Every Western brand selling in China runs on Tmall, JD.com, and Douyin at once — and almost every one of them ranks those channels by GMV. That is the wrong ranking. The channel with the highest gross merchandise value is frequently the least profitable once you subtract platform commissions, merchant-funded promotions, return losses, advertising, logistics, and FX. Channel profitability is what is left after all of that: the true net margin each marketplace contributes to your P&L. Until you can calculate it per channel, you are allocating inventory, ad budget, and headcount to whichever platform shouts loudest — not the one that actually earns.
This is the capstone of the China marketplace close. Once you can reconcile each cost line individually — platform fees, advertising spend, returns, and promotion funding — the next question is unavoidable: which channel is actually making money? This guide explains what channel profitability means on Chinese marketplaces, why GMV misleads, the exact cost stack you must subtract to reach true net margin, how Tmall, JD, and Douyin differ structurally, and how to build a per-channel P&L that a CFO can defend.
What is channel profitability on Chinese marketplaces?
Channel profitability is the net contribution margin a single marketplace generates after every channel-specific cost is deducted from the revenue it produces. It starts from the same GMV-to-net-revenue bridge you use for the whole business, but it is calculated one platform at a time so you can compare Tmall against JD against Douyin on a like-for-like basis. The output is a per-channel contribution margin — revenue minus the variable and directly attributable costs of selling on that channel — which is the number that should drive budget, inventory, and expansion decisions.
It is not the same as gross margin, and it is not GMV. Contribution margin isolates the profit a channel adds after its own variable costs, which is exactly what you need when the same SKU sells across three marketplaces with three different fee, return, and promotion profiles.
Why does GMV mislead brands about which China channel is most profitable?
Because GMV is a top-line vanity number that ignores every cost that varies by platform — and on Chinese marketplaces those costs vary enormously. Two channels can post identical GMV and deliver completely different profit. Douyin can generate huge livestream-driven GMV at a commission and return rate that leaves almost nothing behind, while a lower-GMV Tmall flagship store quietly delivers the bulk of your actual margin. Ranking by GMV inverts the priority order — a real risk given how concentrated and competitive China’s e-commerce market is across these three platforms.
- Commission spread — category take rates differ by platform and by product category, so the same sale nets different revenue depending on where it closes.
- Promotion intensity — Douyin and festival-heavy channels layer far more merchant-funded discounting, which is pure contra-revenue.
- Return behavior — impulse-driven livestream commerce carries materially higher return rates than search-led Tmall purchases, and returns destroy margin.
- Ad dependency — some channels only move volume with heavy paid traffic, so their GMV is effectively rented, not earned.
This is why a Tmall vs. JD vs. Douyin data comparison that stops at GMV is dangerous: it makes the highest-cost channel look like the winner.
What costs must you subtract to get true net margin per channel?
To move from a channel’s GMV to its true net margin, you subtract a stack of costs, each of which must be attributed to the specific platform that incurred it. Skipping or mis-allocating any one of them breaks the comparison. The full stack, in order:
- Merchant-funded promotions — store coupons and cross-store discounts you paid for, booked as contra-revenue (platform-funded subsidies are not subtracted; see the promotion split below).
- Platform commission & fixed fees — the category take rate plus technology and service fees deducted at settlement.
- Payment & transaction fees — Alipay/WeChat Pay and platform payment processing charges.
- Advertising & traffic costs — Alimama, JD Ads, and Douyin paid traffic directly attributable to that channel.
- Returns & refunds — the net revenue and logistics loss from returned units, which varies sharply by channel.
- Fulfillment & logistics — warehousing, last-mile, and reverse logistics for that platform’s orders.
- FX impact — the RMB-to-reporting-currency conversion on the actual settled cash, not the order-date rate.
The promotion line deserves special care: only merchant-funded discounts reduce your revenue as contra-revenue under IFRS 15, while platform-funded subsidies are reimbursed at settlement and must not be treated as your cost. Getting that split wrong is the single most common reason a channel P&L understates margin.
How do Tmall, JD, and Douyin differ in cost structure?
Each marketplace has a distinct economic shape, and understanding it tells you where the margin leaks before you even open the data. These are directional patterns you should confirm against your own settlement records, not fixed rates.
