Open an answer
#In Overview or AEO Insights, open a response card to read the full captured answer. The detail view shows the prompt, model, date, collection method, detected brand mentions, and cited sources when available.
Filter response cards by model or choose Mentioned only to narrow the list to answers that mention your brand. Turn that filter off when you want to investigate where your brand is absent.
Read the surrounding text before judging a mention. A recommendation, comparison, and criticism can all contain the same brand name.
Read collection labels
#The Collected via label records how Searchly obtained the observation. You may see Consumer UI, Claude API, Provider API, API fallback, Legacy UI, or Legacy API.
Those labels distinguish observations captured from a consumer interface from provider API responses and older collection methods. Reports keep a selected methodology separate from other raw history.
No answer displayed records an observed absence of an answer for that prompt. Check that outcome separately from an answer that exists but does not mention your brand.
Check mentions and sources
#The Visibility row shows whether the answer mentions your brand, its recorded position, and whether it also cites your site. The Mentions row shows other detected brands.
Scroll to Sources to open the cited pages. A mention is a reference to a brand in the answer; a citation is a source link. One can appear without the other.
If a source looks relevant, use the Sources report to see how often its domain or URL appears across the selected results.
Export raw answers
#- Set the date, topic, and model filters in AEO Insights.
- Select Export in the header.
- Use the Raw tab to review the matching answer history.
- Select Export raw answers to download the CSV.
The export includes prompt and answer text, model, topic, collection method and quality, outcome, timestamps, and detected brand mentions with position. Summarized answer and citation export tabs are marked Coming soon.
Raw history and report metrics have different scopes. Use collection labels to identify observations before comparing an exported count with a dashboard score.