Douyin passed 800 million daily active users in 2026, and people spend an average of 93 minutes a day inside the app. That number alone explains why every brand I talk to wants to advertise there. But most of them still think of Douyin as a place to run video ads. That view is five years out of date. Douyin is now an interest e-commerce engine. You do not just buy attention. You create demand with content, then capture it with paid traffic and a shop. This guide explains how the machine actually works in 2026, and who should spend money on it.
Written by Olivier Verot, founder of GMA. I have run Douyin campaigns from Shanghai since the platform opened its ad tools to brands, from first video seeding to full Juliang Qianchuan media buying and store livestreams. I have also watched plenty of budgets burn on it, which is why the second half of this article is about who should not bother.
What is Douyin and why it still matters
Douyin is the Chinese version of TikTok, run by ByteDance and launched in September 2016. It started as a short video app built around music, filters and 15-second clips. It is now the second largest retail channel in China. Douyin e-commerce GMV passed 4 trillion RMB, growing more than 30% year on year while the older platforms slowed down. That growth is the reason a video app belongs in a serious China budget.
The audience is broad. Douyin no longer belongs to teenagers. It reaches adults in their thirties and forties, users in lower-tier cities, retirees who shop by livestream. The old cliché that Douyin is only for post-90s youth is wrong now. If you sell almost anything to Chinese consumers, your buyers are on Douyin every day.
Read as well: Introduction to Douyin Marketing
Douyin is interest e-commerce, not an ad network
Here is the mental model that fixes most mistakes. On Baidu or Tmall, a shopper already knows what they want and searches for it. That is search e-commerce. On Douyin, nobody wakes up wanting your product. The algorithm shows a video, the video creates the desire, and the purchase happens in the same session. ByteDance calls this interest e-commerce. It means your job is not to interrupt someone who is looking for you. Your job is to be interesting enough that the algorithm keeps pushing your content to people who did not know your brand existed.
This changes what “advertising” means. A boring product video that you simply pay to broadcast will die. Native content that entertains first and sells second gets picked up by the recommendation feed and earns free reach on top of what you paid. The gap between those two outcomes is enormous. I have seen the same product get 20 times the views just by changing the creative from a TV-style spot to a native Douyin clip.
Juliang Qianchuan: the tool you actually buy media with
When brands say they want to “run Douyin ads”, the tool they mean is Juliang Qianchuan (巨量千川). It is ByteDance’s media buying platform for e-commerce, and in 2026 it sits directly inside the Douyin e-commerce department rather than off to the side. You use it to push traffic to three things: a short video, a livestream, or a product page. You set a goal, usually a return on ad spend target, and the algorithm optimises toward real transactions instead of raw views.
Two facts about Qianchuan matter before you commit a budget. First, it rewards content that already performs. If your organic video converts, paid traffic makes it fly. If your content is weak, Qianchuan just spends faster on something that does not work. Second, traffic is no longer cheap. Average Qianchuan CPM has climbed from roughly 80 RMB in 2023 to around 120 RMB in 2026, close to a 50% rise. The days of buying cheap Douyin reach are over. You need content good enough to earn organic distribution, then paid budget to scale the winners.
Store livestreams (店播) now beat influencer livestreams
For years the story of Douyin selling was the mega-influencer live show. That era is fading. Store-operated livestreams, called 店播, now drive around 70% of all livestream GMV on Douyin. A 店播 is your own brand running its own stream, every day, from its own account, often many hours a day. It is less spectacular than a celebrity host, but it is yours. You own the traffic, the data and the margin, and you are not paying a huge commission to a top KOL who can drop your product next week.
This is the shift most foreign brands miss. They still budget for one big influencer campaign and wonder why sales collapse the day after. The brands that win treat the livestream as a permanent shop channel, staffed and scheduled, fed by Qianchuan traffic. Influencer livestreams still have a role for launch spikes and credibility, but they are the spark, not the engine.
The shelf (货架) captures the demand your content created
Douyin used to be pure feed: content pushed at you, buy now or lose it. Not anymore. The platform built a shelf, called 货架 in Chinese, made of the in-app mall, the search bar and product cards. This is where people go back to buy the thing they saw yesterday. Shelf scenarios grew from about 30% of Douyin GMV in 2023 to around 40% in 2024, and reached close to 50% during big sales like 618.
Why this matters for your ad plan: content and livestreams create the demand, the shelf captures the part that does not convert on the spot. If you run videos but have no proper Douyin shop, search presence or product cards, you are paying to build desire and then letting it leak away. A serious Douyin operation optimises all of it together. Native content on top, Qianchuan traffic in the middle, a well-stocked shelf underneath. Miss one layer and the maths stops working.
Content still wins: the Mai Dong lesson
An old example still teaches the point better than any theory. Mai Dong (脉动), a sports vitamin drink aimed at 18 to 35 year olds, posted a plain promotion video on its official account. It got about 200 likes and a handful of comments. Dead. Then the brand launched a challenge instead: use our background music, add the hashtag, post your own clip, top likes wins an iPad. More than 74,000 people joined, and the top entry pulled in over 63,000 likes.
Same brand, same budget order of magnitude, completely different result. The interruptive ad failed. The native, participatory format worked because it fit how people use the app. In 2026 the tooling is far more advanced, but that principle has not changed. Content that belongs on Douyin earns reach. Content that feels like a TV spot gets buried.
What to consider before you promote on Douyin
- Native creative: film for Douyin, not for TV. Vertical, fast, human, ideally shot the way a normal user films. Polished brand films underperform against rougher native clips.
