Google tells platforms to keep link-only generated pages out of Search by default

On October 26, 2026, Google Search Central updated its existing guide to preventing user-generated spam. The revised introduction explicitly includes hosting platforms and tools that publish shared pages, generated content or web apps, rather than focusing solely on comments and forums. A new 'Only index content meant for Search' section recommends using a noindex robots meta tag or X-Robots-Tag response header, or blocking crawling via robots.txt, by default for pages intended for direct sharing, collaboration or utility rather than public discovery. The monitoring section adds Search Console page-indexing and performance anomalies, and platform-level page-creation/indexing metrics. Google recorded the change in its official documentation changelog. This is advice about reducing abuse, not an announcement that Google will automatically penalise all generated pages.

Small SaaS products increasingly let customers publish forms, dashboards, AI-generated reports, landing pages and app previews under a shared domain. A link intended for one client can accidentally become an indexable public page, and bulk account abuse can create thousands of thin pages that threaten the platform's search reputation. The revised Google guidance gives product teams a concrete default decision: distinguish private/link-only utility pages from deliberately public content and make that distinction explicit in metadata and monitoring. Changing a sitewide robots.txt rule without understanding existing indexed URLs can cause different effects from noindex; builders need to test the right control.

Google's October 26 Search Central documentation update says it expanded guidance to platforms publishing shared pages, generated outputs and web apps. The guide now explicitly discusses automated page generation and platform misuse. This is a documentation revision and should not be described as a newly launched search-ranking system or proof of a broad enforcement action.

For pages made for direct sharing, collaboration or utility, Google's new 'Only index content meant for Search' section recommends a noindex directive or a crawl block by default. A product can still deliberately expose public pages to indexing when they have genuine search-discovery purpose. The advice is conditional on the page's purpose, not a ban on all AI-generated or user-generated content.

A robots meta noindex or X-Robots-Tag instructs crawlers that can fetch the page not to include it in results. A robots.txt disallow prevents compliant crawlers from fetching it, which means they may not see a page-level noindex directive. Google lists these as options, but operators should choose deliberately, especially when URLs are already known or indexed. Authentication and access control are separate requirements for confidential data; noindex does not make a page private.

The updated guide recommends verifying relevant domains or subdomains in Search Console, checking Page Indexing and Search performance for unexpected spikes, and watching creation/indexing metrics for abnormal growth or a few users producing disproportionate indexable URLs. Existing guidance also recommends account reputation signals, moderation, report flows, and untrusted-link rel attributes.

Define an explicit public/indexable state for each generated resource, with link-only as the safe default where appropriate. Emit X-Robots-Tag for non-HTML resources and meta noindex for HTML when crawlers are allowed to fetch; test response headers, redirects and canonical URLs. Separate publicly discoverable templates from private reports and previews. Add creation-rate alerts, moderation queues and Search Console review to the launch checklist. These are implementation implications inferred from Google's documented advice, not new mandatory product rules.