How AI reads this page

Text

The page content in Markdown, without menus or graphics. It is the format closest to how a language model reads text.

# Structured data and AI: which schema types actually matter for GEO

> Google says AI needs no special schema, and an Ahrefs study agrees. Which structured data types to use, where to put them and which mistakes to avoid.

URL: https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono
Language: English
Version IT: https://geosnap.ai/blog/dati-strutturati-schema-geo-quali-servono
Author: Rinald Sefa, CMO Geosnap · Category: Guides
Publisher: Geosnap (Maind Group S.r.l.), https://geosnap.ai

## In short

No schema type gets you cited by AI: Google says so explicitly, and an Ahrefs study of 1,885 pages found no increase in citations after JSON-LD was added. Structured data is still useful to describe your organization, locations and products precisely, as long as it matches the page.

- Google requires no special structured data for AI Overviews and AI Mode: classic SEO rules apply.
- In the Ahrefs study, adding JSON-LD did not measurably increase citations in AI Overviews, AI Mode or ChatGPT.
- Organization, LocalBusiness and Product with Offer matter most, when complete and consistent with your feed and Business Profile.
- Since May 7, 2026 Google no longer shows FAQ rich results: FAQPage markup can stay, but don't add it for AI.
- The worst mistake is markup that differs from the visible text or conflicts with other data about your brand.

Written by the Geosnap AI agent, reviewed and approved by Rinald Sefa.

Translated from the Italian original. [Read the original](https://geosnap.ai/blog/dati-strutturati-schema-geo-quali-servono)

Schema.org structured data shows up in almost every list of tips for "getting found by AI": add the right markup and ChatGPT or Google AI Mode will cite you more often. In 2026 we finally have official guidance and measured data to check how much of that holds up. The short version: there is no schema "for AI", but a few types are still useful, as long as you use them for the right reasons.

## Do you need special structured data to appear in AI answers?

No. On its [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features) page, Google says there are no additional requirements to appear in AI Overviews or AI Mode, and no special Schema.org structured data to add. To be used as a source, a page must be indexed and eligible to appear in Search with a snippet: the same conditions as classic SEO.

The guide to optimizing for generative AI features that Google published in May 2026 points the same way. According to [Search Engine Journal](https://www.searchenginejournal.com/googles-new-ai-search-guide-calls-aeo-and-geo-still-seo/575026/), it repeats that structured data is not required for generative search, that you do not need files or markup built specifically for AI, and that you do not have to break content into small chunks. Markup is still recommended as part of a normal SEO strategy, mainly for rich results.

There is one rule Google keeps repeating, though: structured data must match the visible text. Markup that says something different from the page is a problem, with or without AI.

## What the data says: adding schema does not increase citations

The most robust study so far was published by [Ahrefs in May 2026](https://ahrefs.com/blog/schema-ai-citations/). Across 6 million URLs, pages cited by AI were almost three times as likely to have JSON-LD as other pages. That looks like proof, but it is a correlation: well-maintained sites do many things right at once, schema included.

To isolate the effect, the team followed 1,885 pages that added JSON-LD between August 2025 and March 2026, matching each one with three similar pages from other domains that did not. It then measured citations in the 30 days before and the 30 days after the change:

| Platform | Estimated effect | How to read it |
| --- | --- | --- |
| Google AI Overviews | −4.6% | Small drop; the comparison pages were declining too |
| Google AI Mode | +2.4% | Not distinguishable from zero |
| ChatGPT | +2.2% | Not distinguishable from zero |

The authors list the limits as well: the pages studied were already heavily cited, the different schema types were pooled, a 30-day window may miss slow effects, and only JSON-LD present in the initial HTML was tested. The takeaway is still clear: if a page is already visible to AI, adding schema will not push it higher.

## So what is structured data good for in 2026?

It describes facts precisely that a system would otherwise have to infer from text: who you are, what you sell, at what price, where you are. It is not a lever for citations, but it removes ambiguity in a few concrete cases.

- **Products and prices.** OpenAI explains that for shopping results [ChatGPT considers structured metadata](https://help.openai.com/en/articles/11128490-improving-the-visibility-of-your-store-in-chatgpt-shopping) from first-party and third-party providers, such as price and description, and lets merchants apply to send a direct product feed.
- **Engines that say it helps.** In 2025 Microsoft's Fabrice Canel said, as reported by [Search Engine Land](https://searchengineland.com/microsoft-bing-copilot-use-schema-for-its-llms-453455), that Bing uses schema to help its large language models understand content, Copilot included.
- **Rich results and Google's own data.** In classic Search, markup makes a page eligible for rich results, which remain the main reason Google recommends it. On the same AI features page, Google also asks you to keep Merchant Center and your Business Profile up to date.

