Google ranking factors are not a fixed public checklist. Google’s own guidance emphasizes helpful, reliable, people-first content, strong page experience, crawlability, mobile usability, links, and trust signals. The 2024 Google API documentation leak added useful clues about internal systems — including NavBoost, site-level signals, Chrome-related data references, and author/entity fields — but the leak did not prove exact ranking weights or confirm current active usage. In 2026, the safest SEO strategy focuses on content quality, topical authority, crawlability, internal links, high-quality backlinks, strong UX, brand and entity clarity, and measurable user satisfaction.
What Are Google Ranking Factors?
Google ranking factors are the signals and systems Google uses to determine which pages appear in search results and in what order. Google does not publish a complete list of active ranking signals or their exact weighting. What it has confirmed publicly: helpful content, crawlability, mobile usability, page experience, and links matter. Beyond official guidance, the SEO community draws on court testimony, internal documentation leaks, correlation studies, and practitioner experience to fill in the gaps.
The most significant development in understanding Google’s ranking systems came in May 2024, when Rand Fishkin published approximately 14,000 attributes from leaked Google API documentation — a dataset that revealed internal module names, signal names, and data types Google had never publicly disclosed. Google confirmed the documents were authentic but cautioned they were incomplete, outdated, and should not be used to draw direct conclusions about current Search ranking systems.
Google Ranking Factors at a Glance
| Factor Category | What It Means | Confidence Level | Practical SEO Action |
|---|---|---|---|
| Helpful content | Content satisfies user intent and adds original value | High — official guidance | Add first-hand experience, examples, sources, and useful depth |
| Crawlability and indexability | Google can access and index the page | High — official guidance | Fix robots.txt, noindex, canonical, sitemap, and rendering issues |
| Backlinks and authority | Other sites link to or reference your content | High — long-standing SEO evidence | Earn relevant editorial links and brand mentions |
| Internal links | Site structure helps Google discover and understand pages | High — practical SEO evidence | Link related pages using descriptive anchors |
| Page experience | Users can load, use, and navigate the page effectively | Medium-high — official guidance | Improve Core Web Vitals, mobile usability, and intrusive UX issues |
| Mobile usability | Mobile version is accessible and complete | High — official guidance | Keep mobile content equivalent to desktop |
| Brand / entity clarity | Google understands the business, author, and topic focus | Medium — practical SEO evidence | Use consistent profiles, schema, About pages, and authoritative mentions |
| User engagement systems | Click and user interaction systems may influence ranking systems | Medium — leak and court evidence | Improve titles, UX, satisfaction, and query-match quality |
| Freshness | Recency matters where the query requires updated information | Query-dependent | Update content when facts, tools, laws, prices, or platforms change |
The 2024 Google API Leak — What It Revealed (and What It Did Not)
In May 2024, an anonymous source leaked approximately 2,500 pages of Google’s internal API documentation to Rand Fishkin at SparkToro and SEO consultant Mike King. The documents revealed internal names for ranking modules and signal attributes. Google confirmed the documents were authentic but stated they were incomplete, outdated, and should not be used to draw direct conclusions about current Search ranking systems. (SparkToro: Rand Fishkin’s full analysis) (iPullRank: Mike King’s technical deep-dive)
The key context: the documents referenced many attributes and fields, but they did not reveal active usage, current weighting, or whether every documented attribute is a live ranking input. What follows covers what the leak suggested — not what it confirmed as current practice.
NavBoost: a click-related ranking system
The leaked documentation and antitrust testimony referenced NavBoost as a system connected to click and user interaction data. It appears related to click-based ranking adjustments — processing signals such as satisfied versus unsatisfied user sessions — though Google has not published a public formula for how these signals are weighted today. A page that attracts clicks but fails to satisfy users may perform poorly over time if Google’s systems interpret it as a weak result for that query. The practical implication: user satisfaction after the click matters as much as earning the click itself.
