Google algorithm: how Google ranks pages

What is the Google algorithm?

The Google algorithm refers to all the systems that decide which pages to show, and in what order, for a given search. The singular is misleading. Google itself talks about “ranking systems”, in the plural, that work together on each query.

Its documentation “In-depth guide to how Google Search works” was updated on December 18, 2025. It states that relevance depends on an algorithm based on hundreds of factors. Among them: the user’s location, language and device. Two people who type the same query therefore don’t always see the same results page.

Google does not publish the list of its factors or their weight. Any page that claims to give “Google’s 200 ranking factors” rebuilds a list from patents, tests and statements. It can be useful, as long as you read it as a hypothesis.

How does the Google algorithm work, from crawling to ranking?

Google describes three successive stages. Not every page gets through them. According to the same documentation, the search engine accepts no payment to crawl a site more often or rank it higher.

1. Crawling: robots, including Googlebot, discover URLs by following links and sitemaps, then download text, images and videos.
2. Indexing: Google analyses the content and runs the JavaScript with a recent version of Chrome. It groups duplicate pages and chooses a canonical URL. The page enters the index, or not.
3. Serving results: for each query, the ranking systems draw on the index and order the pages judged most useful.

Order matters in SEO. A page that is never crawled will never be indexed. A page missing from the index cannot rank, however good its text. The causes of non-indexing and their fixes are covered in the guide to Google indexing.

What criteria does the Google algorithm take into account?

On its public site “How Search works”, Google groups its criteria into five families. It is the most official presentation of ranking factors that exists.

Family	Question Google asks	What you can work on
Meaning	What is the user really looking for?	Target the intent, not just the keyword
Relevance	Does the page cover this topic and this intent?	Cover the question and its sub-questions
Quality	Does the content show expertise, authoritativeness and trustworthiness?	Named author, sources, evidence
Usability	Does the page display quickly and properly, on mobile too?	Performance, layout, accessibility
Context	Where, in which language, on which device?	Local version, language, mobile parity

The first criterion, meaning, explains why the same page cannot target two different needs. The guide to search intent shows how to read a SERP before writing. The quality criterion refers to the E-E-A-T framework: experience, expertise, authoritativeness, trustworthiness.

To improve its algorithm, Google says on the same page that it runs hundreds of thousands of experiments every year. Each change is tested before it is rolled out.

What are the main systems of the Google algorithm?

Google maintains a guide to its ranking systems, revised on December 10, 2025. It separates active systems from those that have been retired or merged into the core of the algorithm.

System	Role documented by Google
BERT	Understand how a combination of words changes the meaning of a query
RankBrain	Link the words of a query to concepts, even without an exact match
Neural matching	Match the concepts of a query with those of a page
Passage ranking	Find the section of a page that answers the question
Link analysis and PageRank	Estimate the importance of a page from the links that point to it
Reliable information systems	Surface authoritative pages, demote low-quality sources
Freshness	Favour recent content when the query calls for it
Deduplication and site diversity	Avoid duplicates; in general, no more than two results from the same site
Original content, reviews system	Favour original reporting and reviews based on real expertise

The same guide lists retired systems. Panda, launched in 2011, joined the core systems in 2015; Penguin, launched in 2012, in 2016. The 2022 helpful content system was folded in in March 2024. Lists of “animal updates” remain useful for SEO history, much less for deciding what to do today. The role of links is covered in the article on PageRank.

What are Google algorithm updates?

Several times a year, Google rolls out core updates that reassess all content. Its page on core updates was revised on December 10, 2025. It states that these updates are announced on the Search Status Dashboard, with their start and end dates.

A drop after a Google core algorithm update does not mean a penalty. Other pages were judged more useful. Google recommends a methodical reading:

• wait at least one week after the rollout ends before comparing;
• compare performance before and after, over equivalent periods, in Search Console;
• identify the pages and queries most affected;
• assess those pages with the helpful content self-assessment questions, then improve what needs improving.

Recovery is not immediate. According to the same page, an improvement can have an effect within a few days. Yet the systems can take several months to take it into account.

Spam updates follow a different logic. They enforce Google’s spam policies: keyword stuffing, paid links, content produced at scale with no added value. Since March 2024, that last practice has a name, “scaled content abuse”, whatever tool is used to produce the pages.

The Google algorithm and AI: what changes?

AI Overviews and AI Mode add a layer of generated answers on top of the classic results. Google’s page “AI features and your website” was revised on December 10, 2025. According to it, no special optimisation is required to appear in them.

The Google algorithm remains the way in. The conditions are those of classic SEO: the page must be indexed and eligible to be shown with a snippet. Google also describes “query fan-out” there: the engine runs several searches on subtopics to build its answer. A page that covers the sub-questions of a topic well increases its chances of being cited.

The article on AI Overviews explains how they work, their effect on clicks and how to measure them.

How do you learn SEO from these fundamentals?

Learning SEO starts with the source. Google publishes a starter guide and documentation that is updated continuously: anything that contradicts it deserves checking.

1. Read the official groundwork: Google’s SEO Starter Guide, then the pages on crawling, indexing and ranking systems.
2. Open the Search Console of a real site: the Performance and Page indexing reports. The Google Search Console guide explains what to look at first.
3. Test on a small site: change a title tag, track the effect on impressions and click-through rate, note the delays.
4. Organise your knowledge by pillar: technical, content, popularity. The guide to organic search serves as the overall map.
5. Date everything you read: a rule from 2019 can be wrong in 2026. FID, for example, was replaced by INP in March 2024 in the Core Web Vitals.