How it works
The AI engines, and what grounding means
Presence Scout checks four AI surfaces. They behave differently, especially in whether they search the web before answering, and that difference decides whether you get citations.
The engines
- ChatGPT: the most used AI assistant. Its answers are web-grounded, so they carry citations — the same way ChatGPT searches before answering a question like “best CRM for small business”.
- Google AI Overviews: the AI answer Google shows above its own search results. Always grounded, and the one closest to traditional SEO. Google only shows one for some queries, so “No overview” is a real result rather than a gap.
- Google AI Mode: Google's conversational search tab — a separate surface from AI Overviews, and the one Google is steering searchers toward. It always answers, so there is no “not shown” case.
- Perplexity: an AI search engine. Always grounded, and reliably returns its sources.
- Gemini: Google's assistant, grounded with Google Search.
Grounded vs not grounded
- A grounded answer means the engine searched the web first and answered from what it found. So it can tell you which pages it used.
- An ungrounded answer comes purely from the model's training. It can still name you, but there are no sources behind it.
- Only grounded answers produce citations. If your Citations page is sparse, it is usually because few grounded checks have run.
- Grounding is the expensive part of a check, but every plan gets it — an ungrounded answer isn't a cheaper version of the product, it's a different one.
Enable Perplexity if you want citations on every plan. It always grounds, so it fills the Citations page and the Action Plan reliably.