraint—such as local contracting, hiring, invoicing, platform access, licensing, fulfillment or customer support.
This distinction matters. Setting up an entity can create capability, but it cannot create demand. A weak test also creates risk: if the message, price, channel or service path is not ready, a poor result may say more about the experiment than about China.
The plan below is commercial guidance for market-entry teams. It is not legal, tax, customs or regulatory advice. The right structure depends on the brand's sector, activities, products, contracts, data flows and locations, so specialist advice should be taken before a live commercial test.
What does “local entity” mean in this decision?
Teams often use “local entity” loosely. For this article, it means an enterprise incorporated or registered in mainland China and wholly or partly invested by a foreign investor. China's Foreign Investment Law defines a foreign-funded enterprise as one incorporated under Chinese law within China; it also makes clear that market access, negative-list restrictions and licensing can depend on the activity and sector.
Three other arrangements should not be confused with your own local entity:
An overseas parent selling or marketing into China. The company remains incorporated outside mainland China. Whether a particular activity is permitted or practical depends on the business model, product, channel and rules that apply.
A distributor, importer, agency or other local partner. This may give the brand a route to customers, but the partner is a separate company. Contracts should state who owns the accounts, customer data, trademarks, content, inventory and commercial relationship.
A resident representative office. This is not a shortcut to ordinary selling. MOFCOM's Foreign Investment Guide of the People's Republic of China (2025 Edition) says a representative office has no legal-person status, may not engage in profit-making activities, and is limited to activities such as market research, publicity and liaison within the permitted scope.
That vocabulary helps leadership ask the right question: not “Can we enter China without a company?” in the abstract, but “Which activities do we need to perform, who will perform them, and what structure will support them lawfully and reliably?”
First, define the decision your test must support
A China test should not begin with “let's see what happens.” Write down the decision, the evidence required and the stop conditions before spending begins.
A useful decision brief contains five lines:
Audience: the specific buyer, user or decision-maker you need to learn about.
Problem or occasion: the need, use case or trigger the brand hopes to serve.
Proposition: the localized benefit, product or service you want to test.
Behaviour: what a genuinely interested person should do next—search, save, request a sample, book a call, join a waitlist or place a compliant order.
Advance threshold: the level and quality of evidence that would justify the next test.
Set the threshold before results arrive. It might be a minimum number of qualified enquiries, a target conversion range, repeated purchase intent at an economically viable price, or evidence that two separate audience groups respond to the same problem. The number will vary by category; the discipline should not.
Lotus's view is that a credible entry case needs three kinds of proof:
Demand proof: people in the intended segment have the problem or desire.
Behaviour proof: some of them take a meaningful next step when shown a relevant proposition.
Delivery proof: the brand can serve that step with acceptable compliance, customer experience and economics.
Views alone satisfy none of these.
Stage 1: Map the constraints before you stimulate demand
Before publishing content, collecting leads or taking orders, create an activity map. List what the test will actually do: promote a product, collect personal information, ship samples, accept payment, contract with customers, use creators, store data overseas, or ask a partner to act on the brand's behalf.
Then check each activity against five questions:
Is the sector open to the proposed form of foreign participation?
Does the product, service or claim require a licence, registration, filing or prior review?
Can the intended platform and account type support the applicant and category?
Who is responsible for customer data, advertising claims, payments, fulfillment, returns and complaints?
Who owns the accounts, creative assets, Chinese brand name and customer relationship if a partner is used?
The current 2024 national foreign-investment negative list took effect on 1 November 2024, but a negative-list check is not the whole compliance analysis. Sector licences, product rules, platform eligibility, local market-access rules and the planned transaction flow can still matter.
Testing without your own entity also does not mean testing without obligations. China's Personal Information Protection Law can apply outside China when an organisation offers products or services to people in China or analyses their behaviour. The Internet Advertising Measures apply to commercial advertising through websites, webpages and apps within China. Build data and claim review into the pilot rather than treating them as a final launch task.
The output of this stage is not a long legal memo. It is a red-amber-green activity map that tells the team which experiments can proceed, which need a qualified partner or specialist review, and which should wait.
Stage 2: Build a baseline from evidence that already exists
Before creating new demand, find out what is already happening.
