Deutsch
Retrieval & grounding

Grounding page

A grounding page is a web page that states verifiable facts about one entity, such as a company, person or product, in plain language and in structured data. The idea comes from the SEO community. It is not a term used by search engine providers.

What does a grounding page contain?

The initiative behind groundingpage.com describes eleven building blocks in its specification, among them a clear entity name in the H1, a short lead definition, self-contained H2 sections, visible structured facts, a distinction from similar entities, sources with dates, and JSON-LD that mirrors the visible content. The spec says readability for humans stays decisive and compares the intended tone to an encyclopedia entry rather than a campaign page.

Independent write-ups such as SISTRIX and Seokratie name typical content: name, founding year, location, products or services, a distinction from similar offerings and an FAQ section for common ambiguities.

Retrieval-augmented generation: retrieve first, then answerA question triggers a search in a search index or vector database. The best passages are added to the prompt together with the question. The language model writes the answer from them and points to the sources of the passages. The answer can only cite what was retrieved first.1 QUESTION2 RETRIEVAL3 CONTEXT4 ANSWERQuestionfrom a useror an appRetrievalsearch orvector search findsmatching passagesKnowledge sourcesearch index, web,vector databasePrompt with contextquestion + passages[1] passage · URL A[2] passage · URL B[3] passage · URL CLLMAnswertext withsource links[1][3]Citations point to retrieved passagesSchematic, based on Lewis et al. (2020) and Google’s grounding docs.
In retrieval-augmented generation, the system first retrieves matching passages and passes them to the language model as context. Google calls the link between model output and verifiable sources grounding. The answer can only cite what was retrieved first (an inference from the method, not a Google statement). Sources: Lewis et al. (2020): Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (arXiv 2005.11401); Google Cloud: Grounding overview (Vertex AI); Google Cloud: Grounding with Google Search (Vertex AI).

How does it differ from an About page?

An About page builds trust and tells a story. A grounding page lists facts and avoids slogans, so a model or a crawler that retrieves it gets consistent statements.

The word “grounding” also exists in AI development. There, grounding means a model retrieves current web results before it answers, which Google describes for Gemini as grounding with Google Search. A grounding page is a different thing: it is a page someone hopes will be retrieved in that process.

Open questions

SISTRIX points out that neither Google nor OpenAI has confirmed that such pages are handled in a special way. Consistent facts about an organization across its own site and external profiles are useful with or without a dedicated page.

Example

A machine manufacturer publishes /company-facts/ with legal name, founding year, headquarters, product lines, certifications and a link to its commercial register entry. The same data appears as Organization markup in JSON-LD. The page has no slogans or promotions.

FAQ

Is a grounding page an official Google standard? No. The specification comes from the groundingpage.com initiative, and none of the sources checked says Google has adopted it.

Does a grounding page replace schema markup? No. Schema markup supplies structured fields, and the page adds context in text, as SISTRIX describes it.

Sources

  1. Grounding Page Standard, technical implementation – groundingpage.comgroundingpage.com
  2. Grounding with Google Search – Gemini API docsai.google.dev
  3. How AI Search Grounding Actually Works – DEJANdejan.ai
  4. Grounding Page – SISTRIXsistrix.de

Last updated: