Your sales are growing in China. So why aren’t you making money?

Imagine you’re running a European consumer brand.

After months of preparation, you’ve finally entered China. Your flagship store on Tmall is live, your products are appearing in Douyin videos, and your local partner sends weekly sales reports showing steady growth. The team is excited, and your China strategy seems to be working.

Then, three months later, someone from finance asks an uncomfortable question:

“We’re selling more products. So why aren’t we making more money?”

Selling products in China is one challenge, and generating sustainable profits from those sales is another. The gap between the two may have less to do with the product itself than with the commercial structure behind it.

Consider a product that sells for €99 in Europe and ¥1,100 in China. Of the ¥1,100 paid by a Chinese consumer, how much actually contributes to the brand’s bottom line?

I was involved in expanding a retail startup’s nationwide sales network in China. The business grew to 42 stores, with annual sales reaching approximately RMB 80 million. That experience taught me that sales growth and profitable growth don’t always move together. A store can report impressive revenue while rent, staffing, promotions and operating expenses steadily erode its contribution.

China’s e-commerce market presents a similar challenge. The difference is that the costs are spread across multiple commercial partners, which makes them much harder to see in a single sales report.

This week, let’s follow the ¥1,100.

1. Two Ways to Sell the Same Product

International brands entering China’s online market have several options. Two representative approaches are selling through a distributor and operating a brand-owned store with a third-party service provider.

Model A: Selling Through a Distributor

A Chinese distributor purchases products from the brand and resells them through local channels. The brand earns revenue from its sales to the distributor rather than from the retail price paid by consumers. Depending on the contract, the distributor may handle importing, inventory, warehousing, local sales and distribution, which reduces the brand’s operational complexity and the working capital it needs to commit to China.

The trade-off is that the brand gives up part of the potential retail margin and may have less control over consumer pricing, promotions, distribution channels and customer relationships. Distributors may also negotiate volume rebates, marketing support or territorial rights. And when different distributors pursue conflicting pricing strategies, products can move across channels or territories at unauthorised prices. In China, this long-standing distribution problem is known as cuanhuo (窜货), or channel leakage.

Model B: Operating a Brand-Owned Flagship Store

Alternatively, the brand may operate a flagship store on Tmall and appoint a third-party e-commerce service provider, commonly known as a TP, to manage daily operations. Under this arrangement, the brand typically keeps the inventory and gains more direct visibility into store performance. Platform charges, operating fees, advertising, fulfilment, returns and unsold stock usually remain largely its responsibility too. In this issue, Model B means a domestic Tmall store run by a China-based selling entity under general trade. Selling cross-border through Tmall Global is a different route, with its own commercial and tax arrangements.

A TP can operate an entire sales channel without purchasing a single unit of inventory. The company helping you sell your products, in other words, does not necessarily share the risk of failing to sell them.

Neither model is inherently superior. What matters is how revenue, costs, commercial control and risk are allocated between the parties.

2. Who Gets Paid Along the Way?

A single ¥1,100 online order may involve several commercial participants. The marketplace provides the transaction infrastructure, a TP manages store operations, creators or livestreamers help generate demand, and logistics providers handle delivery and returns. Behind them is the brand that develops, manufactures or sources the product. Each participant may be compensated differently.

Tmall applies category-dependent transaction-related charges. Alibaba has disclosed transaction-based rates typically ranging from 0.1% to 5%, and the relevant merchant agreement also specifies a basic software service fee of 0.6%, subject to applicable platform rules and system settings. Merchants may also be required to provide a refundable security deposit, which ties up working capital while the store operates.

TP compensation is negotiated contract by contract and may involve a fixed monthly service fee, a sales-based commission or both. Commonly quoted market ranges put the base fee at around ¥40,000 to ¥80,000 a month plus 5% to 20% of sales, while commission-only arrangements tend to run at 20% to 30%.

Creator marketing can similarly involve fixed campaign fees, sales commissions and other promotional expenses. Commonly quoted creator commissions fall between 10% and 30% of sales, depending on the category and the creator’s reach. Well-known livestreamers often add an upfront placement fee, known in Chinese as kengweifei (坑位费), ranging from a few thousand yuan for mid-sized accounts to millions for the largest live rooms. It is payable whether the product sells or not.

Actual charges vary by category, service scope, campaign and contract. For brands, the more useful distinction is between three types of financial commitments:

Fixed costs: Expenses that remain payable even when sales disappoint.

Variable costs: Expenses that change with sales volume or transaction value.

Upfront commitments: Money committed before the commercial outcome of a campaign is known.

These costs behave differently as a business grows, and the differences matter most when actual sales fall below forecast.

