Douyin Live Commerce: Why It Is a Retail Job, Not a Marketing Job
The dominant format on Douyin is not the influencer livestream Western teams picture. It is a staffed, scheduled, paid-traffic retail shift that runs most of the day, most days of the year. Brands fail on Douyin because they file it under the wrong department.
Annie Chan··16 min read
Almost every Western brand that approaches Douyin live commerce approaches it through the same mental model: TikTok Live, but bigger. Find a host with an audience, book a slot, ship product, count sales. That model describes a real thing on Douyin. It is just no longer the main thing, and it has not been the main thing for several years.
The format that carries most of the volume is the one Western coverage barely mentions, because it is not interesting to write about. It is a brand running its own livestream room, staffed by its own people, on a shift rota, most hours of most days, with paid traffic pointed at it continuously. In Chinese it is called dianbo, roughly "store broadcast," as distinct from daren bo, the influencer broadcast.
The load-bearing judgement in this piece: dianbo is not a content format. It is a retail operation with a camera in it. Brands that assign Douyin to their social or influencer team are assigning a store to a department that has never run a store, and the failure that follows looks like a creative failure but is an org-chart failure.
What the split actually looks like
Start with the shape of Douyin commerce overall, because the livestream is only one of three surfaces.
DaoInsights, reporting in March 2025 on data attributed to Douyin and to CBNData, described the platform's GMV as splitting roughly 40 percent shelf e-commerce, 30 percent merchant-led livestream, and 30 percent influencer-led livestream. The same report noted that merchant-led broadcasters made up around 70 percent of all e-commerce livestreamers on the platform, and that the number of brands starting merchant-led streaming grew 113 percent year on year in 2024.
The United States Department of Agriculture's Foreign Agricultural Service reached the same conclusion from a different direction. Its GAIN report CH2025-0229, "Selling in China: The Rise of Livestreaming E-commerce," published 4 December 2025, states that in 2024 store-operated livestreaming drove over half of overall livestreaming GMV, outpacing KOL-led sessions. The same report puts China's livestreaming e-commerce market at approximately 807 billion US dollars in 2024, and Douyin's total e-commerce GMV at 487 billion US dollars, with livestreaming accounting for somewhere between 40 and 58 percent of that total.
The number people quote, and why I will not repeat it as fact
You will see a widely circulated figure that nearly 70 percent of Douyin's livestream GMV came from store livestreams between February 2024 and January 2025, sourced to CBNData. You will also see the 4:3:3 framing above, which implies the store and influencer shares of livestream GMV are roughly even. Both cannot be right.
I do not know which is closer to true, and neither does anyone writing in English about this. The definitions differ, the measurement windows differ, and third-party trackers on Douyin estimate rather than observe. What is not in dispute, because three independent sources agree on the direction, is that merchant-operated broadcast overtook influencer broadcast, that it did so around 2023 to 2024, and that the gap has widened since. Plan against the direction. Do not build a business case on the magnitude.
"A number that three sources disagree about is not a number. It is a direction with a decimal point stapled on for credibility."
Why dianbo behaves like a store and not like a campaign
Three structural properties separate a Douyin merchant livestream from anything in the Western social-commerce vocabulary. Each of them maps to a different internal capability, and none of them maps to social media marketing.
It runs on a rota, not a calendar
A Western brand plans livestreams the way it plans campaigns: a date, a host, a hero product, a promotional mechanic. A serious Douyin merchant room plans the way a retail store plans: opening hours, shift handovers, staffing cover, a script rotation so the same product pitch cycles every twenty to forty minutes for viewers who arrive at different moments.
The reason is mechanical. Douyin's recommendation system evaluates a live room continuously on live signals: how long arriving viewers stay, whether they interact, whether they buy. A room that is dark eighteen hours a day starts every session from a cold position and spends the first stretch of every broadcast rebuying an audience it already paid for yesterday. Continuity is not a stylistic preference. It is what keeps the acquisition cost from resetting.
