How search works
How search engines workThe five steps from discovery to rankingComponentsThe parts that make a search engine runCrawlingHow bots find and fetch your pagesIndexingHow a page gets stored so it can be foundCheck if indexedFour ways to confirm a page is on GoogleCrawling and indexing
Web crawlersGooglebot, Bingbot and the new AI botsGooglebotHow Google's crawler reads your siteRobots.txtRules, examples and a free testerXML sitemapCreate one, then submit it to GoogleCanonical tagPick the main version of a pageNoindex tagKeep a page out of search resultsRanking and results
Search engine rankingWhy one page appears above anotherGoogle ranking factorsWhat really matters, grouped simplyPageRankHow links pass value between pagesAlgorithmsTF-IDF, PageRank and modern rankingResults page (SERP)What each block on the page is forAI searchHow AI answers are built from the indexUnderstand
What is a search engineA plain definition with real examplesTypes of search enginesCrawler, meta, vertical, private and AIBrowser vs search engineTwo tools that people often mix upExplore
Best search enginesGoogle, Bing, DuckDuckGo and moreUses of search enginesThe everyday jobs that search handlesHistory of search enginesFrom Archie in 1990 to AI answersSearch engine ranking is the order in which a search engine lists pages for a query. Every time someone searches, the engine compares the pages it has stored, then places the ones it judges most useful at the top. This hub explains how search engines rank pages, which signal families matter, why positions move, then what a site owner can realistically influence.
Ranking is the step where a search engine sorts every matching page for a query, putting the most relevant, trustworthy and usable results first. The order is calculated fresh for each search, which is why the same page can rank well for one query yet poorly for another.
The basics
A results page is a short list chosen from a very large pile. Ranking is the process that decides the order of that list, from the first blue link downwards.
Imagine asking a well-read friend for a good book about gardening. She knows hundreds of titles, yet she only names three, starting with the one she thinks suits you best. A search engine does the same thing at a vast scale. It holds copies of billions of pages in its index, a giant catalogue built in advance, then picks a handful for each person who asks.
The place a page holds on the results page, where position one is the top organic result.
Example: third resultThese are the words someone types or speaks into the search box. Rankings always belong to a specific query, never to a whole website.
Example: best running shoesA listing that earned its place through ranking, as opposed to an ad that someone paid for.
Example: a blog postBecause rankings are tied to a query, it makes little sense to say a site “ranks well” in general. A bakery might hold the top spot for its own name, sit on page two for “cakes near me”, then not appear at all for “wedding catering”.
Two stages
Before any page can be ordered, the engine has to gather a set of candidates. These two jobs are called retrieval and ranking.
| Retrieval | Ranking | |
|---|---|---|
| Main question | Which pages could possibly answer this? | Which of those pages answers it best? |
| Speed | Very fast, often using simple word matching | Slower per page, using many more signals |
| Output | A long list of candidates | A short, ordered list of results |
| Analogy | Pulling every gardening book off the shelf | Choosing the three best ones to recommend |
Retrieval only works on pages that are already stored, so a page has to be crawled and indexed first. Our guide to how search engines work covers those earlier steps, while the components page shows which part of the system handles each job.
What gets measured
Google explains its approach in terms of broad families of signals rather than a single formula. Other engines publish less detail, although their goals are very similar.
The engine interprets the query first, correcting spelling, spotting synonyms, then guessing whether you want a fact, a product or a place.
The most basic check is whether a page contains the words or ideas searched for, including in its headings or main text.
Systems look for content that shows experience, expertise and care, along with signs that other reputable pages rely on it.
Pages that load quickly, work well on phones and avoid intrusive pop-ups tend to give searchers a better experience.
Location, language, past settings or device can change results, which is why “football” means different things in different countries.
A fast page with no useful answer will not rank, while a perfect answer on a broken page may struggle too.
These families describe goals rather than a checklist. Inside each one, the engine relies on many separate systems that are updated over time.
