Buzz marketing moves faster in China than anywhere else I have worked. A clip can pull two million views in under two hours and be gone from every feed by dinner. What survives that speed is rarely the brand’s original plan. Sometimes it is a mistake somebody turned into an opportunity fast enough. Sometimes it is a stunt built for exactly this kind of spread.
Olivier Verot is the founder and CEO of Marketing to China, based in Shanghai since 2012. He has watched buzz spread across Weibo, WeChat, and now Douyin and Xiaohongshu, with brands that got the timing right and a few that got it badly wrong.

Why Chinese buzz still moves differently in 2026
Douyin’s monthly active users passed 1 billion in March 2026, up 14.4% year on year, according to figures reported by Sina Finance. That scale changes what going viral means. A clip does not need the whole country watching. It needs to hit the right interest cluster inside an algorithm that rewards watch time over follower counts.
Most foreign brands miss this part. Buzz in China is a chain, not one event: a spark on Xiaohongshu or Douyin, remixed by ordinary users, then picked up in private WeChat groups where the purchase decision actually happens. Miss one link and the buzz dies as a video nobody bought anything from.
Here are five campaigns that made real noise in China, and what still holds up about them a decade later.
Uniqlo’s fitting room video turned into a hotel chain’s advertising
A video of a couple in a Uniqlo fitting room in Beijing hit close to 2 million views in under two hours on Weibo and WeChat before it got pulled by censors. Nobody at Uniqlo filmed it, approved it, or wanted it. But foot traffic to that specific Beijing store spiked for weeks afterward, mostly people coming to look, not to buy.

The 7 Days hotel chain did the smart thing here: it did not try to own the story, it borrowed it. Its ads compared cabin comfort at Uniqlo to a night at 7 Days. Zero production budget, one joke, riding somebody else’s traffic spike. That instinct, react in hours rather than plan a campaign around a trend, still separates brands that catch buzz from brands that watch it pass.
OnePlus sold scarcity, not a phone
OnePlus launched a smartphone positioned as a cheaper Samsung alternative, then made it almost impossible to buy. An invitation system rationed access, with waits stretching past nine months for some buyers. The scarcity was the marketing. Every rejected request fed curiosity, and every successful buyer became a small advertisement for having gotten in.
The mechanic still works because it does two things at once: it slows demand into a manageable rollout, and it buys the brand time to fix issues on units already sold before the mass market ever sees them. Xiaomi ran versions of the same playbook for years. The lesson for smaller brands is not “create fake scarcity,” it is “release less than demand and say so honestly.” Chinese consumers spot manufactured hype fast, and it burns trust when discovered.
BMW’s crop circles played on an old fear
Mysterious circles appeared in the Gobi Desert in Qinghai, shaped like nothing natural. Weibo, WeChat, and national TV news ran the alien theory for days before BMW revealed the circles were built to promote the new BMW 1 Series, marketed on its “unique” design. The brand let the mystery run before attaching its name to it.
That sequencing matters. A reveal after the curiosity peaks earns more attention than a reveal at launch, because the second wave of coverage, “it was BMW all along,” is a story in itself and gets covered separately from the original mystery. Two news cycles from one stunt.
American Apparel’s 72-year-old face
American Apparel cast a 72-year-old man, Liu Xianping, the father of one of its designers, as the face of a campaign aimed at teenagers. The mismatch was the point. An old man styled like a trendsetter got shared precisely because it broke the expected casting for a youth fashion brand.

This one aged well as a strategy, even if the brand itself did not survive in China. Silver-haired creators built real followings on Xiaohongshu and Douyin through the late 2020s, and several skincare and fashion brands now run permanent 银发 (silver-hair) influencer programs instead of one-off stunts. American Apparel found the format a decade early.
KFC and McDonald’s fought a color war
KFC released a black “Black Diamond” burger and a pink “Rose Cheese” burger. McDonald’s answered with a grey, stone-colored bun stuffed with double pork and Sichuan sauce. Neither chain needed the food to taste dramatically different. It needed a photo that looked wrong enough to screenshot and send to a friend.

The packaging instinct behind it, bright, slightly absurd, built to be photographed, is still the default aesthetic on Xiaohongshu product posts a decade later. Chinese consumers reward food and packaging that reads as “kawaii” or visually strange over food that is simply well made but plain to look at.
The 2026 buzz playbook: what these five brands did not have
None of the five campaigns above had the tools brands use today to manufacture and extend buzz on purpose. Three matter most right now.
Tail-KOC networks instead of one big KOL. iiMedia Research put China’s KOL and KOC “grass-planting” market at 117.2 billion yuan for 2026, inside a wider word-of-mouth economy worth more than 500 billion yuan. The shift worth noting is where the trust sits: tail-end creators with a few thousand followers post interaction rates 3.2 times higher than big-name KOLs, and users trust them 78% of the time, ahead of professional bloggers at 65%. Fifteen small Xiaohongshu creators can cost close to what one mid-tier KOL charges for a single post, and read as opinion rather than advertising.
Open templates that invite remixing. During Qingming 2026, the bubble tea chain Mixue Ice City released the open source design file for its festival poster instead of a finished ad. Xiaohongshu users produced over 2,300 remixed versions in three days, at no extra media cost to the brand. The buzz was not a video people shared, it was a template people wanted to play with.
Getting cited by AI search, not just ranked by it. A growing share of Chinese consumers ask DeepSeek or Doubao what to buy before opening Baidu. GEO, generative engine optimization, means writing so the model can lift and cite you: Doubao favors content in clear H2/H3 sections with tables, DeepSeek rewards longer, detailed articles, and citation rates drop fast once a page goes stale. Buzz gets the spike. GEO keeps a brand showing up in the answer weeks later.