- Own your livestream: plan a 店播 schedule you can actually staff, not a single influencer event. Consistency feeds the algorithm and the shelf.
- Feed the shelf: set up your Douyin shop, product cards and search presence so demand you create has somewhere to land.
- Budget for real CPM: with Qianchuan CPM near 120 RMB, plan for paid traffic that scales proven content, not cheap blanket reach.
- KOC over one big KOL: many mid-size creators and everyday reviewers now beat a single celebrity for trust and cost. Seed widely, then amplify the winners.
Who Douyin is worth the budget for, and who it is not
I say no to Douyin more often than people expect. It is worth the money if you sell a visual, impulse-friendly, mid-price product: beauty, food and drink, small appliances, apparel, health supplements, anything that demos well in 30 seconds and ships easily. If your margin can absorb a 120 RMB CPM and a daily livestream, and you can commit to content for at least six months, Douyin can become your biggest China channel.
It is usually the wrong first move for long, considered B2B sales, ultra-niche industrial products, or brands with a tiny team and no stomach for daily content. If you cannot feed the machine content, you are just buying expensive impressions. For those brands I often point to other channels first, Baidu search, WeChat private domain, or content on Xiaohongshu, and treat Douyin as phase two. Read our take on why brands are moving budget from Tmall to Douyin before you decide the split.
Case study: how a Swedish brand made Douyin pay
Elin runs a Swedish home fragrance and candle brand. She came to us after nine months on Douyin with almost nothing to show: a beautiful account, cinematic brand films, and a cost per order so high the channel lost money on every sale. Her CPM had crept past 130 RMB and her videos averaged a few hundred views. She was ready to quit Douyin entirely.
The problem was not the budget, it was the model. She was buying traffic to interruptive brand films with no shop depth and no livestream. We rebuilt it in three layers. First, native creative: short, unpolished clips showing the candle burning, the scent described by a real person, filmed on a phone. Second, a daily 店播 store livestream, two hours every evening, run by two staff, not a celebrity. Third, we opened her Douyin shop properly with product cards and search so the demand had a shelf to land on. Qianchuan was set to a return on ad spend target and used only to scale clips that already earned organic views.
Over four months, organic reach on her top clips rose past 40,000 views each, and roughly a third of sales started coming from the shelf and search rather than the paid feed. Her cost per order fell by about 60% and the channel moved from loss to profit. Nothing magic, no overnight tenfold. Just the right model instead of paying to broadcast pretty films nobody asked for.
FAQ: advertising on Douyin in 2026
How much does it cost to advertise on Douyin?
There is no fixed price. You buy traffic through Juliang Qianchuan on a CPM or cost-per-action basis, and average CPM has risen to around 120 RMB in 2026, up from roughly 80 RMB in 2023. Beyond media spend you need budget for content production and, ideally, a daily livestream team. For a foreign brand testing the channel seriously, plan for a few months of committed spend rather than a one-off campaign, or the algorithm never has enough data to work with.
What is Juliang Qianchuan?
Juliang Qianchuan (巨量千川) is ByteDance’s media buying platform for Douyin e-commerce. You use it to push paid traffic to a video, a livestream or a product page, with the algorithm optimising toward transactions and return on ad spend. It rewards content that already converts, so it amplifies winners rather than rescuing weak creative. In 2026 it sits inside the Douyin e-commerce division, which tightened the link between ads and actual sales.
Should I use influencers or run my own livestream?
Both, but the balance has shifted. Store-operated livestreams (店播) now drive about 70% of livestream GMV on Douyin. Your own daily stream is the stable channel you control. Influencers still help for launches, credibility and a quick spike, but relying on one big name is fragile. Treat influencer streams as the spark and your own 店播 as the engine that runs every day.
What is interest e-commerce on Douyin?
Interest e-commerce means the platform creates demand instead of just serving existing demand. On a search platform, people look for what they already want. On Douyin, the algorithm shows content to people who did not know your product existed, the content sparks interest, and they buy in the same session or later from the shelf. Your advertising job is to make content interesting enough that the recommendation feed keeps distributing it.
What is the Douyin shelf (货架) and why does it matter?
The shelf is Douyin’s search-and-buy layer: the in-app mall, the search bar and product cards. It captures demand your content created but that did not convert on the spot. Shelf scenarios grew from about 30% of Douyin GMV in 2023 to near 40% in 2024, and close to half during festivals like 618. If you run ads without a proper shop and search presence, you pay to build desire and then let it leak away.
Is Douyin right for every brand?
No. Douyin pays off for visual, impulse-friendly, mid-price products that demo well in a short clip: beauty, food, drinks, small appliances, apparel, supplements. It is a poor first move for long B2B sales cycles, ultra-niche industrial goods, or teams that cannot commit to daily content. If you cannot feed the machine content, you are buying expensive impressions. For those cases, start with search or private domain and treat Douyin as phase two.
We run Douyin advertising for foreign brands in China
GMA is a Douyin and social commerce agency based in China. We build the full stack that makes the channel profitable: native content, Juliang Qianchuan media buying, daily store livestreams and a Douyin shop set up to capture shelf demand. We tell you honestly when Douyin is not your best first bet, and we manage the channels that are. If you want to sell on Douyin instead of just posting on it, contact us for an audit.
Sources and further reading: Dao Insights on store-led livestreams and 36Kr on 2026 content e-commerce. See also our guide to AI virtual hosts in livestreaming and Douyin statistics for 2026.
