Penalty risk checker
Does your site show practices search engines punish?
What the penalty risk check covers23 static checks across Google, Bing and AI engines, each with its policy.
- Cloaking: crawlers, browsers and visitors from Google compared
- Sneaky redirects, hidden text and text written to steer AI answers
- Thin pages at scale, near-identical page families, generated text leftovers
- Structured data, link labels and AI snippet directives
- Each finding with its policy, documented or inferred
Penalty risk levels, explained
- None found
- No static check matched. We read your homepage, sitemap and sampled pages as a browser, as Googlebot, Bingbot and AI crawlers, and saw nothing a written spam policy names. A good sign, not a clearance.
- Low
- A lead worth a look: a heuristic matched, or a risk we infer for AI engines, which publish no spam rules. Plenty of honest sites show one, such as a plan page template reused three times.
- Elevated
- A check backed by a written Google or Bing policy failed, for example a rating of your own company in structured data or affiliate links without rel sponsored. Fix these before the next spam update.
- High
- A practice with no innocent reading was seen, such as crawlers and visitors getting different pages, or several elevated families at once. Treat it as urgent and open Manual actions in Search Console.
What each family means and how to fix it
Crawlers and visitors get different pages
Googlebot or an AI crawler receives other text, another title or other outbound links than a browser, or a visitor arriving from Google is sent to another site. Usually a plugin, a CDN rule or middleware reading the user agent or referrer.
Fix: Serve the same main content and links to every user agent and referrer, and remove those conditions from middleware, edge rules and redirects.
Pages at scale with little value
Hundreds of sitemap URLs holding a locked preview, a few lines of text, or one template with a city, country or keyword swapped. Google calls it scaled content abuse or doorways, whoever wrote the pages.
Fix: Merge near-identical pages into one strong page with 301 redirects, noindex locked previews and thin pages, and stop publishing in bulk.
Hidden text and text aimed at AI answers
Paragraphs pushed off-screen, shrunk to zero or coloured like the background, or instructions asking assistants to keep the brand in mind as a favoured source, in hidden markup, llms.txt or a Summarize with AI link.
Fix: Delete text visitors cannot see unless it is an accessible label or skip link, and keep llms.txt and Markdown copies free of instructions to assistants.
Structured data that does not match the page
A star rating in JSON-LD shown nowhere on the page, a review of your own company on your own site, or FAQ and HowTo markup whose questions and steps are not visible.
Fix: Mark up only what the page shows, drop self-serving ratings, and remove FAQ markup rather than hiding questions to match it.
Generated leftovers, borrowed names and unlabelled links
Unrendered markdown (asterisks around a bold word), template placeholders or translation keys in the page text; another company's brand leading your domain and titles; paid links without rel sponsored; a CMS default of nosnippet or noarchive.
Fix: Review generated text before it ships, lead with your own name, qualify paid and user links, and remove snippet directives you never chose.
Middleware that cloaks, and the same middleware fixed
The rewrite shows crawlers a page visitors never see. Remove the condition, then compare both views with URL Inspection in Search Console.
// middleware.ts
export function middleware(req) {
const ua = req.headers.get("user-agent") ?? "";
if (/Googlebot|bingbot|GPTBot/i.test(ua))
return NextResponse.rewrite(new URL("/seo-landing", req.url));
}// middleware.ts
export function middleware(req) {
// Same response for every user agent and referrer.
return NextResponse.next();
}Questions about penalty risk
- Is this a Google penalty checker?
- It looks for the practices Google's spam policies name, not for the penalty itself. A manual action shows only in Search Console, and algorithmic demotions are never announced. Use this result for causes and Search Console for the verdict.
- Why only one free check every 30 days?
- A penalty risk result is sensitive, so it goes only to the person who asked and is never cached, shared or given a public link. The limit keeps it from being run against other people's sites. Verified owners get it daily.
- Does a clean result mean my site is safe?
- No. Our requests come from our own servers, so cloaking by IP address, timed pop-ups and scripts that spare crawlers stay invisible, and a sudden burst of new pages needs daily history to spot.
- What should I do after a drop during a spam update?
- Fix the families marked high or elevated first and stop adding pages. Recovery usually shows at the next spam or core update, weeks later; requesting indexing again does not speed it up.