How Search Engines Use Links?

Search engines use links to find pages, work out what they are about, and decide how to order them in results. Here is how each step works.
Search engines use links to find pages, work out what those pages are about, and decide how to order them in results. A link is a path for a crawler, a clue about the linked page, and one vote among many ranking signals.
Key takeaways
- Google describes Search in three stages: crawling, indexing, and serving results, and not every page passes each stage (Google Search Central, as of November 2026).
- A crawler finds new pages by following links from pages it already knows, and by reading sitemaps that site owners submit.
- Anchor text and the words near a link help a search engine judge what the linked page is about.
- PageRank is still part of Google's core ranking systems, and it treats links as votes, with votes from strong pages counting more.
- In Ahrefs' January 2025 study of 1,000,000 keywords, referring domains correlated with rankings at 0.255, ahead of total backlinks at 0.248.
- Buying links to pass ranking credit is link spam under Google's spam policies, as of November 2026.
What a link tells a search engine
A link tells a search engine that one page points to another page, and that the linked page may be worth a visit. The link itself carries three kinds of information: the address it points to, the clickable words used for it, and the page it comes from.
The address tells the crawler where to go next. The clickable words, called anchor text, give a short label for the linked page. The source page gives context, because a link from a page on the same topic says more about relevance than a link from an unrelated page.
Link equity is the ranking value that a link passes from one page to another. It is not a public score, and Google does not show it. The value depends on the linking page, the number of links on that page, and how the link is marked.
A single link never decides a ranking on its own. Search engines weigh links with content, relevance, location, language, and device, so a link helps most when the linked page also matches the query well.
How links help crawlers find new pages
Links help crawlers find new pages because a crawler starts from pages it already knows and follows the links on those pages. Each new address it finds goes into a queue of pages to visit.
Google calls this first step URL discovery. There is no central list of all web pages, so Google must keep looking for new and changed pages and add them to its list of known pages. Sitemaps add more addresses, but links remain a main way new pages are found.
Google's crawler is called Googlebot. It decides which sites to visit, how often to visit, and how many pages to fetch. It also tries not to visit a site so fast that it slows the site down.
A page with no links pointing to it is hard for a crawler to find. Internal links from pages Google already crawls give a new page a path into the crawl queue. That is why a new page should be linked from a live page, not left with no incoming links.
Google does not accept payment to crawl a site more often or to rank it higher. Google also does not guarantee that it will crawl, index, or serve a page, even if the page follows its rules.
How links shape indexing and relevance
Links shape indexing because a search engine reads a link and its nearby words while it works out what the linked page is about. The linked page is then stored in the index with that information attached.
Indexing is the stage where Google analyses the text, images, and video files on a page. It also looks at key tags and attributes, such as title elements and alt text. Links add to that picture by showing how other pages describe the linked page.
Anchor text is the clickable text of a link, and it acts as a short label for the page it points to. Clear, honest anchor text that matches the linked page helps a search engine place that page in the right topic. Misleading anchor text does not help, because the page content still has to match.
Google groups pages with similar content and picks one page, called the canonical page, to show in results. Links to the wrong copy can split that signal, so links should point to the version you want shown.
Indexing is not guaranteed. Google states that not every page it processes will be indexed, so a link can bring a crawler to a page and still leave that page out of the index.
How links feed ranking and PageRank
Links feed ranking because Google's systems look at how pages link to each other to work out what pages are about and which pages may help most. A link from another page acts as a sign that the linked page is worth notice.
PageRank is Google's link-based score of a page's importance, set by the number and quality of pages that link to it. Google still lists it as one of its core ranking systems, in a form that has changed a lot since the original version.
The original model used a damping factor of 0.85, so each page passed about 85 percent of its value on through its links. That figure comes from the 1998 PageRank paper by Larry Page and Sergey Brin at Stanford University. Google has said the system has changed since then, so the old formula is a model, not a live map.