Start with the brand's English name, likely Chinese names, product names, category terms, common misspellings and the problems the offer solves. Review:
Chinese search trends and related questions;
social posts, reviews, creator content and comment language;
marketplace listings, reseller activity and authenticity questions;
competitors' positioning, price architecture and customer objections;
distributor, sales and customer-support enquiries already reaching the global team;
official category, trade and consumer data with a clearly stated period and denominator.
Baidu Index can show weighted search-interest trends, related demand and audience patterns for keywords. Baidu's own terms say the reports are reference material and that users should verify accuracy and completeness before making decisions. Treat it as a directional signal, not a measure of total market size.
Look for patterns that survive across sources. One viral post, one enthusiastic distributor or a cluster of copied marketplace listings is not enough. Stronger baseline evidence includes recurring problem-led searches, repeated comparison or price questions, independent creator discussion, identifiable buyer groups and enquiries that match the brand's actual offer.
This stage should end with a small number of testable hypotheses, not a large presentation. For example: “Urban professionals appear interested in the product's space-saving benefit, but trust depends on proof of materials and a clear warranty route.” That is specific enough to test.
Stage 3: Run qualitative learning before a scaled campaign
Use interviews, moderated sessions, customer-support transcripts, creator conversations and partner feedback to understand the language behind the signal.
The purpose is not to ask whether people “like” the brand. Ask participants to make choices:
Which problem feels most urgent?
How do they describe the category in Chinese?
Which proof makes the offer credible?
What would stop them from enquiring or buying?
Which price, service or delivery questions appear first?
Where would they expect to research the brand and verify that it is official?
Separate end customers from intermediaries. A distributor may value margin and supply reliability; a consumer may care about authenticity, product fit, delivery and after-sales support. Combining their answers into one “China demand” score hides the decision.
Continue until the team can explain the target customer's problem and decision criteria in natural Chinese, and until the remaining disagreements are clear enough to test quantitatively.
Stage 4: Test one localized proposition and one next step
The first active pilot should be deliberately narrow. Choose one audience, one proposition, one primary discovery environment and one next step that the business can genuinely support.
Examples include:
a focused Chinese landing page with an enquiry or waitlist route;
a compact content series testing two or three benefits;
a creator-led explanation with pre-agreed usage rights and disclosure;
a partner event or webinar with qualified follow-up;
a small sample or consultation programme;
a controlled commerce test through an eligible route.
Keep enough variables stable to learn. If the audience, offer, price, creative format and channel all change at once, the result will not tell you what worked.
Measure behaviour in layers:
Attention: qualified reach, search lift and relevant viewing show whether the intended audience notices the proposition.
Consideration: saves, repeat visits, comparison questions and detailed comments show whether people are evaluating it.
Intent: completed enquiries, sample requests, waitlist sign-ups and sales conversations show whether interest becomes a concrete next step.
Commercial: eligible orders, price acceptance, lead quality and partner conversion show whether the route can produce value.
Delivery: support load, fulfillment time, returns, complaints and margin after route costs show whether the promise can be served sustainably.
The most important denominator is the qualified audience, not total impressions. A lower-volume test with clear customer fit is more useful than broad reach that generates no interpretable behaviour.
Our related guide to China social media strategy for lifestyle brands goes deeper into a platform-led first 90 days. The principle here is broader: use the smallest credible experiment that can change a leadership decision.
Stage 5: Add a commercial test through a route that fits the category
Once the proposition produces credible intent, test the transaction or handoff that would make the opportunity real.
For a B2B company, that may be a paid diagnostic, sample request, qualified meeting or partner-led sales process. For a consumer brand, it may be a controlled cross-border retail test, an authorised retailer, a distributor pilot or another eligible path.
Cross-border e-commerce is a useful example of why “no local entity” should never be treated as “no local infrastructure.” China's official cross-border e-commerce retail-import framework recognises an overseas-registered enterprise as the cross-border seller, while also setting conditions around the goods, transaction data, domestic platform and service providers, consumer protection and regulatory responsibility. The route is category- and model-specific, not a blanket permission for any product.