## Which schema types to use, by type of site

The practical rule is a few complete, consistent types rather than many half-filled ones. This table sums up the priorities for the most common cases.

| Type | Where | What to fill in | Priority |
| --- | --- | --- | --- |
| Organization | Home page or About page | Name, logo, website, contacts, official profiles in `sameAs` | High for everyone |
| LocalBusiness | Location page or pages | Address, opening hours and phone, identical to your Business Profile | High for local businesses |
| Product and Offer | Product pages | Price, currency and availability, identical to your feed | High for online stores |
| Article | Blog posts | Author, publication date and last update | Medium |
| BreadcrumbList | Inner pages | Navigation path | Medium |
| FAQPage | Pages with real questions and answers | Visible questions only, one answer each | Low |

FAQPage needs a note. Since May 7, 2026 Google [no longer shows FAQ rich results](https://www.searchenginejournal.com/google-drops-faq-rich-results/574429/) in Search, and it has removed the related Search Console report. The markup can stay on your pages without causing problems, but there is no point adding it everywhere in the hope of more citations.

Take a small travel agency: Organization on the home page, LocalBusiness with an address and opening hours that match its Business Profile, and Article on the blog are enough. An online furniture store, on the other hand, should focus on Product and Offer, because price and availability are exactly the data an AI assistant compares.

## The mistakes that do the most damage

The costliest mistakes are not about which schema type you pick, but about consistency between markup, page and the rest of the web.

1. **Markup that differs from the page.** Google's [structured data guidelines](https://developers.google.com/search/docs/appearance/structured-data/sd-policies) forbid marking up content that is invisible or misleading. A violation can lead to a manual action, and the page loses eligibility for rich results.
2. **Conflicting data.** One price in the schema, another in the feed and a third in the text confuse anyone reading the page, people and automated systems alike. The same goes for addresses and hours that differ between your site and your Business Profile.
3. **Schema added only through JavaScript.** The Ahrefs study only looked at JSON-LD in the initial HTML, so it says nothing about markup injected later. If your site builds content in the browser, check that text and markup are already in the HTML crawlers receive, and that crawlers can access your site at all (see our [guide to AI crawlers](https://geosnap.ai/en/blog/il-tuo-sito-blocca-chatgpt-guida-crawler-ai)).
4. **Schema instead of content.** No markup makes up for a page that does not answer the question. A clear text comes first; its structured description comes second.

## How to tell whether schema actually helps you

Measure instead of trusting promises. Ahrefs suggests a simple test: add schema to one group of pages, pick similar pages as a comparison group, and watch both for at least 30 days.

- Choose the questions your customers ask AI assistants and track over time whether and how your brand appears. Geosnap does this on ChatGPT, Gemini and Google AI Mode; what matters is using the same list of questions before and after.
- Check in Search Console that your markup has no errors, and compare impressions and clicks of the changed pages with the comparison group.
- Fix inconsistencies in prices, addresses and names first, and only then add new types.

For the other big technical myth of the year, read what the data says about [llms.txt](https://geosnap.ai/en/blog/llms-txt-serve-davvero-dati-2026). You can see how AI answer monitoring works on the [features](https://geosnap.ai/en/features) page.

## Sources

- [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features) (Google Search Central, 2025-12-10)
- [General structured data guidelines](https://developers.google.com/search/docs/appearance/structured-data/sd-policies) (Google Search Central, 2026-07-10)
- [We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved.](https://ahrefs.com/blog/schema-ai-citations/) (Ahrefs, 2026-05-11)
- [Shopping with ChatGPT Search](https://help.openai.com/en/articles/11128490-improving-the-visibility-of-your-store-in-chatgpt-shopping) (OpenAI Help Center, 2026-10)
- [Google Drops FAQ Rich Results From Search](https://www.searchenginejournal.com/google-drops-faq-rich-results/574429/) (Search Engine Journal, 2026-05-10)
- [Google's New AI Search Guide Calls AEO And GEO Still SEO](https://www.searchenginejournal.com/googles-new-ai-search-guide-calls-aeo-and-geo-still-seo/575026/) (Search Engine Journal, 2026-05-15)
- [Microsoft Bing/Copilot use schema for its LLMs](https://searchengineland.com/microsoft-bing-copilot-use-schema-for-its-llms-453455) (Search Engine Land, 2025-03-20)

## Frequently asked questions

### Does Schema.org markup help you get cited by ChatGPT?

There is no evidence that adding it increases citations: in the Ahrefs study of 1,885 pages, the effect on ChatGPT was indistinguishable from zero. For shopping, however, OpenAI says it considers structured metadata such as product price and description.

### Should I remove FAQPage markup after Google's change?

You don't have to. The markup can stay without causing problems; Google simply no longer shows FAQ rich results. Remove it if the marked-up questions are not visible on the page, because that breaks the guidelines.

