WebSite SearchAction and the retired sitelinks searchbox

What WebSite SearchAction is

WebSite is the schema.org type for a site as a whole, usually declared once on the homepage. Its potentialAction property can hold a SearchAction that points at the site’s own search endpoint: target is a URL template containing a {search_term_string} placeholder, and query-input names that placeholder so a consumer knows what to substitute.

Google used this markup to render a sitelinks search box under a site’s result. It removed that element on 21 November 2024 and deleted the documentation for it, so the search box the markup produced is gone. The markup is still valid schema.org and still describes a real endpoint, which is why this rule reports at info severity rather than failing a page for its absence.

What squirrel checks

The rule reads every script[type="application/ld+json"] on the page and keeps the first WebSite node from the last script that contains one, matching either a plain @type or an @type array. Depending on what it finds it emits zero to two checks:

• website-schema: passes with the message WebSite schema found when the node exists. When it does not and the page is the homepage, meaning a URL path of / or empty, it reports at info status with the message No WebSite schema on homepage and the value Consider adding for sitelinks searchbox. On any other page with no WebSite node the rule emits nothing at all.
• website-search: reports at info status with the message WebSite has no potentialAction and the value Add SearchAction for sitelinks searchbox when potentialAction is absent, or with the message WebSite has no SearchAction when potentialAction exists but holds no node typed SearchAction.
• search-action: warns with the message SearchAction missing target URL when target is absent, warns with SearchAction missing query-input when target is present but query-input is not, and otherwise passes with SearchAction properly configured.

A SearchAction is recognised only when its @type is exactly the string SearchAction; an @type array containing it is not matched. Every property test is a truthiness test rather than a presence test, so a property present but empty, 0 or false counts as missing. The rule checks that target and query-input are truthy and stops there. It does not parse the target template or confirm that the placeholder in query-input matches the one in target.

A potentialAction written as a node reference, {"@id": "…#search"}, is read from the node with that @id anywhere in the page’s JSON-LD, in the same @graph or another script. A relative @id resolves against the page URL. A reference to a node the page does not have is not a SearchAction.

How to fix it

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "WebSite",
  "url": "https://example.com/",
  "potentialAction": {
    "@type": "SearchAction",
    "target": "https://example.com/search?q={search_term_string}",
    "query-input": "required name=search_term_string"
  }
}
</script>

This block goes in the homepage <head>.

JSON-LD wrapped in an @graph array, which is what Yoast, Rank Math and several other generators emit, is read the same as a top-level node: the rule flattens each script before looking for WebSite.

Enable / disable

Disable this rule

[rules]
disable = ["schema/website-search"]

Disable all Structured Data rules

[rules]
disable = ["schema/*"]

Enable only this rule

[rules]
enable = ["schema/website-search"]

enable replaces the default ["*"], so this is the only rule selected. Leave disable out: a rule that matches any disable pattern is skipped even when enable names it. Only an enabled setting under a rule’s own [rule_options] table overrides both lists.

Related rules

• schema/organization: the Organization node that usually sits alongside WebSite on the homepage.
• schema/json-ld-valid: whether the JSON-LD parses and passes schema.org validation at all.
• schema/breadcrumb: the other site-structure markup, checked on every non-homepage URL.
• schema/coverage-outlier: pages missing the structured data the rest of their page type carries.

Structured data findings ship in every audit next to the core SEO, performance and agent experience rules. See Fix SEO issues with AI for how an agent works through a report.

References

• Farewell, Sitelinks Search Box, Google Search Central Blog
• WebSite, schema.org
• SearchAction, schema.org

Check your site

Run squirrel audit https://example.com and open the Structured Data section of the report. The homepage result shows which part of the markup is missing. Signed-out local audits need no account and cost nothing. Signed-in audits spend credits.