The Search Quality Rater Guidelines are a long manual Google writes for the human raters who evaluate its search results. They explain how to judge whether a page is high quality and whether a result actually meets a searcher's need.
Raters don't decide rankings, and the guidelines aren't a ranking formula. But they're the most detailed public description of what Google means by "quality," which is why SEOs, content teams, and now GEO practitioners read them closely.
What are the Search Quality Rater Guidelines?
Google works with external raters to measure the quality of its results. Its How Search Works page on testing explains that raters assess how well pages fulfill search requests, and that Google publishes the guidelines so ratings stay consistent. A Google overview deck from November 2023 put the rater pool at about 16,000 people across regions, working in more than 80 languages.
The full guidelines document is a public PDF. As of October 2026, the version Google serves at that link is dated September 11, 2025, and runs to more than 180 pages, with a change log at the end.
How raters' work is used (and how it isn't)
This is where the most confusion lives, so it's worth being precise.
- Ratings don't directly change rankings. Google states plainly that rater scores do not directly impact ranking.
- Ratings help test changes. In side-by-side experiments, raters compare results with and without a proposed change. That feedback helps Google decide whether a change improves search. Google's testing page lists 719,326 search quality tests in 2023.
- The guidelines aren't a ranking manual. Google's generative AI guidance says so directly, and in May 2026 Google's John Mueller repeated that the guidelines are worth reading but aren't a guide to rankings.
The fair reading: the guidelines describe the outcome Google wants its systems to produce. They don't describe the signals those systems use to get there.
What's inside the guidelines
Page Quality rating
Raters score pages on a scale from Lowest to Low, Medium, High, and Highest, with half steps in between. They consider the page's purpose, the quality of its main content, the reputation of the site and creator, and E-E-A-T: experience, expertise, authoritativeness, and trust. Trust is treated as the most important part.
Needs Met rating
Separately, raters judge how well a result satisfies the query, from Fails to Meet through Slightly, Moderately, and Highly Meets to Fully Meets. Fully Meets is reserved for queries with one clear, specific answer. This is where search intent really shows up.
YMYL topics
The guidelines define YMYL ("Your Money or Your Life") topics as those where poor information could seriously affect health, financial stability, safety, or society. Pages on these topics are held to higher standards.
Low quality, spam, and AI
Recent versions include detailed sections on spam and low-effort content. Google points site owners to two in particular: section 4.6.5 on scaled content abuse and section 4.6.6 on main content created with little effort, originality, or added value. The guidelines also define generative AI, noting it can be helpful for content creation but can also be misused. Search Engine Land reported that the September 2025 update added examples involving AI Overviews and revised the YMYL definitions.
Why the guidelines matter
For SEO, the guidelines are the closest thing to Google showing its work. When a core update moves your rankings, the guidelines help you ask better questions: is the main content substantial, is the creator clear, would a careful reader trust this page?
For GEO, the same qualities tend to matter. AI answer engines need sources they can quote accurately and safely. Pages with clear authorship, cited facts, and substantial original content are easier for those systems to rely on, whether or not they use anything like Google's rater framework.
How to use the guidelines in practice
- Read sections 2 and 3 first. They cover page purpose, main content, YMYL, and E-E-A-T, the foundation for everything else.
- Rate your own pages. Pick your top 10 traffic pages and score them honestly using the Page Quality criteria. Have someone outside the team do it too.
- Check the "who" on every page. Raters look for who created the content and who runs the site. Add real bylines, author pages, and a clear about page.
- Look for low-effort patterns. Compare your pages with sections 4.6.5 and 4.6.6. Mass-produced, lightly reworded, or thin content is the clearest risk.
- Pair it with Google's self-assessment questions. Google's helpful content guide is the shorter, site-owner-facing version of the same thinking.
Common misconceptions
- "A rater can tank my site." No. Individual ratings don't change how a specific site ranks. Manual actions are a separate process, applied when a Google reviewer finds a spam policy violation.
- "E-E-A-T is a score Google calculates." Google says E-E-A-T isn't itself a ranking factor. Its systems use a mix of signals that align with it.
- "Third-party summaries are enough." Plenty of articles claim to describe "new" guideline versions. Check the official PDF and its change log before acting on any summary.
How Butter approaches the rater guidelines
At Butter, we use the rater guidelines as a quality lens, not a checklist to game. When we make on-page recommendations, we're often asking rater-style questions: is the purpose clear, is the main content substantial, is the author visible, and would a careful reader trust this on a YMYL topic? Our schema markup guidance helps make authorship and organization details explicit, and our monthly AI prompt testing shows whether that content is being cited in AI answers. We don't treat the guidelines as a ranking promise, and we tell clients that directly.
Frequently asked questions
Do quality raters affect my rankings?
Not directly. Google says rater scores don't directly impact ranking. Raters help Google evaluate and test changes to its systems, which then affect all sites at scale.
Where can I read the Search Quality Rater Guidelines?
Google publishes the full guidelines as a free PDF on its raterhub domain. As of October 2026, the version served there is dated September 11, 2025, and includes a change log at the end.
Are the rater guidelines the same as Google's ranking algorithm?
No. The guidelines describe what high-quality results look like to a human evaluator. Google's algorithms use many automated signals to try to reach similar outcomes, but the guidelines don't list those signals.