Chrome data references appeared in the leak
The leaked documentation referenced Chrome-related data fields, including what appeared to be site visit data associated with the Chrome browser. That does not prove a simple “Chrome traffic = ranking boost” mechanism. The safer interpretation: Chrome-related data may exist in Google’s internal systems, while its current role and specific weighting in Search ranking remain unclear. A field existing in documentation does not confirm it is actively used as a ranking factor or reveal how much it contributes.
Author and entity fields in the documentation
The leak referenced fields for SourceEntity and AuthorEntity, indicating some form of entity-level tracking in Google’s internal systems. Google’s public guidance encourages clear authorship and trust signals — but this should not be framed as a confirmed author-score ranking system. The existence of these fields suggests Google has structured ways to associate content with entities, not exactly how or whether they directly influence ranking for a given page. Google’s documentation notes that E-E-A-T is not a specific ranking factor, though its systems use a mix of signals to identify content with strong E-E-A-T characteristics. (Google: E-E-A-T documentation)
Site-level quality signals
The leaked documents reference a field called siteAuthority, which suggests Google maintains some form of site-level quality signal. This should not be equated with Moz Domain Authority, Ahrefs DR, or any public metric — those are third-party approximations, not direct Google signals. The practical takeaway: site-level reputation, topical consistency, and link quality likely matter in aggregate, but the exact scoring method is not publicly known and the field may not map to what practitioners typically mean by “domain authority.”
Official vs Leaked vs Observed: How to Read Ranking Factor Claims
| Evidence Type | What It Means | Example | How to Treat It |
|---|---|---|---|
| Official Google guidance | Google publicly recommends or confirms it | Helpful content, mobile-first indexing, page experience | Highest confidence — act on this directly |
| Court testimony / legal evidence | Systems discussed in legal proceedings | NavBoost references in antitrust trial | Useful context — not a complete formula |
| API leak / internal documentation | Internal fields or modules appeared in leaked docs | siteAuthority, click-related fields, author entities | Useful clues — not confirmed weights or active usage |
| SEO studies / correlation data | Patterns observed across ranking pages | Backlinks, content depth, branded search demand | Helpful for prioritization — not proof of causation |
| Practitioner experience | Audit and campaign observations across many sites | Internal links improving crawl discovery | Useful — but site- and query-dependent |
What Google Has Not Confirmed
Google has not published a complete list of active ranking factors, exact signal weights, or a live ranking formula. It has not confirmed that Chrome traffic, direct traffic, author entities, unlinked brand mentions, FAQPage schema, or third-party authority metrics operate as direct ranking factors. These elements can still support SEO indirectly — by improving discoverability, trust, relevance, brand recognition, and user satisfaction — but they should not be framed as confirmed ranking levers.
Google’s documentation also notes that E-E-A-T is not a specific ranking factor or a score Google can directly measure. Instead, Google uses a mix of signals to identify pages that demonstrate strong E-E-A-T characteristics, particularly for YMYL topics. (Google Search Central: Creating helpful content)
Content Quality Signals
Topical authority
Sites with strong topical coverage are often better positioned to rank because they can satisfy more related queries, support internal linking, and demonstrate depth around a subject. Google’s Quality Rater Guidelines describe “expertise” as a quality signal, and sites that cover a subject comprehensively through interconnected content typically outperform sites with isolated articles on the same topic. This is a consistent pattern across competitive SERPs — not a rule that overrides all other signals. The standard implementation is topic clusters: a pillar page linked to cluster articles that each cover a specific subtopic in depth. (Pillar Page Strategy: How to Build Topic Clusters That Rank) (Google: Search Quality Rater Guidelines)
Genuine helpfulness
Google’s helpful content guidance and later core updates emphasize helpful, reliable, people-first content and discourage content created mainly to attract search traffic. Google’s guidance emphasizes whether content provides original value: first-hand experience, genuine analysis, proprietary data, or insights not already available elsewhere. Content that summarizes what other pages already say, without adding a perspective or depth the reader cannot get by clicking the first result, is exactly what this system targets. (Google: Creating helpful, reliable, people-first content)
E-E-A-T: Experience, Expertise, Authoritativeness, Trustworthiness
Google added the second “E” for Experience to its quality framework in December 2022. The full framework — Experience, Expertise, Authoritativeness, Trustworthiness — describes qualities that high-quality pages demonstrate. Trust is the most foundational: a page that is inaccurate, misleading, or lacks named authorship fails on Trust before any other signal matters.