How Google PageRank works in practice is simple to state: each page shares its score across its outbound links. A page with 10 outbound links gives each target a larger share than a page with 100 outbound links gives each of its targets.
Google removed the public 0 to 10 toolbar PageRank score in 2016. No tool outside Google can show a live PageRank value today.
What the large studies show
The large studies show the same pattern: pages that rank well tend to have more links, and more linking sites, than the pages below them. These are correlations, not proof that links alone caused the ranking.
Backlinko's analysis of 11.8 million Google search results, updated in April 2025, found that the top result had 3.8 times more backlinks than results in positions 2 to 10. About 95 percent of pages in that set had zero backlinks.
Ahrefs studied the top 1,000,000 keywords in the United States in January 2025. Referring domains had a Spearman correlation of 0.255 with ranking position, total backlinks had 0.248, and Domain Rating had 0.131. Ahrefs warned that correlation is not causation and called these weak correlations.
| Study | Sample | Main finding |
|---|---|---|
| Backlinko, 2025 | 11.8 million Google results | Top result had 3.8 times more backlinks than positions 2 to 10 |
| Ahrefs, 2025 | Top 1,000,000 United States keywords | Referring domains correlated at 0.255, backlinks at 0.248 |
The next two studies show the same pattern from a different angle.
| Study | Sample | Main finding |
|---|---|---|
| Semrush, 2024 | 16,298 keywords and 300,000 positions | Top 10 median was 6 referring domains and 13 backlinks |
| Ahrefs, 2023 | About 14 billion pages | 96.55 percent of pages got zero traffic from Google |
What the medians and local results tell you
The median is a better guide than the average for most pages. Semrush's 2024 study found the median top 10 page had 6 referring domains, while the average was 112. A few pages with hundreds of links pull the average up.
The same Ahrefs study found links mattered more for local queries, where the correlation reached 0.33. For comparable low-volume searches, the figures had slipped from 0.29 in a 2019 Ahrefs study to about 0.22 to 0.24 in 2025.
Treat every figure in this guide as a correlation, not as proof of cause. A page can earn links because it already ranks well, or because it is useful, so links and rankings rise together. Use the studies to set a target, then check your own pages against the live top 10.
Internal links and external links compared
Internal links and external links do different jobs for a search engine, and both matter. Internal links join pages on the same site, while external links join pages on different sites.
Internal links vs external links differ in who controls them and what they prove. You control your internal links, so they show a search engine how your site is organised. You do not control external links, so they act as votes from other sites.
Internal links help crawlers reach deep pages, spread ranking value from strong pages to weaker pages, and show which pages on a site matter most. A page with no incoming internal links, called an orphan page, is easy for a crawler to miss.
External links help a search engine judge trust and relevance from outside the site. A link from a relevant, respected site says more than many links from unrelated sites. Ten links from 10 different sites also say more than 10 links from one site, because each site made its own choice.
Both kinds of links need honest anchor text and working addresses. A broken internal link wastes a crawl, and a broken external link sends a reader and a crawler to a dead page.
How search engines treat paid and marked links
Search engines treat paid links as advertising, not as editorial votes, and they expect paid links to be marked. An unmarked paid link that passes ranking credit counts as link spam.
Google defines link spam as creating links to or from a site mainly to manipulate search rankings. Its spam policies list buying or selling links for ranking purposes as an example, as of November 2026. Sites that break those policies may rank lower or not appear in results at all.
A paid link is allowed when it carries rel="sponsored" or rel="nofollow", because the mark tells Google not to pass ranking credit through it. Google introduced rel="sponsored" and rel="ugc" on 10 September 2019, and it treats sponsored, ugc, and nofollow as hints, not strict orders.
Nofollow links use rel="nofollow" to say the linking site does not vouch for the linked page. Google began treating nofollow as a hint for ranking on 10 September 2019, and as a hint for crawling and indexing on 1 March 2020.
