Bilibili

Objective

Make advertising the group's first revenue line by increasing what each impression yields, on a young and highly engaged audience but in a Chinese market where advertisers demand demonstrated conversion efficiency. The lever chosen is not more ad load but the quality of the match between the ad and the user.

The deployment

Bilibili is the Chinese community video platform born around animation, video games and interest-based content. In the second quarter of 2026 it reports 117 million daily active users, 371 million monthly active users and 113 minutes spent per day per user. Advertising has become its first revenue line: 3.13 billion RMB over the quarter, up 28% year over year. The official release attributes this growth to the improvement of the ad product offering and to better advertising efficiency. The mechanism behind that efficiency, described by CEO Rui Chen in the prepared remarks of the August 27, 2026 earnings call, is AI-driven matching: the system deepens its understanding of product information, of the creatives supplied by the advertiser, of user interests and of the conversion objective the campaign targets, then serves the ad judged most relevant. The metric the group reports on for this work is CTCVR, the product of click-through rate and conversion rate, which Bilibili defines in its glossary as the measure of an ad's effectiveness at triggering an action. That CTCVR is up 19% year over year in the second quarter. In concrete terms, for a user, this means seeing in the feed or on a watch page an ad chosen on its proximity to their interests and to what the advertiser is trying to obtain, rather than on declared targeting. Bilibili says it is extending AI beyond matching, toward creative production with AIGC tools and toward campaign management with ad placement agents, the stated goal being to open the ad business to less equipped advertisers. Research and development expense for the quarter, 1.01 billion RMB, is up 16% year over year, an increase that CFO Sam Fan ties to continued investment in AI capabilities. The technical architecture of the engine, the models used and the exact scope of the campaigns concerned are not published.

Results Proof A

The figure comes from the prepared remarks of CEO Rui Chen at the second quarter 2026 earnings call, held on August 27, 2026, and the transcript is published by Bilibili itself on its investor relations site (S1), which makes it a primary document from the subject brand and not a pickup. It is corroborated word for word by the independent transcript published by The Motley Fool (S3), which also carries it in its highlights, tying it to the effect of AI on ad matching efficiency. The quote is anchored on S3, whose text is machine readable, where the official PDF exposes no extractable text layer; the sentence there is identical down to one hyphen, S1 writing year over year where S3 writes year-over-year. The official earnings release (S2) does not cite CTCVR but attributes the 28% growth in advertising revenue to improved ad products and better advertising efficiency, which corroborates the mechanism without quantifying it. Two results were set aside because the source does not attribute them to the setup. First, the 28% growth in the quarter's advertising revenue, which aggregates the broadening of advertiser verticals, the opening of new inventory and the matching effect, without isolating AI's share. Then the doubling of search revenue, presented by the CEO in a separate paragraph devoted to extending monetization toward new high-intent scenarios, so to an inventory opening and not to the matching engine. Note also a reading trap in the secondary coverage: the sentence about revenue doubling concerns the vertical of advertisers from the AI sector, that is, AI companies buying advertising, and not advertising revenue generated by AI.

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

1. 1
   Assemble the advertiser material advertiser and ad sales team

   Collect the product information, the campaign creatives and the conversion objective targeted, and make them readable by the engine.

2. 2
   Understand content and creative AI

   The system analyzes what the advertiser sells and what the creative shows in order to connect them to the interests expressed on the platform.

3. 3
   Model user interests AI and data team

   Platform usage signals, content watched, interactions, feed a representation of interests used for matching.

4. 4
   Pick the ad served AI

   At display time, the engine keeps the ad that maximizes the joint probability of click and conversion for this user and this objective.

5. 5
   Produce and adapt creatives AI and advertiser

   AIGC tools take on part of the creative production, alongside what the advertiser supplies.

6. 6
   Run the campaign AI and sales team

   Placement agents handle campaign setup and adjustment, to open the ad business to advertisers that have no dedicated media team.

7. 7
   Report on CTCVR data team and management

   The ad business tracks the product of click-through rate and conversion rate as its efficiency metric and publishes it in quarterly results.