15 years of year-end profits: Zhihu turned things around with web novels and cost-cutting.

```

On March 25, Zhihu handed in its final report card for fiscal year 2025.

According to the financial report, Zhihu’s revenue for the fourth quarter was 644 million yuan, and total revenue for 2025 reached 2.75 billion yuan; more critically, the company recorded a full-year adjusted net profit (Non-GAAP) of 37.9 million yuan for 2025.

This means that this Chinese Q&A knowledge community, founded 15 years ago and long trapped in the rut of "good reviews but poor sales," has finally crossed the life-or-death line of breakeven and achieved annual profitability for the first time.

However, peeling back the surface of this milestone financial report reveals that this is not an explosive victory in revenue scale, but rather a hard-won success built on extreme cost-cutting, business restructuring, and commercial compromises on the community’s tone.

Amid major industry upheaval, with AI reshaping search entry and content platforms eroding user attention, Zhihu’s profitability is both self-proof of successful commercialization and a ticket into the next brutal elimination round.

A new skeleton propped up by web novels and selling courses

Looking at the annual revenue scale of 2.75 billion yuan, Zhihu’s size is not considered large among big internet firms. The adjusted net profit of 37.9 million yuan gives a profit margin of only about 1.3%.

This "profit on the blade’s edge" reflects Zhihu’s fundamental strategic shift over the past year, from chasing scale growth to preserving profits and cash flow.

From the financial report, Zhihu’s biggest contributor to turning loss into profit lies in strong cost control. Reviewing its financial trajectory over the past few quarters shows that Zhihu has significantly reduced sales and marketing expenses.

In today’s high customer acquisition cost environment, Zhihu has proactively abandoned the traditional internet approach of buying large numbers of users for inflated monthly active users (MAU), choosing instead to activate existing users through refined operations.

Meanwhile, AI technology’s role in internal review and distribution has also meaningfully reduced personnel and bandwidth costs. This 37.9 million yuan profit is more a result of “saving” than “earning.”

On the “open source” revenue side, Zhihu’s revenue engine has undergone a substantive switch, establishing a new base focused on paid memberships, vocational education, and marketing services.

The short story reading business, represented by “Yan Yan Stories,” is currently Zhihu’s most stable cash cow.

Zhihu leveraged long-tail traffic to successfully convert some users with entertainment reading needs into paid subscribers. Nonetheless, this is also a commercial compromise from Zhihu’s early “hardcore professional” community tone.

Against the backdrop of a challenging macro employment environment, Zhihu is leveraging its platform’s user profile of highly educated, first- and second-tier city users to enter postgraduate, IT skills, and civil service exams vocational education tracks.

This sector, with high per-customer value and good cash flow, is becoming Zhihu’s key tool to escape a single path of traffic monetization.

Affected by an overall tightening of ad budgets, simple display ads have reached their ceiling. Zhihu’s marketing services are shifting towards professional trust-based “seeding” and deep brand communications, seeking to maintain its base against the traffic dominance of short video platforms.

Escaping the “community trap” with AI

If profitability on the financial report proves Zhihu can survive, its industry positioning in today’s content track will determine how well it survives.

Zhihu is currently in a tough, conflicted environment.

In the competition for user attention, Zhihu faces structural pressure.

On one hand, Xiaohongshu, with its decentralized algorithm and strong “practical life” label, has essentially taken over the younger generation’s consumption decisions and search for life experiences, directly “bleeding” Zhihu’s “experience sharing” section.

On the other hand, video platforms like Douyin and Bilibili have used more visually impactful means to reconstruct knowledge popularization into video formats.

In the era of fragmentation, Zhihu, mainly offering text and pictures, is naturally on the defensive in terms of media form.

The truly disruptive force comes from AI.

Since 2025, various AI-native search engines have become popular, changing users’ interaction logic for finding answers: from browsing many long articles to seek consensus, to directly getting the unique answer distilled by AI. This directly threatens Zhihu’s traffic entry point based on Q&A.

However, users have also felt Zhihu’s value during this process. The Achilles' heel of AI large models is now Zhihu’s biggest strategic asset: an intense demand for high-quality Chinese corpora.

Within a closed Chinese internet ecosystem full of low-quality information, Zhihu’s 15 years of accumulated professional discussion, peer review, and deep articles are the most indispensable training fuel for large models to reduce hallucinations and improve logical reasoning.

This data asset barrier, based on genuine human experience, enables Zhihu to gain a high “water seller” premium in the AI value chain.

This is only one side of the competitive landscape. For Zhihu, the first year of profitability does not mean the road ahead will be easy. Capital markets typically hold even stricter growth expectations for profitable companies.

Zhihu currently faces an unavoidable paradox common to content communities: profitable businesses such as web novels, marketing articles, and course-selling ads are irreversibly damaging the community’s professional and elite atmosphere; yet, if the core community creators leave, Zhihu’s high-quality corpus will dry up.

In summary, Zhihu’s profitability in 2025 is a self-rescue by retreating to advance. It has shown the industry that a knowledge community can run a financial model through extreme efficiency management and commercial trials.

But in the next knowledge acquisition revolution led by AI, the main question Zhihu needs to answer is no longer “how to make money,” but how to use its massive data moat to reconstruct platform value before being disrupted by new technology.

Risk warning and disclaimerThe market carries risks; investment must be cautious. This article does not constitute individual investment advice and does not take into account individual users’ special investment goals, financial conditions, or needs. Users should consider whether any opinions, views, or conclusions in this article fit their particular situation. Anyone investing based on this must bear full responsibility. ```