Zhihu marketing: earning Chinese trust through long-form Q&A
Short version: Zhihu marketing is not a reach play, and if you're measuring it on followers you'll quit too early. Zhihu — China's long-form Q&A platform, roughly 80 million monthly users — is where a buyer goes after they've seen you on Red or Douyin, to be talked out of you before they spend. Win that room and you own the last honest step before purchase. Ignore it and someone else — a rival, a skeptic, or increasingly an AI answer engine — writes your reputation for you.
Every foreign brand I talk to has a Red plan and a Douyin plan. Almost none of them have a Zhihu plan, and when I ask why, the answer is always some version of "it's small and slow." Both true. Zhihu's monthly active users have drifted down toward the low-80-millions — a fraction of Douyin's crowd — and the platform has aged up and skews majority-female now. It is not where you go for a viral week. It is where you go to still be standing there in a year.
What Zhihu actually is (and isn't)
Zhihu launched in 2011 as a Chinese answer to Quora — a question-and-answer site where people post a problem and others write long, argued replies. It grew into something more useful than that: a search-driven library of considered opinion that Chinese buyers treat as a bias-check. The register is different from every other platform. Red is aspirational, Douyin is impulsive, Weibo is loud. Zhihu is the person in the group chat who says "okay, but has anyone actually used this for two years?"
That's why it isn't an awareness channel and shouldn't be judged as one. Nobody discovers your brand on Zhihu the way they stumble onto it in a Red feed. They arrive with your name already in their head and a question typed out: is this brand any good, or am I paying for the logo? Your job on Zhihu is to make sure the best answer they find is honest, specific, and — ideally — yours.
Who it's actually for
Zhihu earns its keep for high-consideration purchases, the kind where a buyer reads before they spend. It's close to useless for low-stakes impulse categories. A rough sorting:
| Strong fit | Weak fit |
|---|---|
| Tech, electronics, appliances, tools | Fast fashion, trend beauty, snack impulse buys |
| Financial, insurance, education, healthcare | Anything bought on a whim from a short video |
| Premium and "invest once" durables (skincare devices, prams, mattresses) | Cheap commodities competing purely on price |
| B2B and professional tools where the buyer researches | Categories with no real "which one should I buy?" question |
The test is simple: type your category into a search bar and imagine the question a wary buyer would ask. If that question is "which X is worth it and why," Zhihu is where the answer gets written. If the question is "ooh, what's that," it isn't.
Why it punches above its 80 million users
Two multipliers make Zhihu worth more than its user count suggests, and both are about what happens to the content after you post it.
The first is search. Baidu has long treated Zhihu as a high-authority source, so a well-written Zhihu answer often ranks on the open web, not just inside the app. One good answer keeps working for months, catching buyers who never open Zhihu at all but land on it through search. That's the opposite of a Douyin video, which is dead within days.
The second, newer multiplier is AI. Zhihu has pushed its own answer engine, Zhihu Direct Answer (知乎直答), which pulls from the platform's archive and — in its updated form — traces responses back to the original answerers. More broadly, Chinese AI assistants lean on the same kind of structured, long-form, opinionated text that Zhihu is basically made of. When a buyer asks an AI "is this brand any good," the model is reaching for exactly the sort of considered answer Zhihu is full of. Getting cited there is starting to look less like PR and more like being in the training set of your own category's reputation. Foreign brands underrate this badly — it's the same shift I wrote about in what AI changes about cross-border marketing.
How a foreign brand actually shows up
You cannot buy your way to credibility here the way you can rent reach elsewhere. But there is a real playbook, and it runs roughly in this order:
- Set up an institutional account (机构号). Zhihu's official brand account requires a Chinese business license, so most foreign brands run it through their local entity or partner. It's your on-record voice — for product explainers, spec deep-dives, and answering the technical questions your category actually gets.
- Find the questions that are already ranking. Don't invent threads. Search your category, find the high-traffic "which should I buy" questions, and answer the ones where you genuinely have something to say. One strong answer on a question that already pulls search traffic beats fifty on dead threads.
- Answer like an expert, not a brochure. The platform punishes marketing language. The answers that win acknowledge trade-offs, name when a competitor is the better pick, and show real knowledge. Counterintuitively, admitting what you're not for is the fastest way to be believed about what you are for.
- Seed through credible professionals, not celebrities. Zhihu weight comes from verified expertise — an engineer, a doctor, a category specialist — not follower count. A profile that lists real credentials carries more here than a famous face.
- Use paid amplification (Zhi+ / 知+) to extend the good answers, not manufacture fake ones. Zhi+ boosts distribution of an existing answer or article through Zhihu's own search and recommendation. Amplify content that's already earning its keep; don't pay to push a brochure nobody would have upvoted.
The mistakes I see foreign brands make
Three, over and over. First, treating Zhihu like Red — dropping in glossy lifestyle posts that get quietly ignored, because the audience came for substance. Second, astroturfing: paying for a wall of suspiciously positive answers, which Zhihu users are unusually good at smelling and which can turn a thread against you. Third, and most common, expecting a campaign timeline. Zhihu is a compounding asset. The answer you write this quarter is still catching buyers next year through search and AI. Measured over one month it looks like a failure; measured over a year it's often the cheapest trust you'll buy in China.
How to measure it
Not followers. Track whether your brand's key questions have a strong, high-ranking answer with your point of view in it — coverage of the decision, in other words. Watch upvotes and saves on those answers (saves especially: a save is a buyer bookmarking your argument for later). Check whether your Zhihu answers show up in Baidu results for category queries. And increasingly, ask the AI assistants your buyers use what they say about your category, and see whether the framing sounds like yours. That last one is a rougher signal, but it's the one that's growing.
Bottom line
Zhihu won't give you a viral moment and it won't move numbers in a month. What it gives you is ownership of the last honest question a Chinese buyer asks before they pay — and, now, a foothold in the AI answers that increasingly settle that question for them. For high-consideration categories that's not a nice-to- have; it's the difference between being researched-and-chosen and being researched-and-dropped. Small and slow, yes. Also durable, and mostly uncontested by the foreign brands still spending everything on reach.
If you're mapping which China platforms actually deserve your team's time — and Zhihu is one of the ones most brands skip too fast — that's the work I do, so reach out. And for the platform on the other end of the funnel, read how to get found on Xiaohongshu search.