PUBLIC RESEARCH / CHINA DIGITAL PLATFORMS

Xiaohongshu Is Becoming China’s Search Engine for Everyday Life. AI Is Changing It Again.

People know Xiaohongshu (RedNote) for lifestyle posts and product recommendations. Its more interesting role is helping Chinese consumers decide what to trust, avoid, and buy.

LucidVector · November 12, 2026 · Public Research

Imagine someone in Guangzhou deciding whether to spend ¥300 on a beginner’s painting class.

She could search for local art schools, browse their websites, and compare advertised prices.

But she might also open Xiaohongshu.

Her first search could be 广州 成人油画体验课 — adult oil-painting trial classes in Guangzhou.

The next might be 成人油画课 避雷 — warnings about adult painting classes.

Then 成人油画体验课 值不值 — is an adult painting class actually worth the money?

Notice what happened. She began by looking for a service. Within a few searches, she was investigating whether the service would disappoint her.

For a business owner, those searches are remarkably revealing. They show the questions that must be answered before somebody becomes a customer.

This is the side of Xiaohongshu that international observers often miss.

The platform, also known as RedNote or Little Red Book, is usually introduced as China’s answer to Instagram. Its influence increasingly depends on something more practical: people use it to investigate everyday decisions.

And in 2026, Xiaohongshu began taking that behavior into a new phase with AI-powered search.

A Social App with a Serious Search Business

Xiaohongshu’s official open-platform website describes a community with approximately 300 million monthly active users and 300 million daily search queries.

These are figures published by the platform itself, rather than independent audience measurements. Even so, they illustrate how central search has become to the company’s commercial identity.

Independent research points in the same direction.

In a 2025 survey of Xiaohongshu users, iiMedia Research found that:

  • 55.10% identified obtaining information as a reason for using the platform.
  • 42.35% cited learning new things.
  • 41.78% cited shopping and consumption.

These are survey responses, not a breakdown of all user activity.

The same research reported that beauty and skincare, skills learning, and travel guides were prominent categories of information searched for by respondents.

None of those categories requires a conventional social network.

They require useful answers.

This is why the description “Instagram meets Pinterest” only gets us so far. Xiaohongshu can entertain, inspire, and connect people, but it also helps users solve mundane problems: choosing a restaurant, hiring a photographer, comparing sunscreen, planning a trip, or finding a suitable teacher.

That gives it a distinctive position in China’s consumer internet.

The Five Questions Hidden Inside Xiaohongshu Searches

A typical product search rarely tells the whole story.

Consider an overseas skincare brand exploring the Chinese market. Searching for “skincare” might show popular products and creators. Looking at how consumers refine their searches reveals a much richer picture.

Search termMeaningWhat it reveals
推荐 (tuījiàn)RecommendationWhat options people want to discover
测评 (cèpíng)Review or evaluationWhat they want to compare
避雷 (bìléi)Warnings; things to avoidWhat disappointments they fear
平替 (píngtì)Affordable alternativeWhat makes a cheaper substitute attractive
回购 (huígòu)Repeat purchaseWhat people say is worth buying again

These five expressions describe different moments in the consumer’s decision.

推荐: What should I consider?

The buyer knows the category but has not necessarily chosen a product.

测评: How do the options compare?

She wants details, limitations, and evidence from use.

避雷: What could go wrong?

A negative experience may be more useful than a glowing recommendation.

平替: Can I get similar benefits for less money?

The original product has become a reference point against which alternatives are evaluated.

回购: Would somebody buy this again?

The buyer is looking for an indication of longer-term satisfaction, although an individual post cannot establish genuine repeat-purchase rates.

This sequence is illustrative. Actual users do not necessarily move through these searches in a fixed order.

But the vocabulary provides a useful research framework.

A foreign brand studying the Chinese market could learn more from recurring complaints and comparisons than from a list of trending hashtags.

China’s “Planting Grass” Economy Is More Complicated Than Influencer Marketing

One Chinese expression is especially important to understand: 种草 (zhòngcǎo).

Literally, it means “planting grass.”

In Chinese internet culture, it refers to generating interest in something that another person may want to try or buy.

A creator showing an attractive pair of shoes might 种草 their audience. So might an ordinary customer describing a memorable hotel stay.

The opposite expression, 拔草 (bácǎo), can refer to finally making the purchase or trying something, although usage varies by context.

English-language marketing discussions often translate 种草 as influencer marketing. That captures only part of the idea.

A celebrity endorsement can generate interest. But so can a carefully documented negative review, a comparison between similar products, or an account of how somebody solved an everyday problem.

The commercial value lies in changing how a person understands an option.

A useful illustration comes from iiMedia Research’s 2025 survey.

Among its respondents, 61.33% identified detailed product descriptions and usage experiences as reasons for being influenced by Xiaohongshu recommendations. Visual presentation and celebrity or influencer recommendations also mattered, but detailed experiences led the reported responses.

Another finding is particularly revealing: 46.60% said they would compare more options before buying after viewing product-recommendation content, while 44.19% reported a relatively high likelihood of purchasing.