3. When a Successful Campaign Becomes an Expensive One

Imagine hiring a Douyin creator to promote your product. For this hypothetical campaign, assume:

• Product retail price: ¥1,100
• Fixed campaign fee: ¥30,000
• Creator commission: 20%
• Orders attributed to the campaign: 100

The fixed fee works out to ¥300 per attributed order and the creator’s commission adds ¥220, so the campaign costs approximately ¥520 per attributed order before product costs, fulfilment and other operating expenses.

Now suppose one-third of those orders are returned, leaving approximately 67 retained orders. If the full commission remains payable, the campaign cost per retained order rises to approximately ¥780. If commissions on returned orders are fully reversed, it is approximately ¥670. The actual outcome depends on the contract, refund arrangements and platform settlement rules. These figures also leave out the other costs that returns create, such as reverse logistics, damaged stock and non-refundable transaction charges.

There is also an attribution problem. Some orders credited to the campaign might have occurred without it, in which case the cost per genuinely incremental order is higher than the cost per attributed order. A campaign report showing sales growth does not, by itself, demonstrate profitability.

So when a commercial partner promises to increase sales, the more useful question is how much it costs to generate those additional sales, and whether the business can sustain that cost.

4. Following the ¥1,100: A Unit Economics Illustration

Let’s compare the two models using a simplified financial example. All figures below are hypothetical. They illustrate how different commercial arrangements affect a brand’s economics and do not represent industry averages, actual supplier quotations or the results of a particular business.

Model A: Distributor. The distributor purchases products from the brand at ¥440 per unit, with a 5% volume rebate. The brand’s assumed product cost is ¥250 per unit.

Model B: Brand-Owned Flagship Store. A China-based selling entity operates a Tmall store under the general VAT calculation method. The product is subject to 13% VAT and sells to consumers for ¥1,100 including VAT, which gives a VAT-exclusive sales value of approximately ¥973. The brand appoints a TP charging a fixed monthly fee of ¥50,000 plus a variable fee of 8% of the VAT-inclusive transaction value. At 1,000 units per month, the fixed TP fee represents ¥50 per unit. Other operating costs are based on the assumptions in the table.

The model is calculated per retained sale, after refunds. The ¥65 allowance covers fulfilment and return-related handling, and refunded orders are not counted as revenue. Marketing and other costs are spread across retained sales in the same way, and the monthly volumes in the next section are retained sales too.

Figures are rounded to the nearest yuan; the contribution calculation uses unrounded values.

Contribution here means the amount remaining after the costs included in this model. It is not the company’s final net profit, and the VAT-exclusive sales value is not the cash the business retains, since VAT is accounted for separately, including any input VAT deductions.

At 1,000 units per month, the two models produce similar illustrative contributions through very different commercial structures. In Model A, the brand receives a lower selling price but may transfer more downstream operating responsibilities to the distributor. In Model B, the brand recognises a higher retail sales value but carries the platform, operating, marketing and fulfilment costs. The difference becomes far more significant once sales volumes change.

5. What Happens When Your Forecast Is Wrong?

Suppose the TP’s fixed fee remains ¥50,000 per month. Whether the store sells 300 units or 3,000, that contractual commitment does not change. Keeping all other unit-level assumptions constant, Model B would produce the following results:

At 3,000 units, the fixed fee is spread across a much larger number of transactions. At 300 units, it absorbs a significantly greater share of each sale, while the simplified distributor model still produces ¥168 per unit.

Real contracts are more complicated. A distributor may require minimum purchase quantities, marketing contributions, rebates or inventory-return rights, and a brand-owned store may see its advertising cost per order change as volumes grow. These assumptions would need to be tested against the economics of an actual business.

In this simplified model, more volume always helps: both the contribution per unit and the monthly total rise as sales grow. Real businesses rarely work that neatly. Additional sales in China are often bought with more advertising and more creator campaigns, and the cost of acquiring each extra customer tends to rise as the easiest buyers are reached first. When that happens, sales can keep growing while total contribution stalls, which is exactly the situation the finance team was asking about. Higher contribution is also not the same as higher net profit, since overhead, further investment and financing costs sit outside the model.

The illustration still highlights a common planning risk: a channel model can look attractive at forecast sales volumes and become commercially unattractive when actual demand falls short. Before approving a China market-entry budget, I would want to see a downside scenario alongside the optimistic forecast.

6. Who Carries the Risk When Sales Slow Down?

When products sell more slowly than expected, someone must hold the unsold inventory and pay for storage. Someone must fund additional promotions, approve discounts and absorb the cost of returns, and someone needs the contractual ability to exit an arrangement that is no longer commercially viable.

