Ximalaya and Chinese podcasts: is audio a real channel for foreign brands?
Short version: yes, audio is a real channel in China — but not the one most foreign brands imagine. "China podcasts, 220 million listeners" makes people picture a US-style host-read ad market. What's actually there is two very different businesses sharing one word. Ximalaya is a mass audio utility — audiobooks, knowledge courses, radio, sleep tracks — with the reach. Xiaoyuzhou (小宇宙) is a narrow, urban-elite podcast salon with the attention. Pick the wrong one and you'll spend real money reaching people who will never buy you. Pick the right one and you get 40 minutes of a hard-to-reach audience's trust for a fraction of what Douyin charges for three seconds.
Audio is having a real moment in China. The podcast audience passed 220 million listeners in 2023, more than doubling over five years, and brand interest has followed — over 180 brands ran podcast advertising in 2024, up roughly 50% year on year, and around 170 branded shows were live. The names doing it aren't fringe: Lululemon, LEGO, L'Oréal Paris, Xiaomi and tea chain Bawang Chaji have all bought in, and heavyweights like Nike, Levi's and LVMH-owned houses have launched their own branded podcasts. So the question isn't whether audio works. It's whether it works for you, on which platform, and for what job.
The two Chinas of audio, in one sentence each
Ximalaya is the reach play. Founded in 2012, it's the largest online audio platform in the country — north of 300 million monthly active users at last public count, and now owned by Tencent Music after a roughly US$2.4bn acquisition. But most of that reach is audiobooks, paid knowledge courses, drama serials, crosstalk, kids' content and sleep audio. People use Ximalaya the way Westerners use a library plus a radio plus an Audible subscription. Penetration is broad and even — all ages, both genders, every tier of city.
Xiaoyuzhou is the attention play. It's small by comparison, but it is where actual conversational podcasts — the two-people-talking-for-an-hour kind — live and where the culture happens. Its user base grew about 50% in 2024, and the demographic is almost comically concentrated: overwhelmingly 16–35, over 80% female, more than nine in ten hold a bachelor's degree or higher, and roughly three-quarters sit in first-tier and "new first-tier" cities. That's the young, urban, educated, high-spend consumer that premium brands spend fortunes trying to reach on Red and can't get to answer a survey.
| Ximalaya | Xiaoyuzhou | |
|---|---|---|
| What it really is | Mass audio utility (audiobooks, courses, radio, sleep) | Conversational-podcast salon |
| Reach | Very large, broad across ages & tiers | Smaller, but growing fast |
| Audience | Everyone — commuters, families, students, seniors | Young, female-skewed, degree-holding, top-tier cities |
| Best brand job | Scale reach in a vertical; sponsor knowledge/parenting/wellness content | Deep consideration for premium, considered-purchase brands |
| Buy shape | Content sponsorship, channel placement, paid columns | Host-read integrations, branded shows, episode partnerships |
Why the economics are genuinely attractive — with a catch
Here's the number that makes marketers lean in: podcast CPMs in China run roughly 50–100 RMB, against something like 600–1,000 RMB on Bilibili and around 1,500 RMB for a WeChat placement. On paper, audio is the cheapest premium attention in the market. And the engagement backs it up — in one industry survey, 72% of Xiaoyuzhou listeners said they were happy for a show they follow to take a brand partner, and 82% reported taking some action after hearing a podcast ad. Audio listeners don't skip the way feed-scrollers do; a host reading your copy in their own voice is closer to a recommendation from a friend than to an ad.
The catch is scale, and it's a real one. Cheap CPMs multiplied by a small, concentrated audience is still a small absolute number of people. This is not a channel that moves volume. Chinese podcasting is also still figuring out how to make money at all — the medium struggles to convert its devoted listeners into paying subscribers, which is exactly why platforms lean on brand deals. If your KPI is units shifted this quarter, audio will disappoint you and you'll blame the channel for a job it was never built to do.
So when should a foreign brand actually bother?
The honest filter is simple: audio pays off when your product is a considered purchase with a story worth telling slowly, and it wastes money when your product sells on impulse and price. Use this:
- Good fit: premium beauty and skincare with a science or ingredient story, wellness and supplements, spirits, tech and gadgets, outdoor and performance, financial and B2B services, anything where the buyer researches before they commit and where trust is the bottleneck.
- Poor fit: fast-fashion, snack impulse buys, low-consideration commodities, or any launch whose whole logic is a discount code. Put that budget in Douyin and Red where the impulse loop closes.
- Right platform: reaching the urban-elite tastemaker → Xiaoyuzhou, host-read or a branded show. Reaching parents, commuters, learners at scale in a vertical → Ximalaya sponsorship.
- Wrong expectation: treating either as a performance channel with a same-week ROAS target. It's an upper- and mid-funnel trust builder that feeds your Red and search demand.
How to run it without lighting money on fire
The brands that get value here don't buy a spot and read a script. They do the unglamorous work first: they listen to the shows for a few weeks to learn which hosts genuinely fit the brand's world, they let the host write the read in their own language rather than handing over corporate copy, and they pick depth over breadth — one credible show whose audience is exactly yours beats ten scattershot placements. A host-read integration that sounds like the host actually uses the thing will out-convert a polished branded episode nobody chose to press play on. And because attribution in audio is even murkier than the rest of cross-border, measure it the way you'd measure PR: watch whether branded search and Red mentions lift in the days after an episode drops, not whether a tracking link fired.
One more discipline: don't over-produce. The reason podcasts work is intimacy. The moment your integration sounds like a TV spot, you've thrown away the one advantage the medium gave you.
Bottom line
Chinese audio is a real channel, but it's a scalpel, not a hammer. Ximalaya buys you reach across a broad audience inside specific verticals; Xiaoyuzhou buys you deep trust with a small, influential, otherwise expensive urban audience. Both reward brands with something worth 40 minutes of a stranger's attention and punish brands that just want cheap impressions. If you're a considered-purchase brand with a genuine story, audio is one of the best-value trust plays in China right now. If you sell on impulse and discount, skip it and spend where the impulse loop closes.
If you're weighing where audio fits in a China channel mix — or whether it fits at all — that's the kind of call I help brands make; reach out. For the platform that audio usually feeds, see how to get found in Xiaohongshu (Red) search, and for another patient, trust-first channel, read Zhihu marketing: earning Chinese trust through long-form Q&A.