An important distinction: E-E-A-T is not a specific ranking factor or a score Google directly calculates. It is a quality framework used by human quality raters to evaluate content. It overlaps with trust, authorship, reputation, sourcing, and content quality signals that Google’s automated systems do assess. For YMYL topics — medical, financial, legal, safety — Google’s quality systems apply especially strict expectations because the stakes for users are higher. For the full E-E-A-T implementation approach, see the E-E-A-T in SEO: Complete Guide. (Google: E-E-A-T documentation)
Content freshness
Freshness matters most for queries where recency is part of the user’s intent — news, current events, product comparisons, software documentation. For evergreen queries (“how to tie a bowline knot”), freshness carries far less weight. Google has long used freshness-oriented systems for queries where recency matters, often discussed in SEO as “Query Deserves Freshness.” Updating content with substantive changes and updating the dateModified field in Article schema helps Google understand recency — but updating dates without substantive content changes does not help and can erode source credibility.
Link Signals
Editorial backlink quality
High-quality editorial backlinks from relevant, authoritative sites remain one of the strongest practical SEO signals. Links help discovery, authority, and trust — three things that contribute to ranking across most query types. Google does not publish exact link weights, and relevance and quality matter more than raw link count. A topically relevant link from a mid-authority industry publication often outperforms a topically irrelevant link from a high-authority general site. Google’s spam policies confirm it devalues links that were not editorially given, including paid links and link schemes. (Google Search Central: Link spam policies)
Link diversity in competitive SERPs
In competitive SERPs, pages with stronger and more relevant referring domains often have an advantage when content quality and intent match are comparable. Multiple ranking factor studies from Ahrefs, Semrush, and Moz have found that referring domain counts correlate consistently with high SERP positions — while correlation does not prove causation, the pattern holds across competitive industries. The practical emphasis should be on earning editorially given links from relevant, authoritative sources, not on raw volume.
Internal linking
Internal links help Google discover pages, understand topical relationships between content, and assess which pages are most important within a site. Important pages benefit from relevant internal links from hub pages, navigation, and closely related articles. Contextual anchor text — text that describes the destination page’s topic — is more useful than generic “click here” links. Internal linking can support crawl discovery, topical relevance signals, and page prominence, but these effects are site- and context-dependent. For a full internal linking approach, see the Internal Linking for SEO Guide.
Technical and Page Experience Signals
Crawlability and indexability
Before any ranking signal matters, Google must be able to crawl and index the page. Common technical issues that block ranking include: pages blocked in robots.txt, noindex tags applied incorrectly, canonical tags pointing away from the target page, JavaScript rendering failures, and pages missing from the sitemap. Fixing these issues is the highest-priority technical SEO work because they determine whether a page is eligible to appear in results at all. For a systematic technical audit covering all of these, see the Technical SEO Audit Checklist: 47 Items to Fix.
Core Web Vitals
Core Web Vitals are part of Google’s page experience signals. The three metrics — LCP (loading), INP (interactivity), and CLS (visual stability) — draw from real-user field data collected through the Chrome User Experience Report (CrUX). Lab data from tools such as PageSpeed Insights and Lighthouse is useful for diagnosis, but should not be treated as a direct replacement for field data in ranking assessments. Core Web Vitals matter most when performance is noticeably poor or when competing pages are otherwise similar in quality and authority. They are more likely a tiebreaker than a primary ranking driver. (Google web.dev: Core Web Vitals) (Google: Page experience documentation)
Mobile-first indexing
Google’s default since 2019. Mobile-first indexing does not mean Google only crawls mobile pages — it means the mobile version is the primary version used for indexing and ranking. If the mobile version has less content than the desktop version — hidden text, collapsed sections omitted from mobile rendering — that content may not contribute to rankings. The mobile and desktop versions should be content-equivalent. (Google: Mobile-first indexing documentation)
HTTPS
HTTPS has been a confirmed ranking signal since August 2014. It is now table stakes — browsers mark HTTP sites as “Not Secure,” which directly harms trust and user engagement. HTTP-only sites are extremely rare among competitive domains. HTTPS is largely a baseline technical and trust requirement now. Once implemented correctly, it is unlikely to be a major differentiator among competitive sites.
