Zhihu seeks the optimal solution between AI and profitability

Author | Liu Baodan

Editor | Huang Yu

AI is a certain opportunity, but it is also a huge cash-burning investment, especially for companies struggling with profitability. How to solve the profitability problem while betting on AI is a topic that needs to be handled with care.

As a content community that has been online for 13 years, Zhihu is providing its own answer through practical actions.

On November 26, Zhihu released its Q3 2024 performance report, which showed that Zhihu's revenue in the third quarter was 845 million yuan, with a net loss reduced by 96.8% year-on-year to 9 million yuan. This is the lowest net loss quarter for Zhihu since its listing in 2021.

Zhihu's reduction in losses has also set a new high since its listing. Recently, Zhihu's founder, chairman, and CEO Zhou Yuan answered the question of when the company would break even with "soon." The loss of 9 million yuan also means that Zhihu is just one step away from achieving complete profitability.

The financial report shows that Zhihu has put in a lot of effort to reduce losses. Compared to revenue growth, Zhihu has focused on improving efficiency and cost management.

Overall, Zhihu's revenue has slowed down. In the third quarter, Zhihu achieved total revenue of 845 million yuan, a decrease of 17% compared to the same period last year.

Zhihu's business model based on content scenarios includes paid reading services, marketing services, and vocational education services, which achieved revenues of 459 million yuan, 257 million yuan, and 105 million yuan respectively in the third quarter. For the first time, the revenue from paid reading services accounted for more than half, reaching 54%, supporting a significant portion of Zhihu's revenue.

In terms of cost efficiency, Zhihu's cost of revenue and operating expenses in the third quarter were 305 million yuan and 625 million yuan respectively, with a year-on-year decrease of over 30%. This is mainly due to reduced content and operational costs, as well as improved technical efficiency leading to lower cloud service and bandwidth costs.

In addition, Zhihu has significantly reduced its revenue in sales and marketing expenses, R&D expenses, and general and administrative expenses year-on-year.

While Zhihu is reducing losses and moving towards profitability, it is also continuously laying out its AI strategy. Unlike many tech companies that are diving into AI large models, Zhihu's AI path is focused on investment and application.

At the end of 2022, ChatGPT became a global sensation. Four months later, Zhihu, as a leading investor, invested in Mianbi Intelligent, and in April this year, Zhihu participated again in its Series A financing. Mianbi Intelligent is a company that accelerates and empowers the application of AI large models. Zhihu's investment in Mianbi Intelligent not only provides more financial flexibility but also lays the foundation for exploring application scenarios.

Zhou Yuan is redefining Zhihu in the AI era and has gradually clarified the direction of AI exploration, which is the search scenario in AI applications, one of the most promising tracks for AI applications.

Referring to Perplexity, a company focused on developing an AI chat-based conversational search engine, it was reported in early November that the company is about to complete a new round of financing, with a valuation reaching 9 billion USD, a 17-fold increase from 520 million USD at the beginning of the year, with investors including NVIDIA and others.

Since the beginning of this year, Zhihu has accelerated its pace in AI applications. In June, Zhihu launched a new AI product "Zhihu Direct Answer," which is based on the real Q&A data of creators and represents the formal productization of Zhihu's AI search function, meaning that Zhihu has officially participated in the battle for the implementation of AI large models "Zhihu Zhi Da" is also known as the "Chinese version of Perplexity." This product launched its "Professional Search" feature at the end of October, introducing professional content sources such as Weipu and Zhihu Selected, covering over 50 million pieces of Chinese and English literature data. It also supports file uploads and long document parsing, targeting productivity scenarios for professionals.

Zhou Yuan stated that in the third quarter, the user traffic of Zhihu Zhi Da grew rapidly, and the "Professional Search" feature launched in October will become Zhihu's differentiated exploration towards deeper scenarios.

However, although Zhihu has balanced well in terms of profitability and AI layout, progressing towards their respective established expectations, this balance comes with considerable pressure.

Zhihu's profitability is almost a certainty, but an implicit prerequisite for this profitability is cost control, which has already affected Zhihu's user growth. According to the comprehensive financial report, since a year ago, Zhihu's monthly active users have begun to decline, and in the third quarter, Zhihu's monthly active users fell below the peak of 100 million.

Zhou Yuan mentioned in the earnings call that although optimizing advertising has led to a decrease in monthly active users, they have increased efforts in enhancing user product experience. However, they cannot yet provide clear numerical growth guidance in this area, but it can be confirmed that Zhihu will continue to increase investment in core user experience.

In terms of AI, although Zhihu has made significant investments, Zhou Yuan recently admitted at the 2024 World Internet Conference in Wuzhen that "whether domestically or globally, in terms of the number of users, frequency, and the scenarios it can cover, AI search is still in a relatively early stage."

In his view, the biggest challenge in developing AI search is that consumers have not fully adapted yet. The difficulty lies in the depth of scenario engagement, which is still far from sufficient to fully meet users' pain points.

After more than three years of being listed, Zhihu is finally about to receive good news regarding profitability. However, in the rapidly approaching AI era, Zhihu still has more challenges to address, which may be a more arduous exploration path