- paste-ready asset types per case
- 6
- mismatch lookback on monitored answers
- ~90d
- generations per site / period by plan
- 1–8
How it works
Flag mismatches and look-alike hosts
From monitored answers, we surface brand mismatches over a recent window (about the last 90 days) and cases where citations or grounding point at sibling-family hosts instead of your canonical host.
Open a disambiguation case
Each case shows the confused-with name, how many monitored answers it appeared in, and example quotes — so you see the confusion, not just a score.
Generate paste-ready assets
On demand: a differentiation brief, Organization JSON-LD, About-page copy, FAQ, an llms.txt line, and an interactive implementation checklist — plus a read-only Wikidata search to help you anchor a canonical entity (you create the Wikidata item yourself).
Track the case to verified
Cases move Suggested → In progress → Verified (or Dismissed). Plan limits per website and period: Launch 1, Growth 3, Scale 8 generations.
Honest about what it does
What you get per case
Paste-ready assets to strengthen entity signals — not a promise that every model will stop confusing you overnight.
- It detects confusion and equips you with deployable assets — it does not guarantee AI stops confusing your brand or that citations will appear.
- Adoption of files such as llms.txt is voluntary; stronger signals help, they do not force every assistant.
- Wikidata lookup is a read-only helper in-product — establishing the canonical item remains your browser action.