Under a genuine distributor buyout, more of the inventory and resale risk may transfer to the distributor. Under a brand-owned store, the brand typically retains greater exposure to inventory and operating risks. In practice, though, the contract matters more than the label. A distributor may negotiate stock-return rights or marketing support. A TP may be paid on sales even when the brand’s contribution is weak. A creator may be paid on orders placed, settled transactions or retained sales.

These incentives do not necessarily align, and a commercial partner may be meeting its contractual objectives while the brand’s overall profitability deteriorates. That is why negotiations should look well beyond commission rates. Minimum service fees, inventory obligations, return settlements, discount authority, termination rights and access to commercial data often have a greater long-term impact.

Data requires particular attention. Operating a flagship store does not automatically give a brand unrestricted access to, or ownership of, customers’ personal information. Platform rules, contractual provisions and applicable Chinese data protection laws determine how such information can be accessed and used. Commercial control, financial exposure and partner incentives need to be evaluated together.

7. Five Questions Before You Sign

Before entering a distribution or e-commerce operating agreement, I would want clear answers to five questions.

1. What is our actual revenue per unit? Start with the revenue the business recognises under the proposed arrangement, rather than the retail price paid by consumers.

2. Which costs are fixed, variable or committed upfront? Monthly service fees, sales commissions and creator booking fees create different financial exposures.

3. Who owns the inventory and carries the downside risk? Pay close attention to slow-moving stock, discounting, returns and termination provisions.

4. What happens if actual sales reach only half the forecast? Evaluate the economics at different sales volumes, including a meaningful downside scenario.

5. Who controls the store, commercial data and customer relationships? Understand who operates the sales channel, who can access transaction information, and what happens to accounts, records and customer relationships when the partnership ends.

These questions cannot eliminate uncertainty, but they can help companies avoid confusing growing market activity with growing commercial value.

The China Question

In Issue #001, we explored how international brands can compete with lower-priced Chinese alternatives. In Issue #002, we examined who actually sells their products in China. This week, we followed the money: once a product is sold, who earns the revenue, who bears the costs, who captures the margin, and who carries the risk.

Much of China’s market complexity lies in how different business models divide the value of the same transaction among different parties. Finding a route to customers is the visible part of market entry. The harder part is building a commercial structure that turns sales growth into sustainable financial returns.

So here is the question I would leave with any international business evaluating the Chinese market:

If your sales doubled tomorrow, would your profits actually improve?

If the answer isn’t clear, it may be worth examining the economics of your distribution model before scaling further.

Sources and Methodology

The channel models and financial calculations in this article are hypothetical illustrations. They do not represent the actual results of any particular brand, industry averages or supplier quotations. Table figures are rounded to the nearest yuan.

In Model A, ¥440 represents the brand’s assumed selling price to the distributor on the stated revenue basis, with tax and delivery responsibilities governed by the model’s assumptions. Model B assumes a China-based selling entity using the general VAT calculation method and a product subject to 13% VAT. Operating costs are expressed on a comparable basis excluding recoverable VAT. The landed product cost includes assumed freight and non-recoverable import-related costs but excludes recoverable import VAT. The ¥50 platform-related charge is a hypothetical combined allowance, not an official Tmall fee quote; actual platform charges can also include points programmes, deposits and return-related items. Model B is calculated per retained sale after refunds.

Model A measures units sold by the brand to a distributor, while Model B measures retained retail sales to consumers. These are different transaction levels and should not be read as identical measures of end-customer demand.

The models assume different product costs reflecting different delivery responsibilities, so a real commercial comparison would require consistent contractual, tax and cost assumptions. The calculations exclude corporate overhead, financing costs, corporate income tax and other potentially material expenses, as well as one-off market-entry costs such as product compliance, labelling, registration or filing where required, trademark protection and systems integration. They also describe profitability rather than cash flow: payment terms, platform settlement cycles, deposits, inventory and advertising prepayments mean that a viable contribution margin does not necessarily mean a healthy cash position.

Platform fees: Tmall Merchant Service Agreement (Chinese); Alibaba Group Annual Report, Form 20-F, fiscal year 2026.

TP and creator compensation: Ranges in section 2 are commonly quoted market figures from public industry sources, not uniform benchmarks. Actual rates vary by category, scope of service and contract. The fee amounts used in the illustrations are hypothetical.

Tax framework: Value-Added Tax Law of the People’s Republic of China (Chinese).

The China Question is an independent business commentary series by Cyan Zhao, founder of BergBlick, exploring the commercial realities behind China’s markets.

The financial examples are illustrative and do not constitute tax, legal or investment advice. Actual results depend on product category, contractual arrangements, applicable taxes and operating performance.

Follow or subscribe for practical perspectives on China market entry, distribution and international business.