How Does Google Rank Pages?
Google uses over 200 ranking signals. You cannot optimise all of them, but understanding the most impactful ones tells you exactly where to focus your SEO investment for maximum return.
Google ranks pages by evaluating three broad areas: relevance (does this page match what the user searched for?), authority (do other trusted sites link to it?) and experience (is it fast, mobile-friendly and trustworthy?). The algorithm uses over 200 signals to score each factor, refined continuously by AI systems including RankBrain and BERT.
The Core Ranking Factors, Ranked by Impact
| Ranking Factor | Impact Level | What Google Is Assessing |
|---|---|---|
| Backlinks (quality and quantity) | Very High | Do authoritative, relevant sites link to this page as a trusted source? |
| Content relevance and quality | Very High | Does this page genuinely answer the search query better than alternatives? |
| Search intent match | Very High | Does the content format and purpose match what the searcher expects to find? |
| E-E-A-T signals | High | Is the author credible? Is the site trustworthy? Is there demonstrated experience? |
| Core Web Vitals / Page Speed | High | Is the page fast, stable and responsive on mobile? |
| On-page optimisation | High | Does the title, H1, meta and content structure reflect the target keyword? |
| Internal linking structure | Medium-High | Do other pages on the site link to this page, distributing authority? |
| User engagement signals | Medium | Do users click the result, stay on the page and not immediately return to Google? |
| Domain authority / history | Medium | How long has this domain been trusted? Does it have a track record of quality? |
| Structured data / Schema | Medium | Does the page have machine-readable markup helping Google understand content type? |
E-E-A-T: The Framework Google Uses for Trust
Since 2022, Google has used E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) as its primary quality framework, particularly for YMYL (Your Money or Your Life) topics: health, finance, legal and major purchase decisions.
Does the author have first-hand experience with the topic? Google increasingly looks for indicators that the person creating content has lived or worked with the subject, not just researched it.
Does the creator have demonstrable knowledge in their field? This is shown through credentials, depth of content, citations and consistency across a body of work.
Is the creator or the website recognised as an authority in their field by other sources? This is primarily measured through backlinks and mentions from other trusted sites.
Is the site accurate, transparent and honest? This includes having clear authorship, accurate information, HTTPS, clear contact information and no deceptive practices.
How Google Uses AI to Rank Pages
Google uses several AI systems as part of its ranking process. RankBrain (2015) interprets queries it has never seen before by understanding context and meaning. BERT (2019) understands the nuance of longer, conversational searches. MUM (2021) processes information across multiple formats and languages simultaneously.
The practical implication is that keyword stuffing and exact-match manipulation are far less effective than they were 10 years ago. Google now understands what a page is actually about and whether it genuinely serves the searcher, rather than just matching text strings.
In 2024, internal Google documents were accidentally leaked online, providing the most detailed confirmed picture of ranking factors ever published. The leak confirmed: clicks and user behaviour DO influence rankings, domain authority IS a real metric (called siteAuthority internally), and the freshness of pages matters more than Google had publicly admitted. The documents confirmed much of what experienced SEOs had observed for years.
Understanding how Google ranks pages is step one. Applying it to your specific site, competitors and keywords is where the work happens. Book a free call and find out exactly what it would take for your site to rank where it needs to.