Send a case for careful reading

Variazione Civica Lab studies how GPT-style answer engines answer Italy-related questions. The focus is on language shifts, freshness lag, source capture and entity substitution: the four anchor patterns that often explain why one business is named and another disappears.

The lab does not sell rankings, certify businesses or promise that a page will be named by an answer engine. Its work is qualitative and observational. It can show where a claim seems fragile, but it does not turn that fragility into a guarantee.

The notes are best read as field evidence. A business owner or consultant can compare the prompt, named entities and source trail against their own pages, then ask whether the answer engine had enough clear evidence to work with.

Yes. The lab welcomes cases involving Italian places, services, food categories, operators, public-facing organizations and ambiguous business names. The strongest suggestions include the exact wording of the prompt and a short explanation of why the answer looked wrong or unstable.

The team reads relevant submissions and replies when the case fits its research scope or raises a useful methodological issue. It cannot investigate every message as a private audit.

Replies depend on fit, clarity and the team’s review queue. A narrow, well-documented case is easier to assess than a broad request such as “check our AI visibility.”

The base site copy is in English. The lab’s research often compares English and Italian prompts, and it may discuss regional phrasing when that wording changes the answer.