Glossary
Grounding
Grounding is tying a model's output to verifiable external sources so each claim in the answer can be attributed and checked. A grounded answer names where its statements came from; an ungrounded one is produced from the model's own parameters and cannot be traced back to any document.
In plain terms
Grounded means the engine can point at the page it got the fact from. That page can be yours.
Why it matters
Grounded answers are where citations live. The more an engine grounds a category of question, the more the work of publishing clear, checkable facts converts directly into visibility.
How to apply it
- State facts with dates, figures and named sources so they are safe to ground against.
- Correct outdated pages: stale figures get grounded too.
- Keep the same fact consistent everywhere it appears.
Example
An AI assistant answering "what is the current U.S. federal minimum wage" grounds its answer in a Department of Labor page rather than stating a figure from memory, allowing the user to click through and verify the claim — the same mechanism a vendor's pricing page can trigger if it states current figures plainly.
Related terms
Back to the full glossary (75 terms).
Ranking is no longer enough
You need to be cited, mentioned, and recommended.
Being cited, mentioned, and recommended are three different outcomes, and most brands only ever achieve the first one. Ranking is no longer enough because AI engines answer buyers directly and name only a short list of vendors as the recommendation — everyone else is cited in passing, if at all. Get a free AI Visibility Report to see exactly where your brand appears today across ChatGPT, Google AI Overviews, Gemini, Perplexity, and Copilot, where competitors are winning the recommendation instead, and what's keeping you from moving up the shortlist.