Those answers describe stated intention, not verified transactions. They also caution against treating every successful recommendation as an immediate sale.

An attractive post may create curiosity. A comparison may sustain it. A negative review can interrupt it.

The consumer is doing research.

For companies accustomed to measuring social campaigns mainly through impressions and engagement, that distinction changes what counts as useful content.

Then Xiaohongshu Added AI to the Search Experience

The next development arrived in 2026.

During the summer, Xiaohongshu expanded its AI search capabilities through a product called 点点 (Diandian, or DotDot).

In August, Chinese media covering the platform’s commercial initiatives reported that the AI search product had been made broadly available in late July.

In September, technology publication Leifeng reported a further consolidation: Xiaohongshu’s earlier AI search feature 问一问, which focused on answering questions from existing user-generated posts, was merged into 点点.

The consolidated product brought search and conversational responses into a more unified experience.

The distinction matters.

A conventional social search might return a collection of posts about beginner painting classes.

A conversational search can address a more complete question:

“I have never painted before, I only have weekends free, and I don’t want to spend much on materials. What kind of class should I choose?”

An AI assistant could help organize relevant considerations and surface material from the platform’s content ecosystem.

That example illustrates the potential use case; it is not a verified transcript from the product.

The significance is straightforward: the person searching can express a problem in natural language instead of guessing which keywords will produce useful results.

The reported early changes

At an August 2026 event, Xiaohongshu representatives described several emerging patterns following the expansion of AI search.

According to the reporting:

  • Longer and more specific search questions were becoming more common.
  • Queries exceeding nine Chinese characters increased by 17% month over month.
  • Long-tail search categories accounted for a reported 16%.
  • Users increasingly asked for explanations, comparisons, current information, and direct answers.

The figures were presented by the platform and describe its reported observations. They do not independently establish that AI caused every change.

Still, they suggest an important shift in how users formulate information needs.

People are becoming more comfortable searching for an answer to a situation, rather than merely searching for a product name.

For businesses, this creates a new problem.

It is no longer enough to ask whether a brand appears when someone searches its name.

A more revealing question is whether the brand’s product, evidence, and limitations are relevant when somebody describes the problem it claims to solve.

Xiaohongshu’s Own Advertising Tools Reveal What the Platform Values

Companies often treat a platform’s ranking algorithm as a mystery.

Xiaohongshu does not publish every detail of its organic recommendation system. However, its official advertising documentation gives an unusually useful look at how commercial search is organized.

In 2026, the company’s 聚光 (Juguang) advertising system documented a set of tools for search campaigns.

Advertisers can research keywords, examine relative demand and competition, identify related searches, and assess how closely a particular post matches the search terms they want to target.

One feature, the relevance precheck tool, evaluates the match between a keyword and an advertising post before campaign delivery.

Another analyzes users’ search paths: the related terms people searched before and after a particular query.

The documentation also describes using AI to identify patterns in these paths.

This is valuable evidence about the platform’s advertising products. It should not be mistaken for a complete explanation of organic search rankings.

But it reveals an important commercial principle.

Xiaohongshu is encouraging advertisers to work backward from user intent.

Its documentation treats a mismatch between a search term and the content of a post as a serious limitation for search-ad performance.

For example, targeting a keyword associated with beginner drawing lessons using generic lifestyle photographs would give a prospective student little reason to trust the result.

A post explaining what beginners actually learn, how much time the class requires, and what materials are included would answer the search more directly.

That does not guarantee better organic reach or a successful paid campaign.

It does make the content genuinely more useful.

What International Brands Often Get Wrong

There are several recurring traps when international companies approach Xiaohongshu as another channel for distributing social content.

They translate the campaign, but not the customer’s question

Suppose a language school wants to promote conversation classes.

Its English marketing might emphasize confidence, flexibility, and expert instructors.

Chinese users may instead search for:

  • 成人英语口语课有用吗 — Are adult English speaking classes useful?
  • 一对一英语口语课价格 — Price of one-to-one English speaking classes
  • 英语口语课避雷 — English speaking classes to avoid
  • 上班族英语口语怎么练 — How should working adults practice spoken English?

The difference is substantial.

The first approach describes what the business wants to sell.

The searches reveal what buyers want to know before committing money.

A stronger content strategy would address those questions directly, with evidence about lesson structure, suitable learners, scheduling, limitations, and expected outcomes.

They mistake attractive posts for evidence of demand

A product can receive attention without creating a profitable market.

On Xiaohongshu, a beautifully presented product may collect views, saves, or enthusiastic comments.

Those signals can be useful, but they do not establish how many people would pay, how much they would spend, or whether a particular foreign business could serve them.

There is also a selection problem: highly visible posts are filtered through recommendation systems, search ranking, promotion, and user behavior.

If a company only studies popular posts, it may mistake popularity for representative demand.

For market research, negative reviews and ordinary discussions deserve attention too.

They forget that RedNote’s international audience is different

In 2026, Xiaohongshu’s domestic service and its international-facing RedNote operations became more clearly differentiated.

An international user’s feed and search results may differ from those seen by a mainland Chinese user.

A brand hoping to reach Chinese domestic consumers should therefore avoid assuming that its own overseas account provides a representative view of the market.

