Pre-flight checks
- Google indexation healthy on money pages; no manual actions; Search Console clean.
- Organization schema with sameAs deployed; Wikidata entry claimed where eligible.
- Google-Extended decision documented (default: allow).
Source priorities
- Own the Knowledge Graph card — consistent entity facts across schema, Wikidata, directories, press.
- Reference-tier presence — Wikipedia-adjacent sources and YouTube (heavily cited classes in Overview analyses).
- Answer blocks on pages already ranking — grounding pulls from the index; structure converts ranking into citation.
30-day plan
| Week | Moves | Checkpoint |
|---|---|---|
| 1 | Entity audit across schema/Wikidata/directories; fix collisions | Knowledge-panel facts correct |
| 2 | Answer-block + FAQ reformat of top-10 ranking pages | Rich results validate |
| 3 | Fan-out coverage: sub-question passages on money queries | Overview citation cards include you on test queries |
| 4 | Conversational testing in the Gemini app; log follow-up behavior | Brand persists through multi-turn answers |
Gemini-specific failure modes
- Knowledge Graph serving stale facts — fix at schema + Wikidata + top directories simultaneously; the Graph corroborates, it doesn't trust one source.
- Ranking without citation — page ranks but never appears in Overview cards: passage shape problem, not an authority problem.
Frequently asked questions
Should I optimize for the Gemini app or AI Overviews?
Both share the index and grounding machinery — one body of work. Test both surfaces: Overviews for one-shot queries, the app for conversational follow-ups. The app is where multi-turn persistence (staying named through a conversation) is won.