If commerce is relevant, read our focused guide to entering China through cross-border e-commerce. For the present decision, the lesson is simple: test the complete customer path, including payment, delivery, service, returns and unit economics—not only the advertisement.
Agree in advance who will issue contracts or invoices, receive payment, hold inventory, support customers, process data, approve claims and carry product responsibility. If the test depends on a partner, record what is portable: data access, account control, creative rights, customer history and the ability to change partners later.
Stage 6: Test repeatability, not just a successful launch moment
One campaign can prove that a campaign worked. It does not prove that the market will support a durable operation.
Repeat the strongest test across at least one meaningful dimension: another time period, city cluster, audience segment, acquisition source, creator cohort or product. Keep the core proposition stable enough to compare.
Ask four questions:
Does the same customer problem keep appearing?
Can the brand generate qualified behaviour again without relying on one creator, partner or unusual event?
Do the economics remain credible after media, partner, fulfillment, service, returns and localisation costs?
Can the global team operate the process with clear ownership and acceptable speed?
A repeatable but operationally awkward test is still useful. It may reveal the exact constraint an entity could solve.
When does a local entity become commercially justified?
The entity decision should follow the bottleneck. It becomes worth a formal legal and tax assessment when evidence is repeating and one or more of these needs is material:
customers or partners require local contracting or invoicing;
the business needs to hire and manage a continuing local team;
platform, store, licence or account capabilities require a qualifying local applicant;
inventory, fulfillment, service, returns or warranty need tighter local control;
the company needs direct ownership of accounts, data, customer relationships and operating assets;
partner margin, speed or governance is now limiting an otherwise proven opportunity;
the economics support the ongoing fixed cost and management attention of a local operation.
These are triggers for professional analysis, not automatic instructions to incorporate. The final choice should compare an owned entity with other viable structures, including a distributor, importer, employer-of-record arrangement where appropriate, service partner, cross-border route or a decision not to scale yet.
Our China marketing roadmap for international brands explains how positioning, discovery, trust and conversion should connect once the business commits to a broader operating plan.
A practical go, refine or stop decision
At the end of the test, do not reduce the evidence to one score. Use three decisions:
Go: demand, behaviour and delivery evidence repeat; the remaining constraint is identifiable; the economics justify the next commitment.
Refine: the problem appears real, but the proposition, price, proof, channel or customer path is weak. Change one major variable and test again.
Stop or hold: demand is anecdotal, intent does not survive a credible test, the delivery route is not viable, or the opportunity is less attractive than another market.
“Not yet” is a valid outcome. A market test has done its job if it prevents an expensive assumption as well as when it reveals an opportunity.
Frequently asked questions
Can a foreign brand advertise in China without a mainland company?
Sometimes, but there is no universal answer. Eligibility depends on the platform, applicant location, account type, sector, product, claims and advertising route. Confirm current platform rules and obtain specialist advice for the proposed activity before spending.
Is cross-border e-commerce enough to prove China demand?
It can prove purchase behaviour through a specific assortment, price, platform and fulfillment model. It does not by itself prove demand across all of China or show that general trade, offline retail or another channel will perform the same way.
Can a representative office sell products or services?
MOFCOM's 2025 guide says a representative office may not engage in profit-making activities. It can support limited research, publicity and liaison activities within the permitted scope, but it should not be treated as a light sales company. Obtain legal and tax advice on the planned functions.
What if the first test performs poorly?
Check whether the test reached the intended audience, used natural Chinese, made a credible promise, offered a workable next step and was supported by price, service and delivery. A poor design can create a false negative. If those elements were sound and meaningful behaviour still did not appear, holding or stopping may be the right decision.
When should we start the entity assessment?
Start it when repeatable demand is visible and the team can name the operating constraint the entity would solve. Legal, tax and regulatory advisers can then assess the proposed activities, location and structure against real commercial requirements instead of a hypothetical launch plan.
Make the next China commitment with evidence
The best early China plan is not the one that creates the most activity. It is the one that reduces the most important uncertainty at a proportionate cost.
Lotus helps international teams connect China market research and consumer insight with a practical market-entry strategy, localized testing and ongoing delivery. If your team is deciding whether China demand is strong enough to justify a larger operating commitment, start with the decision, the evidence and the smallest test that can answer it.
