Kuaishou

Objective

Grow ad revenue without depending on audience growth, which is flattening (412.7 million daily active users, +1.2% year over year). The issue is value per impression and per query: match ad to user better, automate bidding, industrialize creative production for advertisers who do not produce enough of it, and turn merchant search into a conversion surface. Kuaishou is one of the rare players to publish the incremental contribution of these components to its growth.

The deployment

Kuaishou publishes something almost no platform publishes: the quantified contribution of its AI models to the growth of its ad revenue. In the first quarter of 2026, the company states that deepening its generative recommendation and intelligent bidding large models produced roughly 3.0% to 4.0% growth in domestic marketing services revenue on its own. Total online marketing services revenue reached RMB 19.6 billion, up 9.3% year over year: the models' contribution therefore represents a substantial share of that growth. Three other components complete the setup. The UAX placement solution received an Agent feature that takes over tasks previously handled by marketing teams, including campaign unit creation. AI-generated ad creatives have scaled: in March 2026 they carried 10.0% of all short video ad spending on the platform. Finally, merchant search was rebuilt with OneSearch V2, where Kuaishou says stronger inference capabilities and a better search experience produced roughly 3.0% incremental GMV. In the same quarter, the Kling AI video generation business passed RMB 650 million in revenue, growing more than 300% year over year, giving Kuaishou a second AI monetization stream distinct from advertising.

Results Proof A

The figures come from Kuaishou's unaudited first quarter 2026 results announcement, published on May 27, 2026: contribution of the models to domestic marketing revenue growth, incremental GMV from search, share of generated creatives in ad spending. Primary document from the subject brand. Online marketing services revenue is corroborated by the earnings call summary published the same day. Worth flagging as rare: the company itself isolates the incremental contribution of AI, which removes the need for attribution by inference.

The internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.

1. 1
   Creative production AI

   The advertiser produces its ad materials, relying on AIGC generation for a growing share of the volume.

2. 2
   Campaign setup AI

   Campaign units are created in UAX, with an Agent mode taking over previously manual tasks.

3. 3
   Matching and bidding AI

   Generative recommendation and intelligent bidding models decide which ad is served, to whom, and at what price.

4. 4
   Merchant search client

   Users looking for a product go through OneSearch, whose inference capabilities have been strengthened.

5. 5
   Conversion loop AI

   Completed purchases feed back into the ad models and the search ranking.