### JSON-LD, Microdata or RDFa?

Google supports all three but recommends JSON-LD, which is easier to manage and check. Make sure it is in the page's HTML rather than only injected later with JavaScript.

### Which structured data does a local business need?

Organization or LocalBusiness with a name, address, opening hours and phone number identical to your Google Business Profile. Consistency across your site, profile and directories matters more than the number of types you use.

Structured data

Schema.org data (JSON-LD) embedded in the page code. It tells the AI explicitly what the page is about and who published it.

{
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Home",
      "item": "https://geosnap.ai/en/"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "Blog",
      "item": "https://geosnap.ai/en/blog/"
    },
    {
      "@type": "ListItem",
      "position": 3,
      "name": "Guides",
      "item": "https://geosnap.ai/en/blog/#guide"
    },
    {
      "@type": "ListItem",
      "position": 4,
      "name": "Structured data and AI: which schema types actually matter for GEO",
      "item": "https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono"
    }
  ]
}

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "@id": "https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono#article",
  "headline": "Structured data and AI: which schema types actually matter for GEO",
  "description": "Google says AI needs no special schema, and an Ahrefs study agrees. Which structured data types to use, where to put them and which mistakes to avoid.",
  "inLanguage": "en",
  "url": "https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono",
  "mainEntityOfPage": "https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono",
  "image": "https://geosnap.ai/img/blog/p2026-10-11-dati-strutturati-en.webp",
  "articleSection": "Guides",
  "wordCount": 1255,
  "author": {
    "@type": "Person",
    "name": "Rinald Sefa",
    "jobTitle": "CMO",
    "url": "https://geosnap.ai/en/about",
    "worksFor": {
      "@id": "https://geosnap.ai/#organization"
    }
  },
  "publisher": {
    "@id": "https://geosnap.ai/#organization"
  },
  "isAccessibleForFree": true,
  "abstract": "No schema type gets you cited by AI: Google says so explicitly, and an Ahrefs study of 1,885 pages found no increase in citations after JSON-LD was added. Structured data is still useful to describe your organization, locations and products precisely, as long as it matches the page.",
  "datePublished": "2026-10-11",
  "dateModified": "2026-10-11",
  "citation": [
    "https://developers.google.com/search/docs/appearance/ai-features",
    "https://developers.google.com/search/docs/appearance/structured-data/sd-policies",
    "https://ahrefs.com/blog/schema-ai-citations/",
    "https://help.openai.com/en/articles/11128490-improving-the-visibility-of-your-store-in-chatgpt-shopping",
    "https://www.searchenginejournal.com/google-drops-faq-rich-results/574429/",
    "https://www.searchenginejournal.com/googles-new-ai-search-guide-calls-aeo-and-geo-still-seo/575026/",
    "https://searchengineland.com/microsoft-bing-copilot-use-schema-for-its-llms-453455"
  ],
  "translationOfWork": {
    "@id": "https://geosnap.ai/blog/dati-strutturati-schema-geo-quali-servono#article"
  }
}

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "@id": "https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono#faq",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "Does Schema.org markup help you get cited by ChatGPT?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "There is no evidence that adding it increases citations: in the Ahrefs study of 1,885 pages, the effect on ChatGPT was indistinguishable from zero. For shopping, however, OpenAI says it considers structured metadata such as product price and description."
      }
    },
    {
      "@type": "Question",
      "name": "Should I remove FAQPage markup after Google's change?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "You don't have to. The markup can stay without causing problems; Google simply no longer shows FAQ rich results. Remove it if the marked-up questions are not visible on the page, because that breaks the guidelines."
      }
    },
    {
      "@type": "Question",
      "name": "JSON-LD, Microdata or RDFa?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Google supports all three but recommends JSON-LD, which is easier to manage and check. Make sure it is in the page's HTML rather than only injected later with JavaScript."
      }
    },
    {
      "@type": "Question",
      "name": "Which structured data does a local business need?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Organization or LocalBusiness with a name, address, opening hours and phone number identical to your Google Business Profile. Consistency across your site, profile and directories matters more than the number of types you use."
      }
    }
  ]
}

Metadata

The information in the page header, which crawlers and engines read before the content.

Title

Structured data and AI: which schema types actually matter for GEO 66 characters

Description

Google says AI needs no special schema, and an Ahrefs study agrees. Which structured data types to use, where to put them and which mistakes to avoid. 150 characters

Canonical URL

https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono

Language

English

Other languages

Italian: https://geosnap.ai/blog/dati-strutturati-schema-geo-quali-servono

Instructions for crawlers

index, follow, max-image-preview:large, max-snippet:-1

Markdown version

https://geosnap.ai/en/blog/dati-strutturati-schema-geo-quali-servono.md

Index for AI

https://geosnap.ai/llms.txt

Last updated

11 October 2026