Tmall (天猫) — search-led, margin-stable
Tmall demand is largely search- and brand-led, so return rates are lower and revenue is more predictable. Commissions and the annual technology service fee are real, and flagship stores carry Alimama ad costs, but the channel typically delivers the steadiest contribution margin of the three.
JD.com (京东) — logistics-heavy, self-operated nuance
JD splits between marketplace (POP) and self-operated (1P) models, and the two have very different economics — 1P looks like wholesale, POP looks like Tmall. Its logistics strength reduces fulfillment risk but the fee and inventory terms on 1P can compress margin if you do not model them separately.
Douyin (抖音) — livestream volume, high leakage
Douyin is interest- and livestream-driven, generating spiky, high GMV — often at the cost of heavy merchant-funded discounting, elevated return rates, and creator/commission fees. It can be highly profitable or quietly loss-making depending entirely on how disciplined your promotion and ad spend are. It is the channel where GMV most overstates profit.
How do you build a channel P&L for China marketplaces?
You build one contribution-margin statement per channel, sourced from settlement data rather than order data, and you standardize it so the three are directly comparable. The workflow:
- Start from settled revenue, not booked GMV — pull each platform’s settlement statements and reconcile them the same way you would a bank statement. See settlement reconciliation for the mechanics.
- Tag every cost line with a channel dimension — fees, ads, returns, and logistics must all carry a Tmall / JD / Douyin tag so nothing lands in an unallocated bucket.
- Separate merchant-funded from platform-funded promotions — only your funded share reduces channel revenue.
- Convert on settled cash — apply the FX rate on the actual payout, so channel margin reflects real repatriated cash.
- Roll it into the monthly close — fold the channel P&L into your China month-end close so it is produced on a repeatable cadence, not as a one-off analysis.
The end state is a single view where Tmall, JD, and Douyin each show revenue, every cost line, and a contribution margin percentage — the foundation of unified China P&L reporting.
How do you allocate shared costs across channels?
Directly attributable costs — a channel’s own commission, its own ads, its own returns — are assigned to that channel with no judgment required. Shared costs need a consistent allocation basis so you do not distort the comparison:
- Warehousing & inventory holding — allocate by units shipped or storage footprint per channel, not by revenue.
- Team & agency retainers — allocate operational headcount and TP (天猫合作伙伴) agency fees by time or order volume where you can attribute them.
- Cross-channel brand advertising — brand campaigns that lift all channels can be held in a shared line rather than force-allocated, so channel contribution margin stays clean.
The rule of thumb: keep contribution margin at the level of costs you can defend as channel-specific. Push genuinely shared, above-the-line costs into a separate corporate layer so a channel is never penalized for a cost it did not drive.
What data and tooling do you need to measure channel profitability?
You need channel-tagged, settlement-grade data unified into one model — and that is exactly where most brands stall, because each marketplace exports its costs in a different structure, in RMB, on its own schedule. Doing this in spreadsheets means re-keying three settlement formats every month and hoping the tags survive. The scalable answer is a data layer that ingests each platform’s settlement, fee, ad, and return feeds, normalizes them to a common schema, and preserves the channel dimension end to end.
This is the problem Digate is built to solve for global brands running China ecommerce operations: it consolidates Tmall, JD, and Douyin data into a single, reconciled model so contribution margin per channel is a standing report, not a quarterly fire drill. When channel profitability updates automatically with each settlement cycle, budget and inventory decisions stop being guesswork. It is the same foundation behind clean marketplace-to-ERP data and a trustworthy unified P&L.
Channel profitability in China: key takeaways
- Rank channels by contribution margin, not GMV. The highest-GMV marketplace is often the lowest-margin one once fees, promotions, returns, and ads are subtracted.
- Attribute every cost to a channel. Fees, advertising, returns, logistics, and FX must each carry a Tmall / JD / Douyin tag or the comparison is meaningless.
- Split promotion funding correctly. Only merchant-funded discounts reduce channel revenue; platform-funded subsidies are reimbursed at settlement.
- Work from settled cash. Build the channel P&L from reconciled settlement data and settled-cash FX, not booked orders.
- Make it repeatable. A channel P&L is only useful if it is produced every close on a unified data model — not reconstructed by hand each quarter.