Click and engagement signals
The 2024 leaked documentation and antitrust testimony referenced NavBoost as a system connected to click and user interaction data. Based on available evidence, it appears related to click-based ranking adjustments — where pages that consistently satisfy users may strengthen for a query and pages that consistently fail users may weaken. A page that attracts clicks but fails to satisfy users may perform poorly over time if Google’s systems interpret it as a weak result. The practical implication: matching user intent precisely and delivering a satisfying experience after the click matters more than CTR alone.
Brand and Entity Signals
Branded search volume
When users search for a brand by name, it can indicate real-world recognition. Rising branded search volume is associated with increased organic visibility in correlation studies by Semrush and Ahrefs. The mechanism is indirect rather than a confirmed direct ranking input: strong brands earn more links, mentions, clicks, and trust over time — which feed into established signals. Building a brand through PR, advertising, and content distribution generates branded search as a secondary SEO benefit.
Repeat visits and direct traffic
Repeat visits, branded search, newsletter traffic, and direct visits can indicate real audience demand, but these should not be framed as confirmed direct ranking levers. The 2024 leak referenced Chrome-related data fields that may include site visit patterns, but the exact role in ranking is not confirmed. The practical SEO value is that strong brands earn more links, mentions, searches, clicks, and trust over time — all of which do affect rankings through established signals.
Unlinked brand mentions
Google’s entity understanding systems may associate brand mentions in text even without a hyperlink. Unlinked brand mentions can support brand visibility and may help users and systems connect your brand with a topic — but they should not be presented as a confirmed ranking factor or guaranteed Knowledge Graph input. PR coverage, podcast appearances, and co-citations with authoritative brands build brand association over time.
Ranking Factor Priority Matrix
| Priority | Factor | Why It Matters |
|---|---|---|
| Critical | Crawlability and indexability | Pages cannot rank if Google cannot access or index them |
| Critical | Search intent match | Content must satisfy the query better than competing pages |
| High | Helpful content and topical depth | Drives relevance and user satisfaction at the page level |
| High | Internal linking and site architecture | Helps discovery, topical context, and PageRank distribution |
| High | Editorial backlinks | Supports authority and trust at the page and site level |
| Medium-high | Page experience and Core Web Vitals | Supports usability, especially when performance is noticeably poor |
| Medium | Brand and entity clarity | Helps recognition, trust, and source confidence across AI and traditional search |
| Query-dependent | Freshness | Matters when users expect current information — less so for evergreen content |
Ranking Factors and AI Search Visibility
Google’s AI Overviews and AI Mode are connected to Google’s broader Search systems, but AI-generated answers place additional practical emphasis on structured answers, source credibility, entity clarity, and extractable content. Pages that rank well in traditional Search may be more likely to be cited in AI-generated answers, but the relationship varies by query, source availability, and AI interface. For the AI-specific optimization layer, see the Generative Engine Optimization guide and the Google AI Mode SEO guide. For visibility in ChatGPT Search and other AI answer engines, crawler access and structured content matter differently than in Google — see the ChatGPT Search SEO guide.
