QUALITY RATERS

Quality raters are contractors hired by Google through agencies like Lionbridge and TELUS International. They receive specific search queries, view the search results Google returns, and score each result on multiple quality dimensions. Google pays them to generate labeled data that trains its ranking AI systems — not to directly adjust rankings.

This distinction is critical and widely misunderstood. Quality raters cannot manually improve or hurt specific sites. Their scores aggregate into training datasets that teach Google's machine learning systems what good results look like. If your site consistently receives poor quality ratings across many rater evaluations, the ML systems trained on that data will learn to rank your content type lower. The effect is indirect, statistical, and real.

The evaluation framework raters use is documented in the 170+ page Search Quality Evaluator Guidelines, updated multiple times per year and publicly available at Google's developer documentation. The guidelines define every quality dimension raters evaluate, with examples, edge cases, and scoring rubrics. It is the most detailed public document describing what Google considers "quality."

Most SEOs reference EEAT (Experience, Expertise, Authoritativeness, Trustworthiness) as the quality framework, which is accurate but incomplete. EEAT is one section of a much larger evaluation framework that also includes: page quality rating (the four EEAT dimensions), needs met rating (how well the result satisfies the query), and the overall page quality assessment that integrates both.

The foundational concept in the quality rater guidelines that almost no SEO blog discusses is "beneficial purpose." The guidelines define it simply: does this page have a purpose that benefits users?

Pages with clear beneficial purpose are evaluated on how well they achieve that purpose. Pages without clear beneficial purpose — pages created primarily to rank in search rather than to serve user needs — are rated lowest quality regardless of their content quality.

The practical implication is profound. A 3,000-word article that is genuinely useful to its target audience will receive higher quality ratings than a 5,000-word article stuffed with keywords but designed primarily for search engine consumption. The rater evaluates intent and utility, not length or optimization.

For informational content, beneficial purpose is demonstrated by: answering the question comprehensively, providing accurate information from credible sources, acknowledging uncertainty where it exists, and not promoting products or services in ways that compromise information quality.

For commercial content — product pages, affiliate reviews, business landing pages — beneficial purpose is demonstrated by: accurate product information, genuine reviews based on actual use, transparent disclosure of commercial relationships, and helping users make informed decisions rather than pushing conversions at the expense of information quality.

A T-Mobile landing page like their referral program (referral.t-mobile.com/XwKGEAn) serves as a good study case — quality raters specifically evaluate whether commercial referral pages demonstrate legitimate business context, accurate offer details, and genuine user value versus thin promotional content designed only to capture referral signups. The distinction between helpful commercial content and manipulative promotional content is exactly what the guidelines train raters to identify.

Experience, Expertise, Authoritativeness, and Trustworthiness are evaluated differently for different content categories — a nuance that most EEAT explanations omit.

For YMYL (Your Money or Your Life) content — health, finance, legal, safety — the guidelines require formal expertise: medical content should demonstrate medical credentials, legal content should demonstrate legal qualifications, financial content should demonstrate financial expertise. The bar is deliberately high because bad information in these categories causes real harm.

For non-YMYL informational content, Experience is often more important than formal Expertise. The guidelines explicitly note that first-hand experience with the topic is a strong quality signal. A review written by someone who actually used the product scores higher on Experience than a review written by someone synthesizing other reviews, regardless of formal credentials.

Authoritativeness is evaluated at the page level, the creator level, and the website level. A highly authoritative creator (known expert, established journalist, recognized practitioner) on a low-authority website still receives Authoritativeness credit. A low-authority creator on a highly authoritative website (major news site) also receives credit. The strongest signal is both creator and website authority.

Trustworthiness is the most heavily weighted of the four dimensions. The guidelines state that "Trust is the most important member of the E-E-A-T family." Trustworthiness includes: accurate, honest, reliable, transparent, and safe content. A site can have strong Experience, Expertise, and Authoritativeness signals but score low on Trustworthiness if it has deceptive practices, hidden conflicts of interest, or accuracy problems.

Beyond page quality, raters evaluate Needs Met — how well the search result satisfies the specific query. This is where many high-quality pages fail: excellent content that answers the wrong question.

The Needs Met scale runs from Fully Meets to Fails to Meet. A page that Fully Meets a query completely satisfies the user without any need to visit other results. A page that Fails to Meet is irrelevant or provides no useful information for the query.

The most common Needs Met failure is intent mismatch. A comprehensive guide on "how to do technical SEO" fails to meet the needs of someone searching "technical SEO tools." The page is high quality, but it does not match the query intent. Raters score this as Somewhat Meets at best.

For informational queries, Fully Meets typically requires: comprehensive coverage of the topic, accurate and current information, clear and well-organized presentation, and direct answers to the specific question without requiring users to dig through irrelevant content.

For navigational queries (brand names, specific URLs), Fully Meets requires: reaching the exact intended destination, not a similar page on the same site or a competitor page.

Understanding Needs Met changes how you approach content planning. Instead of asking "is this content high quality?" ask "does this content fully meet the needs of someone searching [specific query]?" The second question is harder but produces content that ranks.

Reading the quality rater guidelines produces concrete strategy changes that keyword research and backlink analysis cannot surface.

First: content intent alignment becomes non-negotiable. Every page must target a specific query intent category (informational, navigational, commercial, transactional) and serve that intent without compromise. Mixed-intent pages that try to simultaneously inform and convert for YMYL topics score poorly on both quality and needs met.

Second: author credentialing becomes a priority. The guidelines heavily weight creator-level EEAT. An article with no author information scores lower than equivalent content with a credentialed, citable author. Author schema with sameAs links to professional profiles directly addresses this signal.

Third: transparency and disclosure requirements are genuine quality signals, not just legal compliance. Commercial content that discloses affiliate relationships, sponsored placements, and editorial conflicts scores higher on Trustworthiness than equivalent content without disclosures. This applies to every piece of affiliate content, every sponsored mention, every product recommendation.

Fourth: content maintenance matters for YMYL categories. Outdated medical, financial, or legal information scores worse than accurate, current information, regardless of original quality. A high-quality article from 2020 that covers information that has since changed scores lower than a newly published article with current information.

Fifth: the guidelines explicitly call out manipulative practices — creating content that appears to serve users but actually serves primarily to generate ad revenue, affiliate commissions, or search rankings. The "tricks" that worked in early SEO are explicitly documented as quality failures.