How Snippet Controls Limit What Search Features Can Use

When Google or Bing shows a page in search, it may display a short text preview called a snippet. The search engine chooses that text, often from the page itself, and can choose different excerpts for different queries. A meta description can inform the preview, but it does not guarantee that Google will use that wording, as its snippet documentation explains.

Search engines use snippet controls to limit what they display and, for some controls, what page content their AI Search features can use to compose answers. Each search engine defines how these controls work. Google explicitly connects nosnippet and max-snippet to its AI Overviews and AI Mode; its documentation does not make the same explicit claim about data-nosnippet. A search engine applies these controls only if its crawler can read them. The controls do not govern every AI service or make a page private.

What each snippet control does

A crawler is an automated program that visits a page and reads its instructions. For HTML pages, publishers can put instructions in a robots meta tag or mark a specific section. For files such as PDFs, an HTTP response header called X-Robots-Tag can carry the instruction. Google’s robots meta tag specifications describe these options.

The distinction matters if you want to exclude only a particular paragraph. Google documents data-nosnippet as a snippet control, while its documentation expressly connects page-wide nosnippet and max-snippet rules to direct input for its AI Search features. Bing’s announcement of data-nosnippet support describes its effect on snippets and AI-generated previews. Those statements do not establish how every AI product handles the marked text.

Google also requires the page to be in its search index and eligible to appear with a snippet before it can qualify as a supporting link in AI Overviews or AI Mode. That means a page-wide snippet restriction has broader consequences than hiding one excerpt. Google’s AI features guidance describes the eligibility conditions.

Choose the control that matches your goal

Use data-nosnippet when a particular section should stay out of supported search previews while the rest of the page remains available for snippets. For example, a publisher could mark a short passage containing details they do not want repeated in a preview:

<div data-nosnippet>
  Text to exclude from supported snippets.
</div>

Use max-snippet:N to limit a page’s preview length; for Google, the same limit applies to direct input for AI Overviews and AI Mode. Google has not specified a character limit that guarantees a page will never appear as a featured snippet, so use nosnippet to block Google text snippets rather than relying on a very low maximum.

Google offers a separate Search Console setting to exclude content from specified Google Search generative AI features while keeping it eligible for ordinary Search. As of November 2026, Google says the setting covers AI Overviews, AI Mode, and generative AI features in Discover. It excludes the property’s content and links from those features, but does not affect other parts of Google Search or AI training. Check Settings → Search generative AI in Search Console, including whether the property inherits a setting from a parent property. The Search Console help page describes the setting and its scope.

For confidential information, use authentication, remove the information, or otherwise restrict access. Snippet directives govern search presentation and specified Search features; they do not prevent people or services from accessing a publicly available page.

Check what the crawler receives

Check the deployed page or file first. A content-management setting alone does not show what the crawler receives. Inspect the live HTML for robots meta tags and data-nosnippet attributes. For a PDF or another non-HTML file, inspect the response headers for X-Robots-Tag. Check whether another directive or header adds a stricter rule. Google says conflicting robots rules follow the more restrictive instruction.

Next, check whether the search crawler can fetch the URL. Check robots.txt for a block. Also review CDN, firewall, and hosting rules that could block the URL. Google says it must be able to access a page to read its page-level instructions. If robots.txt blocks the URL, Google may not see a nosnippet directive on the page. A blocked page is therefore not a reliable way to communicate a snippet restriction.

Use Google Search Console’s URL Inspection tool to see how Google received and rendered the URL. Bing Webmaster Tools also offers URL Inspection. These tools help check what the search engine can see; they do not prove that an AI feature used the page or will cite it. Google’s crawling troubleshooting guidance explains how to investigate access and fetch issues.

Finally, check search results for several relevant queries. Record the date, and inspect the text preview and AI feature separately. Snippets can change by query, and a single result does not establish how every query or later crawl will behave.

Where snippet controls fall short

A directive only helps when it is present in the version the crawler reads. For example, a CMS setting might not appear in the deployed HTML, or a rule might apply to the wrong URL. Malformed HTML could also make a section-level attribute behave unexpectedly. Google advises against adding or removing data-nosnippet from existing elements through JavaScript because extraction can happen before or after rendering.

Google says limits on automatically extracted page content do not necessarily apply to content supplied separately through structured data or a license agreement. If a fact must be excluded from a Search feature, inspect structured data as well as the visible page.

A Google directive also does not set policy for Bing or independent AI services. Google’s Search Console generative AI setting covers specified Google Search features, not AI training or other products. Bing’s published statement about data-nosnippet covers its snippets and AI-generated previews; it does not establish how every Microsoft AI system handles marked text.

Google says recrawling for preview-control changes can take several days to several months. Its Search generative AI setting has a separate, shorter stated processing estimate, with possible delays from caching. Bing’s October 2025 announcement said data-nosnippet changes could take from a few seconds to up to a week to appear, depending on its crawl schedule. Treat these as platform estimates, not deadlines.

Key Takeaways

• Match the directive to the scope: Use data-nosnippet for selected text. Set max-snippet to limit preview length. Use nosnippet to block Google text snippets and direct input to its AI Overviews and AI Mode.
• Distinguish Google’s documented limits: Google clearly connects nosnippet and max-snippet to direct AI input, but its documentation does not make that same explicit claim for data-nosnippet.
• Inspect the live response: Check deployed HTML or file headers, crawler access, and the search engine’s URL Inspection report before concluding that a directive works.
• Keep private information private: Snippet settings control specified search displays and features; authentication or removal is needed to restrict access.
• Allow time and verify results: Recrawling and processing take time, and one observed preview does not establish behavior for every query or AI product.

What I'm building

Delegate tasks. Get software.

Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request.

Take a look at vroni.com