Spam Policies
Spam policies are Google's rules about which techniques count as deceiving users or manipulating search systems. The definition from the documentation: spam is techniques used to deceive users or manipulate Search systems into featuring content prominently. Google detects policy-violating practices through automated systems and, as needed, human review that can result in a manual action. Sites may rank lower or not appear at all. Below: what exactly is described for each violation, what is acceptable alongside it and a few common claims the documentation does not support.
The basis is Google's spam policies for web search. What happens in a manual review is in the lesson on manual actions; what Google counts as helpful content is in the lesson on helpful content.
How Google finds spam
The overview of ranking systems says Google employs a range of spam detection systems, including SpamBrain, to deal with content and behaviors that violate the spam policies; they are constantly updated. Claims like "a neural network trained on billions of examples" or "the main pattern of the year" are not in the documentation. The policies cover common practices, but Google may act against any type of spam practices it detects, not only the listed ones.
What is forbidden and what is acceptable alongside it
| Practice | What the documentation says |
|---|---|
| Cloaking | presenting different content to users and search engines with the intent to manipulate and mislead. A paywall is not cloaking if Google can see the same full content as a person with access and you follow the Flexible Sampling guidance |
| Doorways | sites or pages created to rank for similar queries that lead users to intermediate pages that are less useful |
| Hidden text and links | white text on a white background, text behind an image or off-screen, zero font size. Tabs, accordions, sliders, tooltips and text for screen readers are acceptable |
| Keyword stuffing | filling a page with keywords or numbers: lists of phone numbers without value, blocks of cities and regions, unnatural repetition |
| Sneaky redirects | redirecting to show users and search engines different things. Legitimate: moving a site, consolidating pages, redirecting after login |
| User-generated spam | spam added by visitors through a channel intended for user content: comments, forums, file uploads |
On how to use JavaScript and images legitimately without falling under cloaking, see the lesson on JavaScript and SEO, and on legitimate redirects the lesson on redirects.
Link spam: what is a violation and what is not
The simplification "buying, selling and exchanging links" is imprecise. The violation is creating links to or from a site primarily to manipulate rankings. Examples: buying or selling links for ranking (money, goods, services, a product in exchange for a post with a link); excessive link exchanges and partner pages only for cross-linking; automated link creation; advertorials or native articles with links that pass ranking credit, and optimized links in guest posts and press releases; links in widgets, footers and templates distributed across many sites. At the same time Google understands that buying and selling links is a normal part of the economy of the web for advertising and sponsorship, and it is not a violation as long as the links are qualified with a rel="nofollow" or rel="sponsored" attribute. See the page on qualifying outbound links and the lesson on backlink auditing.
Content at scale, scraping and expired domains
- Scaled content abuse. This is when many pages are generated for the primary purpose of manipulating search rankings and not helping users, no matter how it's created. Examples: generative AI tools to produce many pages without adding value; scraping feeds or search results, including with automated synonym replacement or translation; stitching together content from other pages; creating several sites to hide the scale; pages whose text makes no sense but contains keywords. AI is one example, not a separate violation, as the post on content and AI says. See the lesson on AI in content planning;
- Scraping. Republishing other people's content without added value, even without citing the source, or with slight changes such as synonym swaps; sites that merely compile embedded material with no substantial value;
- Expired domain abuse. Buying a domain that has expired and repurposing it primarily to manipulate rankings with content that gives little to no value. Examples in the documentation: affiliate content on a former government agency site, casino content on a former school site. The point is hosting low-value content, not "link manipulation" in general;
- Hacked content — placed without permission through vulnerabilities: code or page injection, hidden links and cloaking, redirects. It is a separate category, not a violation by the site owner.
Site reputation abuse
Not every third-party content on an authoritative domain is a violation. The policy applies where third-party content is published on a host site mainly because of that host's already-established ranking signals, earned primarily from its first-party content, so that it ranks better than it could on its own. Explicitly not inconsistent with the policy: wire services and press release sites, syndicated news, user-generated content on forums and comments, editorial columns, native pieces whose purpose is to share content directly with readers, and affiliate links treated appropriately. An example of a violation: a page of sponsored payday loan reviews that a third party distributes to many sites.
Malicious practices and more
Malicious practices include malware, unwanted software and back button hijacking. "Misleading functionality" covers fake generators or sites that promise a tool but lead to ads. Machine-generated traffic is described separately: automated queries to Google, including scraping results for rank-checking, without permission violate the policies and the Terms of Service (see the lesson on manual actions and the topic of rank tracking). And among other things: given a significant volume of valid copyright removal requests, Google can demote other content from the site too.
What is not in the documentation
- Details of how SpamBrain works, such as "trained on billions of examples";
- "The main spam pattern of the year" and a ban on AI as such: the purpose and added value are judged;
- A ban on all buying and exchanging of links: links qualified with nofollow and sponsored are acceptable;
- A definition of "parasite SEO" as any content on an authoritative domain.
Checklist
- Users and search engines are shown the same thing.
- Paid and advertising links are qualified with nofollow or sponsored.
- There is no hidden text, keyword stuffing or doorways; tabs and accordions are acceptable.
- Pages at scale are made to help and add value.
- Third-party content is hosted on merit, not for the domain's authority.
- The public areas of the site are protected from user spam; the site is protected from hacking.
- Only legitimate redirects: a move, consolidation of pages, after login.
Mistakes
- Treating any purchase of a link as a violation and any AI text as spam.
- Hiding text "for SEO" and calling it design.
- Hosting other people's pages on an authoritative domain for its ranking signals.
- Buying an expired domain to host low-value content.
- Collecting positions with automated queries to search without permission.
What comes next
Next is the audit course: start with the audit framework, and for how the spam policies connect to concrete measures see the lesson on manual actions.