Marketing departments do not have the headcount model for this. A social team of four cannot cover a rota. A retail operations team that already staffs stores, or a customer service organisation that already runs shifts, can. That is the single most useful thing to know before you write the brief.
Its traffic is bought, and it is bought as media
The second thing that surprises Western teams is that the audience in a mature merchant live room is largely paid. Douyin's live commerce ad products exist precisely to push in-market users into a live room and measure the conversion inside it. Organic reach exists, but a room scaled on organic reach alone is a room with a ceiling.
So the correct budget line for a Douyin livestream programme is not content production. It is performance media, with a content cost attached. Teams that budget it as content set aside enough money to make the broadcast look good and nothing to make anyone watch it. The room then underperforms, and the post-mortem blames the host.
Its inventory is live, and its pricing is public
In a live room, price is visible, comparable, and screenshotted. Whatever you show is now your reference price across every channel you sell in, including your distributors, including your offline partners, and including grey-market resellers who will buy from your own livestream to arbitrage against you.
The USDA report is blunt about this: it advises exporters to develop a pricing strategy that preserves margin for MCNs, influencers, distributors and platform promotions, and warns that heavy discounts during major festivals can push consumer-facing prices below distributor costs, disrupting supply chains. That is a supply-chain governance problem, and it lands on whoever owns channel pricing. Usually that person has never been in the room where the livestream discount was agreed.
The influencer route looks familiar, and that is the trap
Daren bo, the influencer livestream, is the format Western brands recognise, and it is therefore the format they buy first. It is a legitimate channel. It is just not the channel most brands think they are buying.
Commercially it behaves like distribution, not like marketing. The typical structure is a slot fee paid up front for a place in the host's session, plus a commission on sales, plus a price that has to be at or below the lowest price you offer anywhere, because that exclusivity is what the host is selling to their audience. Add the agency or MCN margin on top.
Run that through your P&L honestly and you will usually find you have paid to place product with a distributor who takes a fee for shelf space, takes a cut of sales, sets your price, keeps the customer relationship, and gives you no data you can use next quarter. Those are distributor terms. There is nothing wrong with distributor terms. There is a great deal wrong with signing them believing you have bought brand marketing.
The test I would apply: after this influencer session ends, what do I own that I did not own before? If the answer is "a sales number and some clips," you bought distribution. If the answer includes followers on your own account, a shop with review volume, and search terms you now rank for, you bought an asset.
This is why the sequencing matters more than the split. Influencer broadcasts are a demand shock. A merchant room is where demand lands and keeps landing. Running the shock without the landing surface is how brands generate an impressive first month and nothing in month four.
Four things that break when you port TikTok Live directly
1. The entertainment-first assumption
TikTok Live in Western markets is a broadcast format with shopping bolted on. Douyin's merchant rooms are the inverse: a shop with presentation bolted on. The content register is closer to a home-shopping channel or a market stall than to a creator's livestream. Product, price, proof, urgency, repeat. Western teams find this aesthetically embarrassing and soften it, and the softened version converts worse than either extreme.
2. The missing shelf
Roughly 40 percent of Douyin's commerce GMV sits in shelf e-commerce, meaning search results, category pages and the brand's own store page. Livestream traffic does not all convert in the room. A meaningful share of it searches for you afterwards, and if your store page is thin, your review count is low and your listings are untranslated, that traffic evaporates into a competitor's page. A livestream programme without a maintained shelf presence is a bucket with the bottom missing.
3. Session structure calibrated to Western attention
A two-hour Western livestream with a scripted arc assumes an audience that arrives at the start. Douyin audiences arrive continuously and stay for minutes. The structure that works is loop-based rather than arc-based: a repeating cycle in which any entry point within a few minutes reaches a complete pitch. Brands that write a linear script find that 80 percent of arrivals land in the middle of an explanation they have no context for.
4. The account is not the store
Douyin separates the content account from the commerce entity. A brand can have a large, healthy content account and no ability to sell, because opening a store is a separate qualification process with separate documentary requirements. Teams regularly build audience for months before discovering the commerce side is blocked on paperwork nobody started.
