The Search Quality Rater Guidelines (SQRG) are a set of instructions used by human evaluators to assess the quality of Google's search results and the performance of ranking systems. While these ratings do not directly cause a website to rank higher or lower, they serve as a benchmark for how well Google’s automated systems deliver helpful, people-first content. For SEO practitioners, these guidelines provide a framework for self-assessing content quality based on the same standards Google uses to evaluate its own algorithmic improvements.
What are the Quality Rater Guidelines?
The Search Quality Rater Guidelines are a publicly available document, currently over 170 pages long, that provides a uniform definition of content quality. Google employs a global workforce of more than 10,000 search quality raters to conduct ongoing evaluations.
Raters use these instructions to determine if a search result is relevant and comes from a reliable source. The guidelines define both high-quality content, characterized by expertise and trustworthiness, and low-quality content, such as pages that seek to deceive users or spread hate.
Why Quality Rater Guidelines matter
Google uses these guidelines to verify that its search improvements actually help users. The practical outcomes of these evaluations include:
- System Benchmarking: Ratings help Google measure how well its ranking systems perform across different languages and locations.
- System Improvements: In 2023, Google's rigorous evaluation process led to more than 4,000 improvements to Search.
- Quality Control: Raters participated in 719,326 search quality tests in 2023 to ensure results meet a high bar for reliability.
- Content Strategy: By understanding E-A-T (Expertise, Authoritativeness, and Trustworthiness), marketers can create content that aligns with Google's definition of "helpful" information.
How Quality Rater Guidelines work
The evaluation process relies on human feedback to verify atmospheric changes in search performance. The mechanism involves several specific testing types:
- Side-by-Side Experiments: Raters are shown two sets of search results: one with a proposed change and one without. In 2023, Google performed 124,942 side-by-side experiments to determine which set of results users prefer.
- Needs Met Rating: Raters use a scale to judge how well a specific result satisfies the user’s intent. This scale was simplified in November 2023 to make ratings more consistent.
- Quality Assessment: Evaluators check for E-A-T. For example, a medical site is expected to feature content from doctors or medical institutions to be considered high-quality.
- Modern Content Evaluation: Guidelines now include specific instructions for rating newer formats, such as short-form video and discussion forums.
Best practices
- Focus on E-A-T: Ensure your content demonstrates expertise, authoritativeness, and trustworthiness. For instance, a doctor writing about health is rated higher than a non-expert.
- Prioritize helpfulness: Create content intended to help people rather than content designed solely to rank in search results.
- Research your reputation: Raters are instructed to look at external sources to verify a website's reputation. Maintain a positive presence on third-party review sites and forums.
- Simplify for "Needs Met": Answer the user's query immediately and accurately to ensure the result highly satisfies the search intent.
- Audit for different devices: Ensure content is useful across various formats, including mobile and video-heavy platforms.
Common mistakes
Mistake: Assuming a high rater score will immediately increase your rankings. Fix: Use the guidelines for long-term quality benchmarking, but realize ratings are used to improve general algorithms, not to rank individual sites.
Mistake: Neglecting reputation research. Fix: Regularly monitor what external sites say about your brand, as raters use these independent reviews to judge your trustworthiness.
Mistake: Focusing only on text-based SEO. Fix: Incorporate modern formats like short-form video, as the November 2023 update added specific guidance for these content types.
Mistake: Providing content that is misleading or deceptive. Fix: Ensure all claims are factually accurate, especially for YMYL (Your Money Your Life) topics like health and finance.
Examples
- High-Quality Scenario: A medical university publishes a study on a new treatment. Raters would likely give this a high E-A-T score because the source is a known authority with proven expertise.
- Low-Quality Scenario: A website with no listed authors or contact information gives financial advice that contradicts general consensus. Raters would identify this as untrustworthy.
- Needs Met Scenario: A user searches for "how to boil an egg," and the top result provides a clear, step-by-step instruction list at the top of the page. This would receive a high "Needs Met" rating.
FAQ
Do rater ratings directly affect my website's rank? No. Human ratings do not result in immediate rank changes for specific URLs. Instead, the data is aggregated to help engineers determine if an algorithm update is working correctly. If raters generally prefer the results after a change, Google is more likely to launch that update.
How often does Google update the guidelines? The guidelines are updated periodically to reflect changes in how people use the web. Major updates occurred in October 2021 and November 2023. These updates often add new examples or clarify language regarding website reputation and modern content formats like forum discussions.
What is the "Needs Met" scale? The "Needs Met" scale is a metric used by raters to evaluate how well a search result satisfies the intent of a query. In November 2023, Google simplified the definitions of this scale to help raters provide more consistent and accurate assessments of whether a page solves a user's problem.
Who are the search quality raters? They are an external group of over 10,000 people located all over the world. They represent various cultures and languages to ensure that search results are evaluated from a diverse range of perspectives.
