Yichi Garden
In AI category recommendations produced with web search enabled, Yichi Garden appears in third position.
Source: AI category recommendation results with web search enabled.
Client
Yichi Garden is a café and restaurant chain, and a client of the AI growth (China) line, delivered as a GEO advisory engagement.
The problem
Hospitality is a recommended category: people do not ask whether a specific brand is good, they ask which places in a city are worth going to. The model answers with a ranked list, and anything below the first few entries is invisible. The governing metric here is rank, not mention rate.
What we did
GEO focused on that one question shape: make the brand’s description consistent across the web-retrievable sources, complete the category keywords and store-level facts, and give the model enough quotable material to place the brand when it assembles a recommendation list.
Limits of this case
- Rank is less stable than mention rate. Web-connected results shift with the live source pool, and the same question can produce a different order at a different time.
- Category recommendations lean heavily on third-party review sites, media and local sources. A brand’s own content has less leverage here than in B2B.
- The result corresponds to a specific question phrasing, city and category. It does not represent nationwide or all-category performance.