Google uses Schema structured data to understand page content, gather information about people or companies, and display rich results when requirements are met. Structured data provides explicit, machine-readable clues, but it does not guarantee that a page will get a special appearance in search results. Google does not support every schema.org type; eligibility for a particular search feature still depends on that feature’s required properties, content policies, and quality requirements.
Helping Google understand the information on a page
Structured data describes and classifies content in a standardized format, helping Google interpret pages more accurately. In a recipe, for example, ingredients, cooking time, and calories can each be marked up rather than described only in text.
Google’s recipe documentation explains that this markup allows users to search for recipes by ingredients, calories, cooking time, and other criteria. It can also be used to display recipe rich results. Structured data does more than affect how search results look: it helps systems understand what each piece of information represents.
A Schema definition does not mean Google offers a corresponding feature
The schema.org vocabulary covers many services and uses, but Google Search supports only part of it. When choosing markup, first identify the search feature you want to use, then follow the requirements in Google Search Central’s documentation. Do not assume that information supplied through a property will appear in search results simply because schema.org defines that property.
- Required properties: These must be provided to meet the feature’s eligibility requirements.
- Recommended properties: These can supply additional information, but completeness and accuracy should take priority over the number of fields.
- Other properties: These may be useful to other services, but that does not mean Google Search requires them.
Valid markup does not guarantee rich results
Google explicitly states that using structured data does not guarantee that the associated features will appear in search results. Along with having correct syntax, markup must comply with general guidelines and feature-specific policies.
For example, markup should be placed on the page where the information belongs and describe that page’s content. You must not create a separate blank page containing only structured data. Nor should you mark up information users cannot see, even if it is accurate. The Rich Results Test can help validate markup, but issues such as spam or inappropriate markup may be unrelated to syntax, so the tool may not detect them.
JSON-LD is easier to maintain, not a shortcut to rich results
If your website’s setup allows it, Google recommends JSON-LD because it is generally easier to implement and maintain at scale and less prone to implementation errors. Unless otherwise stated, Google also supports Microdata and RDFa. As long as the markup is valid and correctly implemented according to the feature’s documentation, none of the three formats has an advantage in terms of search feature support.
Google can also interpret JSON-LD dynamically inserted through JavaScript or CMS widgets, provided that the rendered HTML contains the structured data. During implementation, check not only the code but also the page content Google actually receives. After deployment, continue monitoring for template or service issues that could cause the markup to stop working.
Google structured data: perspectives from across the web
Do Google’s AI search features require special Schema markup?
No. Google says that existing SEO fundamentals apply to AI Overviews and AI Mode, with no need to add special schema.org structured data or AI-specific machine-readable files. To appear as a supporting link, a page must still be indexed, be eligible to appear in search with a snippet, and meet technical requirements. Even meeting these conditions does not guarantee display. Structured data can still be used for other search features, and its content should match the text displayed on the page.
How does enriched search use structured data?
Enriched search allows users to search by the properties of marked-up items, such as finding recipes within a particular cooking time. It is based on structured data and search ranking algorithms; recipes, job postings, and events are supported types. Pages must provide the required properties for the relevant type and comply with quality and content policies. Enriched search applies to pages describing a single item in detail, rather than category pages that merely list multiple items.
Can structured data still be used when search snippets are restricted?
It may be. Google explains that most structured data is not subject to snippet restrictions set by robots meta tags. For example, recipe structured data may still be used in a recipe carousel even when preview text is restricted. For structured data describing creative content, max-snippet can be used to limit the preview length of the description. Data may still be used in search results even if it is inside a data-nosnippet element. If you do not want to provide a particular piece of data, modify or remove the markup directly.
Which does Google use when image structured data conflicts with IPTC metadata?
If both methods are used to provide image information but their contents conflict, Google says it will use the structured data. IPTC photo metadata is embedded in the image itself and can travel with it across different pages. Where the two serve similar purposes, keeping the information consistent can prevent different fields from giving contradictory descriptions of the same image. This precedence applies to image metadata and should not be treated as a general rule for all data sources.
Can Google still update structured data when a website temporarily returns a 503 status code?
When pages return a 503 status code, Google’s systems cannot refresh the website’s titles, descriptions, metadata, or structured data. If this information is changed during a service suspension, you should therefore not expect Google to update it at the same time. Google’s guidance on disabling websites also says to continue allowing crawling and avoid having robots.txt return a 503 status code, which could block crawling of all pages.
References
- 產品子類結構化資料 (ProductGroup、Product) | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Source used in the article above.
- 多元化的互動式搜尋結果 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: How does enriched search use structured data?
- Google 搜尋生成式 AI 功能最佳化指南 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: Do Google’s AI search features require special Schema markup?
- 電子商務網站的相關結構化資料 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Source used in the article above.
- 結構化資料標記的運作方式簡介 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Source used in the article above.
- 漫遊器中繼標記規格 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: Can structured data still be used when search snippets are restricted?
- AI 功能與您的網站 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: Do Google’s AI search features require special Schema markup?
- Google 圖片 SEO:圖片中繼資料 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: Which does Google use when image structured data conflicts with IPTC metadata?
- 利用 JavaScript 產生結構化資料 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Source used in the article above.
- 暫停或停用網站 | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Discusses: Can Google still update structured data when a website temporarily returns a 503 status code?
- Google 搜尋的生成式 AI 內容使用指南 (適用於網站) | Google 搜尋中心 | Documentation | Google for Developersdevelopers.google.com — Source used in the article above.

