Internal Linking: The Link Building You Already Control

Every other link building guide is about links you have to ask strangers for. This module is about the links you already own. An internal link, in plain terms, is a hyperlink that points to another page on the same website: the one part of link building where nothing has to be pitched, bought, or inspected for a sponsored tag. Most sites of a few hundred to a few thousand pages have never done this work deliberately, which means orphan pages nobody links to, money pages linked only from the footer, anchors that say "click here", and hrefs that point at redirects.

By the end of this module you will have crawled the site, listed orphans and under-linked money pages, chosen targets with proper anchors, placed the links, fixed broken and redirected hrefs, and measured the before and after. That is the whole job, and it costs time and no money.

One honest limit sits over everything below. Internal linking will not rank a page Google thinks should not rank. What it does is tell Google which of your pages you think matter: a clear structure helps Google understand which pages you consider the most important, even though a clean structure alone does not make a site rank better overall. On most sites that message is currently noise. Your task this month is to turn it into a signal.

What internal links actually do

Google's own documentation gives internal links three jobs, and you should learn all three before you touch anything. First, discovery: "Google uses links as a signal when determining the relevancy of pages and to find new pages to crawl." The SEO Starter Guide puts the scale on it: "In fact, the vast majority of the new pages Google finds every day are through links, making links a crucial resource you need to consider to help your pages be discovered by Google and potentially shown in search results." If no other page on your site links to a page, Google may never find it at all.

Second, relevance. Anchor text, which is the visible text of a link, tells people and Google something about the page being linked to. "Link text (also known as anchor text) is the text part of a link that you can see. This text tells users and Google something about the page you're linking to." Descriptive anchors help both readers and Google make sense of your site and find other pages on it.

Third, importance. Google's bar for this is one link: "Every page you care about should have a link from at least one other page on your site." That sentence is the pass mark for the entire module. Behind it sits PageRank, Google's long-running link-analysis system and still one of its core ranking systems, much evolved since launch. Google's current framing is worth quoting exactly, because vendors routinely overstate it: "While PageRank uses links and is one of the fundamental algorithms at Google, there's much more to Google Search than just links. We have many ranking signals, and PageRank is just one of those." It is one signal among many, not the scoreboard.

Before you start: what you need

You need three things, and you should gather all three before you run anything.

First, a crawler. Any of these will do the job on a site your size, with prices as listed on 7 September 2026. Screaming Frog SEO Spider crawls 500 URLs free; a £199-a-year licence removes the limit and adds JavaScript rendering, the Search Console and Analytics integrations, crawl comparison, and scheduling. Sitebulb Desktop covers 10,000 URLs per audit on Lite and up to 500,000 on Pro, with a 14-day Pro-equivalent trial and Cloud plans from £95 a month. Ahrefs Site Audit is included from Lite at €119 a month, with 100,000 crawl credits a month and up to 25,000 pages per project. Semrush Site Audit sits on the SEO plan at $117.33 a month billed annually, or $139 month to month.

Second, Search Console access to the site. Its Links report is free, and it shows what Google has actually sampled, which is a different thing from what your crawler sees. You will use both.

Third, a written list of the pages that make the site money: service pages, category pages, flagship guides, whatever converts. Aim for ten to thirty URLs. Every later step in this module asks which pages you care about, and if you cannot name them now you will spread links evenly across pages that do not deserve them.

Lesson 1: crawl the site and export the links

Run the crawl first and interpret it second. Enter your homepage, let the crawler follow your internal links the way Googlebot discovers hyperlinks in HTML, and wait for it to finish. On a few hundred URLs the free Screaming Frog tier may be enough; on a few thousand pages you need the licence. If your site renders important links with JavaScript, enable JavaScript rendering so the crawl sees what visitors see.

When the crawl completes, open the Internal tab in Screaming Frog and learn four columns. Inlinks counts every internal link to a URL, including repeats. Unique Inlinks counts each linking page once, so a post linking to your pricing page three times is one unique inlink and three total inlinks. Crawl Depth counts clicks from the homepage, with the homepage itself at zero. Link Score appears after Crawl Analysis runs: Screaming Frog's own 0-to-100 model of internal PageRank, covered in Lesson 3. Also pull the bulk export of internal outlink anchors, which you will need when you audit wording in Lesson 4.

In Ahrefs, open Site Audit and work the Links section, including the "Orphan page (has no incoming internal links)" flag, the Internal link opportunities report, and the Internal pages section for 4XX targets. In Semrush, open the Internal Linking report in Site Audit and work its named issues: broken internal links, internal redirects, chains and loops, "Too many internal links", "Pages with only one internal link", "Orphaned sitemap pages", and depth beyond three clicks.

Then open Search Console's Links report and read it for what it is. The Top internally-linked pages table shows which of your pages is linked the most from within your own site, and selecting a URL shows which of your other pages link to it. But this report is a sample, not a census: "This report isn't a comprehensive list of every link on your site. It shows a sample of internal and external links to help you understand your site's overall link profile." Pages are grouped by canonical, tables are capped at 1,000 rows, dead links may linger, and nofollow is not flagged. Use it to confirm what Google has seen. Never use it as your crawl.

Lesson 2: find the orphans and the under-linked pages

An orphan page is a page no internal link path reaches from your crawl's start point, usually the homepage. It can still have links from other orphan pages, which does not rescue it. These pages get no internal ranking credit passed to them, and Google may never discover them except through a sitemap or an outside backlink. On a site that has never done internal linking deliberately, expect to find some.