GMV is not revenue, and in some categories it is not close
This is the part of Douyin live commerce that Western brands consistently fail to model, and it is where the money actually goes.
Yicai Global reported that returns from wholesale suppliers serving Douyin livestream vendors in women's apparel had risen to between 75 and 80 percent, up from around 60 percent the prior year, and described a supplier whose roughly 100 million yuan of gross sales netted out to around 20 million after returns. Those figures are specific to women's clothing sold through livestream, which is the highest-return corner of Chinese e-commerce, and they should not be generalised to other categories. But the mechanism generalises.
The mechanism is that Chinese platforms compete on consumer-side friction removal. Douyin, following Pinduoduo, Taobao and JD, made instant-refund policies mandatory for merchants, meaning a refund can be granted to the consumer before the goods come back, or in some cases without them coming back. Combine that with a purchase decision made in ninety seconds under time pressure and you get a return rate structurally higher than anything in your Western baseline.
Practical consequence: every business case for Douyin should be built on net GMV after returns, and the return assumption should come from category data on the platform, not from your existing DTC numbers. If your finance team signs off on a plan using your European return rate, the plan is wrong by a factor you will discover in month three.
One counterweight worth knowing: the USDA report notes that private-domain livestream rooms, meaning closed groups of existing customers rather than open algorithmic traffic, show refund and return rates as low as 3 percent, far below platform averages. The trade-off is obvious. Open traffic gives you scale with terrible retention economics. Private traffic gives you excellent economics with no scale. Most durable programmes end up running both, using open rooms for acquisition and private ones for margin.
The paperwork moved, and it moved against you
Two years ago, opening a cross-border store on Douyin was an administrative task. It is now a qualification exercise, and the direction of travel is one way.
Reporting on the 2025 and early-2026 rule updates describes a platform-wide crackdown on what Chinese platforms call "fake foreign brands," meaning entities registered offshore purely to present as imported. The consequences for genuine foreign brands are real: pure trading companies without brand ownership are broadly no longer accepted, trademark evidence or a notarised exclusive authorisation naming the platform and the Greater China territory is expected, and platforms increasingly want verification of overseas manufacturing and distribution rather than just an offshore certificate of incorporation. ChemLinked reported in 2026 that Douyin was consulting publicly on further tightening of international brand qualification requirements.
I would treat every specific requirement in that paragraph as directional and confirm current rules before committing, because they have changed repeatedly. The strategic point is stable: the qualification burden is rising, it is front-loaded, and it sits on a critical path that content production cannot compress. Start it before you start anything else.
There is a second-order effect worth naming. Rising qualification barriers are, on balance, good for brands that can clear them. A platform that has purged fake imported brands is a platform where being genuinely foreign is worth something again. The cost of entry went up and so did the value of being inside.
A sequence that survives contact
If I were scoping a first year, I would order it like this, and I would resist every attempt to reorder it for speed.
Qualification and entity work first, in parallel with nothing. It gates everything and it is the only step with an external clock you do not control.
Shelf before stage. Store page, listings, reviews, search terms. A livestream pushes traffic at a shelf, and an empty shelf converts at zero.
A small, ugly, consistent merchant room before any influencer spend. Two to four hours a day, six days a week, for at least eight weeks. This is where you learn your real conversion rate, your real return rate, and which of your SKUs actually demonstrates well on camera.
Paid traffic layered onto the room only once the room converts organically. Buying traffic into a room that cannot hold it is the fastest way to spend a China budget with nothing to show.
Influencer broadcasts last, sized against the room's capacity to absorb the aftermath, and negotiated knowing you are buying distribution.
Most teams want to run this in reverse, because the reverse order produces a headline number in week two. It also produces a programme with no owned asset, no learning, and a price floor you cannot recover from.
Who should own this internally
The recommendation follows from the whole argument. Douyin live commerce should report to whoever owns retail operations or direct-to-consumer trading, with marketing supplying creative and media buying as services into it.