Jing Daily’s rundown of 2026 China consumer trends makes a point worth sitting with: shoppers increasingly reward brands that show cultural fluency and emotional understanding, not just a loud stunt. A weird poster still opens the door. What keeps people inside the store is whether the brand behind it seems to actually get them.
How Siobhan turned a mocked launch into a real sales recovery
Siobhan runs a small Irish skincare brand that entered China through a local distributor two years ago. Her distributor’s launch video mispronounced a key ingredient name on camera, and the clip got mocked on Xiaohongshu as cheap and out of touch. It picked up 40,000 shares in 48 days, mostly comments making fun of the accent, not the product. Her product page conversion rate dropped from 3.1% to 1.4% inside a week.
The first move was the obvious one: pull the video, post a formal apology in polished Mandarin. It did nothing. Conversion stayed flat around 1.6%, and a few comments accused the brand of being embarrassed by its own market. Deleting the joke did not kill it, it just took away the brand’s only voice in its own story.
What worked was leaning into it instead of fighting it. Her team hired fifteen tail-KOC skincare reviewers, accounts between 2,000 and 8,000 followers, and paid them to post honest “I tried the product everyone is laughing at” videos. Total cost across all fifteen notes came to roughly 13,000 yuan, a fraction of a single mid-tier KOL post. The reviews did not defend the mispronunciation, they ignored it and talked about texture and results instead, which is exactly what tail-KOC trust is built on: it reads as a peer’s honest opinion, not brand damage control. Within six weeks conversion recovered to 3.4%, slightly above the pre-crisis baseline, and direct sales for the quarter closed 18% above the prior quarter.
The mechanism, not the story, is the part worth copying. Bad buzz dies fastest when the brand fights to erase it. It fades naturally when the brand lets real people talk over it instead.
What can go wrong with buzz marketing in China
Buzz marketing has a failure mode most brands underestimate: a joke legal would not sign off on, a cultural reference that lands as mockery, a campaign read as disrespecting a sensitive topic. China’s platforms move fast enough that a misjudged campaign becomes a national story before a foreign head office wakes up. Trademark squatting on a suddenly-viral name is another real risk once a campaign moves; worth reading how to protect a brand in China before, not after, it takes off.
Paid amplification has its place once organic buzz starts fading. Boosting a clip through Douyin’s ad system at the right moment can stretch a three-day spike into three weeks of traffic, but only while the content still feels current. Push it stale and the algorithm treats it, and the brand, as trying too hard.
FAQ
How fast does buzz marketing actually move in China?
Fast enough that a video can peak and start declining within 48 hours. Douyin and Xiaohongshu’s algorithms reward early engagement heavily, so the first two to six hours decide most of a clip’s total reach. Brands that plan a “buzz campaign” over weeks usually miss the window; the ones that catch buzz react to something already moving.
Can a small or new brand trigger buzz without a big budget?
Yes, and 2026 makes it easier than a decade ago. Fifteen tail-KOC posts on Xiaohongshu can cost less than 15,000 yuan combined and generate more trusted engagement than one expensive KOL post. An open, remixable template or a genuinely funny in-house post costs even less. Format matters more than budget.
What is the biggest mistake foreign brands make with buzz campaigns in China?
Importing a campaign built for a Western audience and assuming the humor or shock value translates. It rarely does. The other common mistake is panicking and deleting content the moment a campaign gets mocked, which usually extends the story rather than ending it. A calm, honest response beats silence or an over-formal apology.
Does buzz marketing still work now that AI search platforms are part of the picture?
It works differently. Buzz still drives the short-term spike in awareness and traffic. But platforms like Doubao and DeepSeek increasingly shape what a consumer believes about a brand weeks later, based on what got written and indexed about the campaign. A brand that never writes anything substantial for AI models to cite loses most of the long-term value of the spike.
Buzz marketing in China rewards brands willing to move fast and look a little strange in public. It punishes brands that try to control every detail of a story that was never fully theirs to control in the first place. The five campaigns above prove the instinct still works. The tools for building it on purpose, tail-KOC networks, remixable content, and AI search visibility, are simply better than they were in 2016.
Marketing to China runs Chinese social buzz and influencer campaigns for foreign brands entering the market, from first Xiaohongshu seeding to full Douyin amplification. We build the KOC networks, we write for AI search visibility, and we tell clients honestly when a stunt is more risk than it is worth. If a launch needs a push that actually lands in China, get in touch with our team.