Account eligibility, commercial tools, and content distribution may also depend on region and operating entity.

For a fuller explanation of those differences, see our guide: https://lucidvector.tech/what-is-rednote-xiaohongshu-guide-2026/.

How to Research Chinese Consumer Demand Using Xiaohongshu

You do not need a large advertising budget to begin learning from consumer questions.

A simple research process can already reveal useful patterns.

Consider an overseas business evaluating a premium adult drawing course for Chinese consumers.

Start with the category.

Search 成人绘画课 — adult drawing classes.

Look at the different offers and the language used to describe them.

Add a decision-related term.

Search 成人绘画课 测评 to look for evaluations, or 成人绘画课 避雷 to explore complaints and warnings.

Narrow the context.

Search 零基础 成人绘画课 — beginner adult drawing classes — to examine concerns specific to people without prior experience.

Look for alternatives.

Search 成人绘画 自学 还是 报班 — whether adults should learn drawing independently or enroll in a class.

This last query could be particularly interesting for a course provider.

If people are weighing independent learning against paid instruction, the business must explain what the paid experience offers that free resources do not.

Separate evidence from promotion.

Before drawing a conclusion, compare posts from course providers, individual students, commercial creators, and people describing disappointing experiences.

Check dates, prices, locations, and the specific features being discussed.

An individual complaint is a lead to investigate. Repeated complaints across independent sources are a stronger signal. Neither automatically quantifies market demand.

Record questions, not just trends.

A useful research note might say:

“Potential customers appear concerned about whether beginner classes provide enough individual attention and whether materials are included in the advertised price.”

That is a hypothesis worth testing through additional evidence.

It is much more actionable than writing that adult drawing is trending.

AI Search Makes Evidence More Important

There is a tempting conclusion to draw from AI search: if platforms answer questions automatically, brands should produce more content designed to be quoted by AI.

That is a narrow way to look at the opportunity.

An answer is only as useful as the information supporting it.

If users increasingly ask detailed questions about results, prices, safety, or effectiveness, businesses need credible material that addresses those concerns.

For a course, this might mean a clear curriculum, current prices, instructor qualifications, and examples of actual student work.

For a physical product, it might mean specifications, limitations, appropriate use cases, and comparisons grounded in evidence.

For a service, it might mean explaining who it is suitable for and what a customer should realistically expect.

There is no public basis for claiming that adding particular phrases guarantees inclusion in Xiaohongshu’s AI responses.

The stronger long-term approach is to publish material that deserves to be used as evidence in the first place.

AI search may change how people encounter information. It does not eliminate the need to earn trust.

The Opportunity Beyond Xiaohongshu

Xiaohongshu illustrates a broader change in digital commerce.

Consumers do not always begin their purchase journey on a retailer’s website.

They discover options through social content, investigate them through search, compare personal experiences, and sometimes ask AI systems to organize the information.

For an international business, the practical challenge is learning where those conversations take place and how customers express their concerns.

In China, that requires understanding the local language of purchase decisions.

Words such as 测评, 避雷, 平替, and 回购 are small pieces of vocabulary, but they reveal different reasons why somebody might buy — or walk away.

The real opportunity is to understand those reasons before investing heavily in marketing.

Xiaohongshu happens to be one particularly rich place to observe them.

Further Reading

LucidVector publishes independent research on Chinese consumer behavior, digital platforms, and market opportunities for international businesses.

Our free research includes:

For a broader commercial framework, see China Signal 01: Marketing on the Chinese Internet, our research report on Chinese platforms, consumer discovery, and digital market entry.

https://lucidvector.tech/china-signal-01-overview/

https://lucidvector.tech/get-access/

Sources and Research Notes

  1. https://ad-market.xiaohongshu.com/ — Platform-published audience and search figures.
  2. https://www.iimedia.cn/c400/108275.html — Survey findings on information-seeking, purchase intention, and recommendation content. The publicly available summary does not provide enough methodological detail to treat these percentages as population-wide estimates.
  3. https://ad.xiaohongshu.com/next_help/docs/d6e54d83d20628f893180ec49b790cbd — Keyword planning, search-path analysis, and relevance precheck tools.
  4. https://ad.xiaohongshu.com/next_help/docs/1662aca60c71ebc1eda69e0482f775da — How search-ad relevance is evaluated and used in advertising delivery.
  5. https://finance.sina.com.cn/tech/roll/2026-08-07/doc-inimnvkm3120107.shtml — Reporting on Xiaohongshu’s AI search rollout and platform-reported changes in query behavior.
  6. https://www.donews.com/news/detail/4/6696438.html — Reporting on consolidation of Xiaohongshu’s AI search features into Diandian.

Methodology: This article combines official platform documentation, a third-party survey, Chinese technology reporting, and LucidVector’s analysis. Search examples illustrate a reproducible research method; they are not presented as observations from a controlled experiment or representative consumer sample. Information about advertising tools does not establish how Xiaohongshu’s organic or AI search ranking systems operate.

Enjoyed this research?

Get more independent insights into Chinese markets, platforms, consumers, and business opportunities.