The reason is capability, not politics. The job requires shift staffing, inventory allocation against unpredictable demand, real-time pricing discipline, returns processing at rates your logistics partner has not seen, and a customer service function operating in Mandarin during Chinese hours. Marketing organisations do not have any of those. Retail operations have all of them and are usually not consulted.
This connects to a broader pattern covered in the China market entry strategy piece: the failures that get diagnosed as strategy failures are usually ownership failures, where the China question sits across four departments and belongs to none of them.
Frequently asked questions
What is the difference between dianbo and daren bo on Douyin?
Dianbo is a merchant-operated livestream, broadcast by the brand's or seller's own staff from its own account, typically running long hours on a shift rota. Daren bo is an influencer-operated livestream, where a host with their own audience sells your product in their session, usually for a slot fee plus commission. Dianbo behaves commercially like running a store. Daren bo behaves commercially like placing product with a distributor.
Which one generates more sales on Douyin?
Merchant-operated livestreaming overtook influencer-led livestreaming and now accounts for the larger share of livestream GMV. The USDA Foreign Agricultural Service GAIN report CH2025-0229, dated December 2025, states that store-operated livestreaming drove over half of overall livestreaming GMV in 2024. Estimates of exactly how much more vary widely between sources, so treat the direction as reliable and the precise share as unsettled.
Can a foreign brand sell on Douyin without a Chinese company?
Yes, through the cross-border programme, but the qualification requirements have tightened substantially since 2025. Expect to need an overseas operating entity with trading history, trademark ownership or a notarised exclusive authorisation covering the platform and territory, and in many cases a mainland guarantor arrangement. Rules have changed repeatedly, so verify current requirements rather than relying on any guide, including this one.
Why do Western brands fail at Douyin livestreaming?
Most often because the programme is owned by a marketing team that cannot staff a rota, budgeted as content rather than performance media, planned as a campaign rather than a continuous operation, and modelled on gross GMV without a realistic return assumption. The creative problems are real but secondary. The structural problems are what end the programme.
How high are return rates on Douyin livestream sales?
It depends heavily on category, and the widely cited extreme figures come from women's apparel, which is the worst case. Yicai Global reported livestream-linked returns in women's clothing reaching 75 to 80 percent among wholesale suppliers, up from around 60 percent a year earlier. Other categories are far lower. The important discipline is to source a category-specific figure from platform data before building any business case, rather than importing your home-market return rate.
Do I need paid traffic to make a Douyin live room work?
At any meaningful scale, yes. Douyin's live commerce advertising products exist specifically to feed live rooms, and mature merchant rooms run on a substantial paid component. Organic distribution can validate whether your room converts, which is the right thing to test first, but it will not carry volume on its own.
The one-line version
If your Douyin plan can be executed by a social media team, it is not a Douyin plan. It is a TikTok plan with the wrong logo on it, and the market it is aimed at stopped rewarding that format several years ago.
"The brands that win on Douyin are not the ones with the best livestreams. They are the ones that treated the livestream as a shop floor and staffed it accordingly."
If you want the layer where this becomes visible before it shows up in your sales data, the Xiaohongshu marketing guide covers the discovery side, and China social media marketing maps how the platforms fit together for a brand entering from outside.
Share:
Free · 24 pages
Read Chapter 1 of the Playbook
AI Foundation Models — how DeepSeek and Qwen crossed a billion downloads while Western boards were counting Mag7 capex, and which layers of the AI stack the rotation has not reached. No excerpt tricks: it is the full chapter.
You will also get the weekly newsletter. Unsubscribe any time — the link is in every email.
The Book · $99
China Market Entry Playbook 2026
456 pages across 22 industry chapters — AI, semiconductors, EVs, payments, logistics, beauty, F&B, healthcare, pets, toys. Every chapter runs the same five questions: the anchor number, the rotation underneath it, which layers of the global stack it has reached, and what to do about it. English and Simplified Chinese editions included.
Ex-Transsion Global Digital Marketing Director · Effie Awards Greater China jury member · Guest lecturer, Peking University School of Economics · Writes Annie Chan Talk, your insider lens on China.
