Massive Google Search Algorithm Leak: What Publishers Should Know

You may have heard about the Google Search Algorithm Documentation leak in early 2024, revealing 14,014 ranking attributes. These attributes show that Google may not been entirely honest about some search ranking factors.

This secrecy isn't new, but this leak is significant for publishers relying on search traffic.

Leak Summary

On March 13, 2024, an automated bot named yoshi-code-bot released Google’s internal Content API Warehouse on GitHub

Rand Fishkin, SparkToro co-founder, and Michael King, iPullRank CEO, analyzed these documents extensively, detailing their discovery and analysis process. 

You can read Fishkin's and King's posts for more details.

What the Leak Reveals

Over 14K Ranking Attributes

As of March 2024, Google’s ranking algorithm consists of 2,596 modules and 14,014 attributes. While the exact weighting of these features is unspecified, their existence provides insights into Google’s ranking considerations. 

We don't know which attributes are prioritized, but we know they influence search results.

Site Authority and Brand Matter

Google has previously claimed that site authority doesn't affect page rankings, implying new websites and bloggers can compete equally with larger brands. 

However, the leak reveals an attribute called "siteAuthority," indicating that Google does consider overall domain authority. This suggests that websites with higher domain authority or bigger brands have a better chance of ranking than newer, smaller sites.

Re-ranking Functions and Attributes

Twiddlers are re-ranking algorithms that run between major updates, affecting SERP rankings. Your content can be promoted or demoted based on:

  1. Click Logs – More 'good' clicks mean a higher position.
  2. Freshness – Newer content ranks higher.
  3. Anchor Mismatch – Links that don't match their target site are demoted.
  4. SERP Demotion – Signals from the SERP suggesting user dissatisfaction lead to demotion.
  5. Location Demotions – "Global" and "super global" pages may be demoted, indicating Google's attempt to rank pages based on location.
  6. Porn Demotions – Obvious demotions for adult content.

Links are Crucial

Link diversity and relevance are key, and PageRank remains a significant factor. PageRank for a website’s homepage is considered for every document. 

This means links are important. But we just don't know precisely how these features are weighted.

Page Title Can Make or Break

The page title should closely match search terms to rank higher. Google uses a feature called titlematchScore to measure how well a page title matches a query.

EEAT Score May Not Be Real

While Google tracks authors across the web and within a website, there is no attribute or module specifically highlighting how Google measures experience, expertise, authoritativeness, and trustworthiness (EEAT). It appears that only the author and author bio are important."

"I’m a bit worried that E-E-A-T is 80% propaganda, 20% substance," said Fishkin. 

Other Interesting Findings

  1. Chrome Data for Ranking: A module called ChromeInTotal indicates Google uses data from its Chrome browser for ranking.
  2. Navigation and Exact Match Domains: Poor navigation and exact match domains can lower your rankings.
  3. Page Versions: Google stores the last 20 versions of your web pages. To have a "clean slate," you may need to update your page more than 20 times, though the extent of changes needed is unclear.
  4. Font Size and Text Weight: Google tracks font size for links and text weight. Larger links are positive, and bold text is read differently, which also improves accessibility.
  5. Keyword Stuffing Score: Google tracks keyword stuffing on websites. Keyword stuffing is still bad.
  6. Video Site Classification: If videos are on more than 50% of a site's pages, it is classified as a video site. It’s unclear if these videos need to be indexed, in a specific area of the post, or natively uploaded.
  7. YMYL Content: Your Money, Your Life (YMYL) content, including medical, financial, and health/safety topics, has a separate scoring system.
  8. AI Overviews: AI content and use of AI tool are not mentioned in the document.
  9. Internal Links: There is no clear mention of internal links as an attribute.
  10. Whitelists for Specific Topics: Three topics have "whitelists" requiring approval to be shared: travel, Covid, and politics. It’s unclear if these are for general SERPs, Google's travel section, or related widgets. 
  11. Author Identification: Google can identify authors and treat them as entities, meaning building an online influence as an author may benefit rankings.

SEO Checklist for Publishers Post-Leak

1. Diversify Backlinks

Publishers already value backlinking and have seen its positive impact on SERP rankings. This leak confirms the importance of a solid backlinking strategy.

  • Build high-quality backlinks to your pages.
  • Ensure backlinks come from a variety of relevant domains.
  • Make this an ongoing practice.
  • Don’t worry about link exchanges with sites you've linked to; no module suggests Google tracks this practice.
  • Focus on avoiding poor-quality links from new websites with low domain authority and little content.

2. Match Page Titles with Keywords

Improving page titles for SEO is another well-known practice among publishers to boost rankings.

  • Keywords remain crucial.
  • Avoid keyword stuffing, but use clear titles that include relevant keywords.

3. Increase Domain Authority

Increasing domain authority can be challenging, especially for small to medium-sized publishers.

  • High domain authority is essential for ranking for certain keywords.
  • Boost it by publishing more content, obtaining high-quality backlinks, and promoting your pages on social platforms.

4. Update your Content NOW

The leak confirmed at least 3 date-based attributes for measuring freshness: bylineDate, syntacticDate, and semanticDate

  • Freshness significantly impacts page rankings.
  • Update old content by refreshing page titles, images, data points, and other aspects to give it a renewed appearance.

5. Continue using AI Tools to Improve Your Content

Publishers need not hesitate to use AI tools.

  • As of March 2024, there's no mention of the AI factor.
  • Feel free to use your preferred AI tools to refine your content for your audience.

How Admiral Empowers Publishers to Enhance Search Performance

Pagespeed, visitor experience, and ranking are critical aspects of the online publishing business. Admiral pursues a high bar for performance, and helps publishers in the following ways:

  1. One tag is better than five: Admiral's one-tag, horizontal visitor engagement solution can eliminate the need for multiple tags and vendors to enhance all visitor touchpoints. Rather than separate solutions for subscriptions, adblock recovery, newsletter signups, donations, registrations, consent, social media, etc, publishers can keep the code brief and quick with Admiral's one-tag approach.

  2. Page speed commitment: Admiral vigilantly monitors tag performance and impact. Admiral has engaged multiple third-party testing specialists to measure the impact on Core Web Vitals measures for customers using the Admiral tag, and written up white papers showing the detailed results.

  3. Growing Visitor Relationship and Retention: Admiral's Visitor Relationship Management (VRM) solution helps publishers grow retention and engagement with their visitors. Admiral VRM drives more social follows, newsletter signups, donations, subscriptions, or sharing of first-party data. This results in more loyal and engaged visitors over time, more pageviews, and additional signals of site value to Google's algorithm.

  4. Sharing Best Practices: Admiral maintains a focus on search ranking factors and news, sharing insights and helpful tips with our customers, such as:

Want free tips to help grow your business?

Subscribe to Admiral's Log; a digest of ideas for driving publisher revenues and growing visitor engagement.

Get a Free Account Now with Revenue Analytics Dashboard

Get Admiral Free