Interactive · Signal mixer
Drag the sliders to decide how much each signal matters, then watch three sample pages swap places for the same search.
How much each signal counts
This is a teaching model with four made-up numbers per page. Real engines blend hundreds of signals, and nobody outside can see the exact weights.
Ranking factors
People often search for a complete list of ranking factors with exact weights. That list does not exist in public form, which is worth understanding before trusting any article that claims otherwise.
Google has confirmed some individual signals over the years, such as HTTPS as a lightweight signal in 2014. It also publishes a guide to its named ranking systems. What it does not publish is how much each signal counts, because the balance changes with the type of query. Freshness matters a lot for news searches, for example, yet very little for a question about the boiling point of water.
Myth: there are exactly 200 ranking factors.
The number is often repeated, yet no current official list exists. Modern systems combine far more inputs than any fixed count suggests.
Myth: one trick can push a page to the top.
Ranking systems weigh many signals together, so changing a single element rarely makes a weak page competitive.
Myth: a longer page will always rank higher.
Length helps only when the query needs depth. A short, precise answer can outrank a long page for a simple question.
Myth: paying for ads improves organic positions.
Google states that advertising does not influence organic ranking, because the two systems are kept separate.
Movement
Positions on a results page are rarely fixed for long. Even when nothing on your own page changes, the world around it keeps moving.
| Cause | What happens | What to do |
|---|---|---|
| Algorithm updates | The engine adjusts how it weighs signals, sometimes in a broad core update | Review whether your page is still the most helpful answer |
| New competition | Another site publishes a fuller or fresher page on the same topic | Compare their page to yours, then fill the real gaps |
| Intent shifts | Searchers start wanting something different from the same words | Look at the current top results to see what now ranks |
| Technical problems | A page breaks, slows down or is accidentally blocked | Check Search Console for crawl or indexing errors |
| Seasonality | Interest rises or falls with holidays, weather or events | Compare the same period from earlier years |
Google confirms its larger updates on the Search Status Dashboard, while smaller changes happen almost every day without an announcement. A drop that lines up with a confirmed core update usually means the engine has reassessed relevance or quality across many sites, rather than penalising one page.
Beyond Google
Every engine builds its own index and its own ranking systems. That explains why the same query can return a different order on Google, Bing or a privacy-focused engine.
Google publishes guidance through Search Central, including its guide to ranking systems and its notes on core updates.
Bing runs Bingbot alongside its own ranking systems, then publishes Webmaster Guidelines that cover similar ground.
Engines such as DuckDuckGo avoid personal profiles, so context signals depend more on location settings than on search history.
AI Overviews or answer engines like Perplexity still rely on retrieving pages first, then summarise what they find.
For site owners
Nobody controls rankings directly, yet the basics below give a page a fair chance. They follow the signal families described above.
For a full grounding in every concept behind these steps, our search engine basics guide walks through crawling, indexing and ranking with interactive demos. You can also study the results page itself to see which features compete with organic listings.
Go deeper
Signal families describe what the engine cares about. Algorithms are the methods it uses to measure those things.
Formulas such as TF-IDF or BM25 score how well a page’s words match the query.
RelevanceSystems like RankBrain or BERT help match the meaning of a query, even without exact words.
MeaningEach of these layers is explained with simple examples on our page about search engine algorithms, which also covers how core updates fit into the picture.
FAQ
They retrieve pages that match the query, then score each one on meaning, relevance, quality, usability and context. The page with the strongest overall score for that specific search appears first.
Google does not publish a ranked list. Its documentation points to helpful, relevant content as the core, supported by links, page experience and the searcher’s context.
Common reasons include a core update, a competitor publishing a stronger page, a shift in what searchers want or a technical fault. Check Search Console first to rule out errors.
No, paid ads appear in separate, labelled slots, while organic positions are decided by ranking systems that advertising does not affect.
It varies widely, from days to many months. Competitive topics usually take longer, because established pages already answer those queries well.
Keep learning