Google's link spam policies also cover large-scale link exchanges and automated link building. The December 2022 link spam update, which started on 14 December 2022, used Google's SpamBrain system to neutralise spammy links and cancel the credit they passed.
How to check links on your own site
You can check links on your site in about 30 minutes with free tools, then repeat the check once a month. The steps below use Google Search Console, which is free for sites you own and verify.
- Open Google Search Console, pick your property, and open the Links report to see top linked pages, top linking sites, and top linking text.
- Export the full list from the main Links page, not from one table, because tables stop at 1,000 rows while the export reaches 100,000 rows.
- Check that each key page has incoming internal links, and add links from live pages to any key page with none.
- Open the top linked pages in a browser and check that each link still loads, points to the right page, and uses honest anchor text.
- Compare your page with the top 10 pages for your target keyword, and note the referring domains of each page.
Tools, time and cost to check links
A backlink audit goes further than this quick check, because it reviews the full list of linking sites for spam, relevance, and lost links. Do the quick check monthly, and run the full audit when traffic drops or after a site move.
Paid tools such as Ahrefs and Semrush add competitor data that Search Console cannot show, because Search Console only covers sites you own. Check the vendor pricing page before you buy, as of November 2026.
Measure new referring domains each month, not raw link counts. A referring domain is one unique site that links to you, so 10 links from one site count as 1 referring domain. That is the metric that led the link measures in Ahrefs' 2025 study, at 0.255.
Common mistakes
- Counting raw links instead of referring domains: 10 links from one site count as 1 referring domain, so a big link total can hide a narrow base. Track new referring domains each month.
- Leaving key pages with no incoming links: a page no other page links to is hard for a crawler to find. Link each key page from at least one live page on your site.
- Using misleading anchor text: anchor text that promises one topic and lands on another does not help the linked page. Use words that match the page the link points to.
- Buying links to pass ranking credit: buying links is against Google's guidelines when the links pass ranking credit. Mark paid links sponsored or nofollow, as of November 2026.
- Judging links by one number alone: Domain Rating had only a 0.131 correlation with rankings in Ahrefs' 2025 study. Use it with referring domains, relevance, and traffic, not on its own.
Frequently Asked Questions
How do search engines find pages through links?
A crawler starts from pages it already knows, downloads each page, reads the links on it, and adds new addresses to a queue of pages to visit. Sitemaps add more addresses, but links remain a main path. A page with no links pointing to it is hard to find, so link new pages from live pages.
Do links still matter for Google rankings in 2026?
Yes, links still matter, but they matter less than they used to. Ahrefs' January 2025 study of 1,000,000 keywords found referring domains correlated with rankings at 0.255 and total backlinks at 0.248. Those are weak correlations, and Ahrefs warned that correlation is not causation. Links are one signal among many, with content and relevance beside them.
What is the difference between a backlink and a referring domain?
A backlink is one link from one page to another page. A referring domain is one unique website that links to you, however many links it sends. If one blog links to you 10 times, that is 10 backlinks but 1 referring domain. Referring domains matter more, because each new domain is a separate site choosing to link to you.
Do nofollow links help rankings at all?
Nofollow links generally do not pass ranking credit, because Google treats nofollow as a hint to consider or exclude the link. Google began treating it as a hint for ranking on 10 September 2019. A nofollow link can still bring visitors and help a crawler find a page, but it is not a ranking vote.
Can a search engine rank a page with no backlinks?
Yes, a page can rank with no backlinks, but it is rare for competitive keywords. Backlinko's analysis of 11.8 million Google results found about 95 percent of pages had zero backlinks, and most of those pages get little traffic. A page with no links needs strong content, clear relevance, and internal links from pages that do have links.
Sources
The BacklinkPrice Team researches backlink prices, publisher sites and link building tools. We check prices and plans against each provider's own pages and date them, so you know how current the numbers are.
More from BacklinkPrice Team




