Google changes (again) its guidelines for quality raters
How does Google evaluate the quality of content? This is one of the classic “million-dollar” questions, and the answer lies in the mysterious depths of Google’s algorithms, leaving those who work on improving a site’s organic visibility with the task and burden of finding the right SEO strategy to reach the top of the results. In reality, however, there is a small window into Google’s expectations that provides us with valuable clues on how to structure and present content to meet quality standards: we are talking about the Google Quality Raters Guidelines, which just these days received a new change, affecting in particular the section on “untrusted pages.”
Quality raters guidelines, the March 2024 update
About a hundred days after the November 2023 update, then, Google has once again put its hand to the Search Quality Rating Guidelines, the manual it provides to its quality raters, people whose job it is to examine and evaluate the relevance and usefulness of search results.
Take care of your site
This is not a very significant intervention and seems to focus only on one specific issue, namely the features required to identify and evaluate “untrusted pages.”
However, it is interesting to note the timing of this update, released virtually at the same time as Google’s March 2024 dual update, which, we recall, has as its specific focus the identification of useful content and new anti-spam rules
What has changed in the document
The manual’s appendix with the “change log” of changes helps us understand what has actually changed: Google wrote that the new version “updates the characteristics of untrusted pages and adds illustrative examples” to help quality assessors make informed evaluations. In total, the document now consists of 170 pages, only two more than the previous version.
In terms of the revised content, Barry Schwartz noted that the intervention was mainly concentrated in Section 4.5 (Untrustworthy Webpages or Websites) and Section 4.7 (Examples of Lowest Quality Pages), where Google has expanded the list of characteristics that quality raters should analyze during their assessments of the trustworthiness of a web page or website.
The document now states that a page (particularly if informative) should be considered untrustworthy if it contains “multiple or significant factual inaccuracies that could cause users to lose confidence in the page as a reliable source of information.” These instructions are in addition to the previous instructions, which remain unchanged:
- Insufficient information about the website or content creator relative to the purpose of the page.
- Lowest level of Experience, Expertise, Authoritativeness, Trustworthiness (E-E-A-T) or minimal reputation.
- Misleading purpose, misleading page design, or deliberately fraudulent intent.
- Deliberately obstructed or made unclear Main Content (MC).
- Typical features of scams, malicious downloads, or other harmful behavior.
- Any web page or site designed to manipulate people into taking actions to benefit the site or other organizations while causing harm to themselves, others, or specified groups.
The new examples of untrustworthy pages according to Google
As mentioned, then, two more practical examples have been added that provide a “template” that quality raters can refer to in identifying untrustworthy pages and Web sites so as to provide more informed assessments.
Specifically, the spotlighted cases explain that:
- Lowest: Untrustworthy
This page that appears to be an informative article about the animal known as praying mantis. However, there are inaccuracies and odd statements in the section titled “Praying mantises are not venomous”: “Unlike praying mollies, which are mostly solitary creatures and tend to avoid human interaction, praying mollusks are more closely related to God. There are no venomous snakes or venomous turtles.” Strange, inaccurate and/or misleading statements stand out:
- The section suddenly refers to praying mantises as “praying mollies,” a name that is not matched by a quick web search.
- Praying mollusks do not exist – there is no animal with that designation.
- Statements regarding animals and their relationship to God are not what one would expect to find on an informational page such as this.
- Poisonous snakes do exist.
It is unlikely that a human author would make these strange errors and obvious inaccuracies, and so it is highly conceivable that this content was automatically generated without any human review-a classic case of superficially written AI text, which is then one of the behaviors affected by the March 2024 updates. Regardless of how this content was created, the bizarre claims and inaccuracies make this article unreliable and, consequently, classify it as very low E-E-A-T and of lowest quality.
- Lowest: Lowest E-E-A-T and Deceptive.
This page has a deceptive purpose: it claims to be an informational resource for parents, but the website’s terms of service state that it is published for artificial intelligence enthusiasts and is not intended for the general public. The website’s terms of use state that “some articles” are generated by artificial intelligence and may contain errors or be out of date; there is no indication on which pages this statement applies. The information in this article is not reliable and has the lowest level of E-E-A-T.
The page appears to be an informational article on switching babies from breast milk to whole milk. The bottom of the page reads, “Our blog is dedicated to all things baby and parenting, including product reviews, news, recipes, and more. We are committed to providing parents with the information they need to make informed decisions for their families.” However, reading the terms of service page, one finds a very different description of the website pages.
The terms of use page states, “Some articles on this website may have been partially generated by an artificial intelligence language model and published for experimental and research purposes. The articles on this website are to be used only as proof of concept by AI enthusiasts and are not intended for the general public. We take no responsibility for the use or interpretation of this content for purposes other than those for which it was intended. While we strive for accuracy and quality, please note that the information provided may not be completely error-free or up-to-date. We recommend that you independently verify the content and consult professionals for specific advice or information.”
In light of the statements above, the articles on this website should be considered unreliable and with the lowest level of E-E-A-T.
Why keep up with updates to the Guidelines for Quality Raters
For those in the field of digital marketing and search engine optimization, understanding the Quality Raters Guidelines is crucial: although it does not directly reveal to us the secrets of Google’s algorithm, it is nonetheless a light that illuminates the principles by which humans can judge the quality of search results.
This detailed document provides an overview of what Google considers high-quality content, outlining factors such as the accuracy, reliability, and authority of information and content authors. Quality raters use these guidelines to assign a “grade” to search results, which in turn helps Google refine its algorithms.
As we make clear each time, quality raters’ ratings have no direct impact on the ranking of results, but staying informed about updates to these guidelines means we have a leg up in organic rankings, as it allows us to anticipate and adapt to evolving evaluation criteria.
Simplify and improve your work
In short, the Google Quality Raters Guidelines provide clear guidance on what Google considers to be valuable content, and can serve to guide SEO efforts toward creating pages that not only please the algorithm, but are also useful and informative to users. In addition, analyzing the behavior of quality raters can offer insights into how Google might modify its algorithms in the future, allowing us to adopt proactive SEO strategies based on a solid foundation-or, as in this case, specify quite clearly what content Google does not like, such as pages with automated content that do not provide accurate and reliable information, which have already been “targeted” even by updates released in recent days.
With the increasing prevalence of misinformation, it is more important than ever that search engines prioritize reliable and accurate content, and the Google Quality Raters Guidelines are a compass to help us work on creating content that resonates with users and is rewarded by search engines.