What’s Declining in 2026
Exact-match anchor text manipulation
Real-time Penguin filtering, rolled into the core algorithm in 2016, evaluates anchor text distribution continuously. A link profile with an unnaturally high proportion of exact-match anchor text is a manipulation signal. Natural link profiles have mostly branded and generic anchors with some partial-match anchors. Deliberate exact-match anchor text link-building at scale is risky and may trigger spam systems or reduce trust in the link profile. Google’s spam policies address anchor text manipulation directly. (Google: Spam policies — link spam)
Keyword density
Google’s semantic understanding systems — built on BERT (2019), MUM (2021), and subsequent neural language models — evaluate pages for topical relevance based on meaning, not keyword repetition. Repeating a phrase 15 times does not improve relevance signals. Writing naturally about a topic, covering related concepts and entities, does. Keyword density as a measured optimization target is obsolete.
Thin or scaled low-quality content
Google’s March 2024 core update and subsequent updates targeted large-scale content generation producing pages with no original value. Many sites that relied heavily on thin, scaled, or unoriginal content saw volatility after these updates — particularly where pages lacked first-hand experience, originality, or clear user value. Google’s helpful content guidance specifically warns against producing lots of content mainly to attract search traffic, using extensive automation to generate content at scale without genuine expertise, and updating dates without substantive content changes. (Google: Creating helpful content)
Frequently Asked Questions
How many ranking factors does Google use?
Google does not publish a complete list of active ranking factors or their weights. Google’s guidance describes automated ranking systems that use many different signals, but the exact list and weighting are not public. The 2024 leaked API documentation referenced approximately 14,000 internal attributes — many of which are sub-signals within larger systems, not independent ranking factors. The number itself is less useful than understanding which signals carry the most practical weight for a given query type and competitive landscape.
Is the 2024 Google leak reliable?
Google confirmed the documents were authentic but stated they were incomplete, outdated, and should not be used to draw direct conclusions about current ranking systems. The SEO community’s assessment, informed partly by Google’s own antitrust trial testimony referencing the same internal system names, is that the documents are genuine. How closely current ranking systems match the leaked version is unknown — the leak captured a point in time, not current architecture. Treat it as useful contextual evidence, not a live ranking manual.
What is the single most important ranking factor?
No single factor dominates across all queries. For informational queries, content depth and topical authority tend to matter most. For competitive commercial queries, backlink quality and diversity tend to be the primary differentiator when content quality is comparable. For local queries, Google Business Profile completeness and proximity matter most. The most important factor is usually the one that differs most between your page and the page ranking above it — which is why a competitive content gap analysis is more useful than an abstract ranking factor list.
How does AI-generated content affect rankings?
Google’s position is that it evaluates content quality regardless of how it was produced. AI-assisted content that is accurate, original, demonstrates genuine expertise, and serves user needs well can rank. Thin, repetitive, or unoriginal AI-generated content — produced at scale without genuine human expertise — is what Google’s 2024 updates targeted. The evaluation question is not “was this written by AI?” but “is this genuinely helpful, accurate, and trustworthy for the user asking this query?”
How do I improve E-E-A-T practically?
Five concrete steps: (1) add named author bios with credentials and links to LinkedIn or external profiles; (2) cite primary sources — studies, official documentation, named experts — inline in your content; (3) add a clear About page describing the organization’s expertise and mission; (4) earn coverage and mentions from recognized publications in your niche; (5) show first-hand experience — product reviews that show you have used the product, case studies from your own work, data from your own research. For YMYL topics, adding named expert reviewers and citing peer-reviewed research or primary guidelines carries additional weight.
Sources and References
- SparkToro: Google API leak analysis, Rand Fishkin (May 2024)
- iPullRank: Google algorithm leak deep-dive, Mike King (May 2024)
- Google Search Central: Creating helpful, reliable, people-first content
- Google: Search Quality Rater Guidelines
- Google Search Central: E-E-A-T documentation
- Google web.dev: Core Web Vitals
- Google Search Central: Page experience documentation
- Google Search Central: Mobile-first indexing
- Google Search Central: Spam policies — link spam
Google ranking factors are not a fixed public checklist. Google’s own guidance emphasizes helpful, reliable, people-first content, strong page experience, crawlability, mobile usability, links,...
