LEARN
Chinese AI Search Starts With Ecosystem Ownership

IN ONE SENTENCE
On the Chinese side, AI platforms and content platforms often share an owner, which gives citation preference a visible ecosystem bias.
A structural feature of the Chinese side is that AI platforms and content platforms frequently belong to the same group. That gives citation preference a visible ecosystem bias — a group's own content platforms are likelier to become its AI's sources.
OUR POSITION
So order investment on the Chinese side by ecosystem ownership rather than platform fame. To be named inside a given AI, first have content on the content platforms its group owns.
What ecosystem ownership changes
Within one group, the data path between content platform and AI product is shorter and content is easier to retrieve. This is not speculation — practitioner observation of citation sources consistently points this way.
The implication is that the same content can have very different visibility across ecosystems. Publishing inside one ecosystem means being visible mainly to that ecosystem's AI.
Two measurement limits
Most Chinese platforms do not list citations per answer, so citation share can only be estimated — and must be labelled as an estimate.
Automated string matching produces more false positives in Chinese: there are no word boundaries, so brand names run into adjacent characters or get captured by same-name terms. Externally reported figures need manual review.
One thing about access
Working the Chinese ecosystem requires allowing the relevant crawler identifiers; some platforms publish none, leaving server logs as the only source of evidence.
⚠️ No published identifier does not mean no crawling. Do not assume absence because no matching agent appears in robots.txt.
Sources
- [1]GEO: Generative Engine Optimization.Aggarwal et al., KDD 2024.2024
Updated 2026-08-10