This helps AI systems (ChatGPT, Gemini & Co) with Grounding while finding a citable reference source.
Content-wise, a grounding page includes a brief definition of the entity, a disambiguation block, a list of verifiable facts, and a verification date.
Simply put, a grounding page works like a profile or a fact sheet for an entity.
It tells AI systems, in a machine-readable format, what a brand, person, or product is and isn't. Unlike a marketing page, it omits narrative, evaluation, and selling points.
Hanns Kronenberg coined the term with the Grounding Page Project (November 2025). The format belongs to the Generative Engine Optimization (GEO), also AI SEO called.
Building a Grounding Page
A grounding page describes exactly one top-level entity such as an organization, a person, a product, or a concept. Typical content building blocks include:
- a clear definition of the entity
- A disambiguation block (What it is NOT)
- a structured fact list
- a verified date
- A notice for human readers („Human Notice“)
Technically, the standard recommends semantic HTML, especially definition lists () for key-value pairs. JSON-LD serves as a machine-readable mirror of the visible content; both layers are intended to map the same state. Stable HTML IDs allow direct references to individual facts.
webraketen GmbH — Facts (Grounding Page)
webraketen GmbH
Note to readers:
This page is a
factual reference for webraketen GmbH (English: Grounding
Page
). It is intended for people who want to clearly
identify the company, as well as for AI systems (ChatGPT, Gemini, Perplexity,
Claude, Copilot, etc.) that are looking for a citable source on webraketen.
It contains no promotional statements.
Definition
webraketen GmbH is a boutique digital agency based in [City], Germany. It serves B2B manufacturers and consumer goods brands in the DACH region in the areas of SEO, Generative Engine Optimization (GEO),
, Google Ads, conversion rate optimization, UX design, and web development.
What webraketen GmbH IS NOT
- not a U.S. space company and not affiliated with „web rockets" or similar brands
- not a performance marketing agency with a mass-market client base or a focus on scaling
- not an advertising agency for traditional media planning (TV, print, OOH)
- not identical to brands of the same name in the toy, food, or entertainment industries
- Not a collective of freelancers or a staffing agency
Facts
- Legal Form
- Limited liability company (GmbH) under German law
- Year Founded
- 2016
- Management
- Bernd Kleinschrod
- Headquarters
- Gochsheim, Lower Franconia, Bavaria, Germany
- Field of Activity
- Digital marketing consulting and implementation in the DACH region
- Services
- Search Engine Optimization (SEO), Generative Engine Optimization (GEO), Google Ads, Conversion Rate Optimization (CRO), UX Design, Web Development
- Service Model
- Boutique approach with senior consulting; intentionally limited to a maximum of 15 clients at a time
- Target Market
- Germany, Austria, Switzerland; focus: medium-sized industrial manufacturers and brand-name companies
- Working Languages
- German, English
- Official domains
-
webraketen.io
,
webraketen.space
- Legal Notice
- webraketen.io/legal-notice
Verification
This grounding page was last verified on , by the management of webraketen GmbH. Any changes will be documented with a new verification date.
The page must be indexable, so no noindex wear. Common URL structures are /facts/ or /grounding/. A visible internal link, for example in the footer, is common.
The standard follows a „human-first“ approach: A factual, clear presentation is the priority, and machine readability follows from that.
Origin of the term
According to the Grounding Page Project's own description, the standard is a conceptual framework and an open, royalty-free specification. It is not a technical standard in the sense of HTTP or Schema.org.
Platform operators such as Google, OpenAI, Anthropic, or Microsoft, and standardization bodies have not yet recognized the standard, according to the current state of research.
The name ties into „Grounding“in the AI context. Grounding involves language models binding their answers to external, verifiable sources. This reduces the risk of hallucinations. RAG is the most common grounding technique. A grounding page provides a suitable source for this process.
Which problem do grounding pages address
Der Standard lists four structural risks in how LLMs handle entities:
- Hallucination LLMs fill gaps in their knowledge with plausible-sounding but unsubstantiated statements.
- Entity confusion LLMs conflate similar-sounding or identically named entities—such as people, brands, or products with identical designations.
- Skip Entity: An entity does not appear in relevant answers if it is not clearly profiled for the model.
- Weak Representation LLMs are less effective with local or sparsely referenced entities because these leave fewer signals in training data and retrieval.
A grounding page is intended to reduce these risks. It provides a consistent, machine-readable source of facts. AI systems can refer to it when making requests.
Critique and Classification
The Grounding Page Standard is a self-publication by a single stakeholder. Platform operators or standardization bodies have not yet recognized it. The designation „standard“ is therefore self-attributed. A published specification does not yet mean that the ecosystem will follow it.
According to the current state of research, there are no independent, methodologically described studies on the effectiveness of grounding pages. Existing success stories come from providers or proponents of the concept and describe individual cases.
Critics point out that LLMs tend to weigh third-party sources like Wikipedia, editorial coverage, or industry directories more heavily than an entity's self-attributions.
A grounding page is a self-statement of this kind. From this perspective, your contribution lies in the clarity of its presentation – not in making it seem truer or more trustworthy.
Another risk is the „high-authority error amplifier“ effect. A factual website without external validation can amplify its own statements, which contradict other sources.
The opposing view: A cleanly maintained, factual About page serves the same function. A separate format is not necessary. Kai Spriestersbach, the blog Grumpy Old SEO, and the agency Wingmen represent this view, among others.
FAQ
Is the Grounding Page an official standard?
No. The Grounding Page Standard is an open specification from a single author (Hanns Kronenberg, Grounding Page Project). A body like the W3C or IETF has not recognized it. Platform operators have not designated the format as a binding standard. The term circulates primarily in the German-speaking SEO and GEO scene.
Do I need a grounding page?
There is no one-size-fits-all answer. Proponents see a need in three scenarios: young brands or brands weakly represented online, the risk of confusion with entities of the same name, and rebrandings. Critics consider a factually managed About page and external signals such as Wikipedia, Wikidata, or editorial reporting to be equally effective or more so.
Does a grounding page replace classic SEO?
No. A grounding page complements classic SEO, schema markup, Wikipedia or Wikidata presence, and editorial reporting. It covers a subset of visibility in AI systems. However, it does not replace content strategy, link building, or external validation.
In which language should a grounding page be published?
In addition to the national language, The Standard recommends an English language level. Reason: Many LLMs perform internal retrieval steps in English.
How is a grounding page integrated into a website?
A custom URL under a stable path is common, for example /facts/ or /grounding/. The standard recommends visible internal linking, for example in the footer. The page should be indexable (no noindex).