Cloaking

Cloaking means serving different content or URLs to Google than to people. Google treats it as spam. If you run a small site, avoid anything that could look like bait and switch.

By , founder of Porteur · Updated 14 September 2026 · Markdown

What counts as cloaking

Cloaking is intentional mismatch: Googlebot gets one thing, users get another. Examples: a text‑heavy page for Google, but an image or thin page for visitors. Or sending Google to /guide while users are redirected to /signup.

Google’s spam systems, including SpamBrain, target it. Spam updates roll out and can hit the whole site, not just one URL. A manual action can also be applied after human review.

What does not count as cloaking

  • A paywall that uses the paywalled content markup, and shows the same headline, intro and structured data to Google and users.
  • Responsive design that adapts layout to screen size while keeping the same main content.
  • Geotargeting or language choice that changes currency or wording, not the topic or intent of the page.
  • A or B tests that serve one of a few variants to all users and Google equally, are short‑lived, and do not hide content from users.
  • Server‑side rendering or hydration that results in the same content for Google and users.

How to check your site

  1. Compare what Google fetches

    Open Search Console’s URL Inspection for a page. View the crawled page HTML and screenshot. Compare it to what you see in a logged‑out browser.

  2. Test user agents

    Load the same URL with a normal browser and with a crawler user agent using a crawler or curl. The core text, links and canonical should match.

  3. Check redirects

    Use a redirect checker. Make sure Googlebot and users both get the same status code and final URL, for http to https, www to non‑www and mobile.

  4. Look in server logs

    Sample access logs. Confirm that Googlebot IPs and user IPs are served the same path and do not split to different content.

  5. Review changes and scripts

    Disable JavaScript, cookie walls and A or B test scripts. The core message and offers should still match what Google saw.

Fixes and safe setups

  • Remove any IP or user‑agent rules that swap content for Googlebot.
  • If you must pre‑render, serve the exact same HTML to users. Do not add extra text only for crawlers.
  • For paywalls, add the paywalled content structured data and avoid showing more to Google than to subscribers.
  • Keep A or B tests short and symmetric. Use 302 or 307 for temporary redirects in tests, not 301.
  • Unify redirects. For example, both users and Google go from / to https://yourproduct.com/ with the same 301 chain.
  • Document geo and device variants. Ensure the topic and claims are the same, and link alternates with hreflang where relevant.

A fixed page looks the same for Googlebot and a logged‑out user: same title, H1, body copy, links and canonical.

Penalties, manual actions and recovery

If Google detects cloaking, traffic can drop after a spam update or at any time. A manual action appears in Search Console with the affected pages.

  1. Confirm the issue

    Open the Manual actions report. Read the type and examples. Check the Security issues report too, which is separate.

  2. Fix fully

    Remove all mismatches and sneaky redirects. Make the live pages match what Googlebot will fetch.

  3. Request review

    From the Manual actions report, file a reconsideration request. Explain the root cause, the changes and the date deployed. Reviews take days to weeks.

  4. Expect timing

    Algorithmic hits from spam systems or core updates may need a later update to reassess the site as of 2026.

Questions

Sources

Check my site, free

Worried a template or redirect might look like cloaking, run your URL through the free check to compare what a crawler and a user will see in about thirty seconds from Porteur.

  • Free check, no card
  • Read-only, your own accounts
  • Readable by your agent

Read next