Zhihu's Pivot: From Q&A Diligence to AI Content and Expert Data Asset Builder

A small-cap Chinese content platform sees its AI bets—content assets, comic dramas, and expert data—start to validate even as core marketing revenue still declines.

ZH · Earnings Call · 2026-08-26

From Q&A Diligence to AI Asset Builder

Zhihu's second-quarter report, released on August 26, tells a story of a company that has fundamentally reframed what it sells. Total revenue of RMB 690 million fell 3.7% year-over-year but grew 5.9% sequentially—evidence that the decline is decelerating, while the mix shifts decisively toward content and IP monetization. The headline is that IP operation is now the engine, with marketing services fading into a structural recovery. CFO Wang Han noted: “The sequential growth was primarily driven by content and operations.” — Wang Han, Chief Financial Officer · 2026-08-26 That is the key change: the company is no longer just an ad platform; it is a content asset factory.

CEO Zhou Yuan framed the new strategy clearly in his prepared remarks:

AI is helping extend these capabilities into new applications and commercial scenarios.
Yuan Zhou, Founder, Chairman and Executive Officer · 2026-08-26

This is not a pivot away from the community—it is a pivot toward using the community's intellectual property and expert network as inputs for AI-native products. The company is building AI to be a AI content asset—a set of brandable, licensable, professional assets that can be cited, reused, and monetized across AI channels. Early results show client count for this offering up 50% sequentially, though it remains too early to call it a scalable revenue stream.

Comic Dramas and IP as the Bridge

The most concrete proof of the new direction is the growth in comic drama licensing. COO Zhang Ronghua was bullish: “We are very positive on the AI comic drama market and believe the industry is still in a clear growth trend.” — Zhang Ronghua, Chief Operating Officer · 2026-08-26 Zhihu's Yanyan Story IP has become a top provider for AI comic dramas on TikTok and ranked in the top three on Hongguo in the first half. This is a validation that the content library has real value beyond the platform itself. The company is also exploring in-house production but remains disciplined, preferring licensing where economics make sense. This is a long-term moat build, not a short-term revenue jump.

The other major AI bet is expert data solutions—essentially a data lab for frontier models. The company is using its expert network to design training data, evaluations, and complex task environments for clients, pitching itself as a research-driven partner rather than a simple data provider. This is a higher-margin, higher-difficulty business that aligns with Zhihu's core strength, but it will require heavier compute and R&D investment. Management is explicit that the investment profile will differ from the traditional content business, and that near-term profitability may fluctuate.

Marketing Services: Structural Recovery, Not Yet Restored

The legacy advertising business remains the weak spot. Marketing services revenue declined 10.7% year-over-year, though it improved 4% sequentially. The company's earlier narrative from August 2025 was more optimistic: “Now we can say that this phase is nearing its end, and we expect to see a stabilization and a recovery in the coming quarters with a return to positive year-over-year growth.” — Unidentified Company Representative, Company Management (likely senior executive or spokesperson) · 2025-08-27 That has not yet materialized; the recovery is happening but largely driven by performance-based advertising and key verticals, not a broad rebound. The CEO's guidance for the second half is cautious: “marketing services remain in a period of structural recovery like we mentioned before and may continue to be affected by changes in client budgets and industrial demand.” — Yuan Zhou, Founder, Chairman and Executive Officer · 2026-08-26 In other words, the fat is still being trimmed, and the market should not expect a sharp V-shaped recovery.

The most telling shift is in Paid content and IP: revenue grew 4.4% year-over-year, with average monthly subscribers steady at 13.1 million and licensing revenue up 600% year-over-year. That is the engine of the future. The company is also aggressively buying back stock—$77.9 million repurchased to date—while its market cap ($270 million) remains far below its net cash of RMB 4.4 billion. This suggests management sees the current price as deeply undervalued, and the AI narrative as the unlock.

Looking ahead, the company expects some quarterly volatility as it balances new investments with cost discipline. The long-term goal remains sustainable profitability, but the path now runs through AI-enabled content and expert services rather than advertising alone. For a small-cap player, this is a genuine strategic pivot with measurable early traction—worth watching for anyone interested in how knowledge platforms will monetize in the AI era.