Find them with the crawler, not with Search Console. Search Console's Links report lists your most-linked pages; it has no orphan list, so a URL's absence from its tables proves nothing. In Screaming Frog, connect your sitemaps plus the Analytics and Search Console integrations, run Crawl Analysis, then check the Orphan URLs filters or pull Reports > Orphan Pages. A quicker check on the Internal tab is a blank Crawl Depth, meaning the crawl never reached that URL by following links. In Ahrefs, open Site Audit > Links > Issues and read the "Orphan page (has no incoming internal links)" rows. In Semrush, open the Internal Linking report > Notices > "Orphaned sitemap pages".

Then widen the net from zero links to too few. Sort the crawl by Unique Inlinks, lowest first, and lay your money-page list alongside it. Any page you care about with zero, one, or two incoming internal links is under-linked, whatever the site average says. Semrush agrees: "Pages with only one internal link are harder to find and may seem less important to search engines." Flag every money page sitting more than three clicks deep as well, since depth and inlink count usually fail together. We recommend closing this lesson with two columns on paper — pages with no incoming links, and pages you care about with barely any — because those two lists are the input to Lesson 4.

Lesson 3: find where the equity sits

Some pages on your site carry more weight to pass on than others, and your next move is to find them. Every crawler scores this differently, so learn what each number actually is before you sort by it. None of them is Google's number. John Mueller has said of do-it-yourself internal PageRank calculators, "I wouldn't see it as something that is reflected one to one at Google." Treat every score below as a way to rank your own pages against each other, never as Google's verdict on them.

Screaming Frog's Link Score is the closest to a pure internal model: an internal PageRank-style calculation on a 0-to-100 logarithmic scale, computed on the Internal tab after Crawl Analysis, where the strongest page scores 100. It counts internal links only, and in this model a nofollowed link passes nothing while still counting as an outlink. That evaporation is Screaming Frog's modelling choice, not Google's instruction, and Lesson 6 explains why you must not design around it.

Ahrefs gives you two separate 100-point scores, and you should not confuse them. Site Audit's Page Rating runs 1 to 100 on a logarithmic scale and measures a URL's combined internal and external backlink profile, splitting each page's rating equally among the URLs it links to. URL Rating is a different 100-point score in Site Explorer that also counts internal and external links together, built on the same basic PageRank principles. Ahrefs states plainly that nobody outside Google knows exactly how PageRank is calculated. Both scores mix outside backlinks with internal links, so a high score may reflect external authority rather than internal position.

Semrush names no 0-to-100 internal equity metric, so do not go looking for one. Use the Internal Linking report's inlink counts and crawl-depth flags instead. Search Console's "Top linked pages from within your own site" is likewise a count of internal links in Google's sample, not an equity score.

Sort whichever column your tool gives you, highest first, and write down your ten strongest pages. Your linking plan in the next two lessons runs from these pages toward the orphans and money pages from Lesson 2. We recommend starting there because a link from a page that already carries weight moves more than a link from a page nobody visits.

Lesson 4: choose targets and anchors

Targets first, anchors second, because the wrong target with perfect wording still fails. Your targets are the pages you care about: the money list from Lesson 2 plus any orphan worth keeping. Mueller's guidance is direct: "Internal linking is super critical for SEO. It's one of the biggest things you can do on a website to guide Google and visitors to the pages that you think are important... What you think is important is totally up to you." Pick ten to twenty targets at most. A plan that targets everything targets nothing.

Now give each target its anchor text. Anchor text is the visible text of the link, and Google's rule for it is short enough to memorise: "Good anchor text is descriptive, reasonably concise, and relevant to the page that it's on and to the page it links to." Write each anchor as naturally as possible, naming the topic in words a reader would expect. The warning sits beside the rule: "Write as naturally as possible, and resist the urge to cram every keyword that's related to the page that you're linking to (remember, keyword stuffing is a violation of our spam policies)." Keyword stuffing in anchors is a spam-policy violation, not an optimisation.

Google also shows you what bad looks like: empty anchors, "click here", "read more", and "website". If the link is an image, its anchor is the image's alt text, so audit those too. Screaming Frog can filter non-descriptive anchors in internal outlinks and bulk-export them, which turns this step into a spreadsheet task.

One finding complicates the picture, and you should know both sides of it. A Zyppy study of 23 million internal links across 1,800 sites and roughly 520,000 URLs, compared against Search Console clicks, found that "Pages with at least one exact match anchor had at least five times more traffic than pages without." Its author calls the work a correlation study whose conclusions are directionally useful rather than scientific truth, and correlation is not causation. Set that against Google's instruction to write naturally and resist cramming keywords. The course position reconciles them: give each target one descriptive anchor that names its topic plainly, then vary the wording across further links instead of repeating one phrase. Semrush's "five words or fewer" anchor rule is Semrush's house rule, not Google's, and Ahrefs' Internal link opportunities are its keyword matcher's suggestions, not Google's instructions.

To find placements, search your own pages for unlinked mentions of each target's topic. Screaming Frog's Custom Search does exactly this with its Page Text No Anchors search, which excludes navigation and footer mentions by default. Give every target at least one such placement before you add seconds.

Lesson 5: where the links go

Put your new links in body copy first, in context, from pages on the same topic. Google's link documentation advises linking to other on-site resources that help a reader, in context, and its sitelinks guidance says to "link to your important pages from other relevant pages" with concise, relevant anchors. A relevant paragraph beats a sidebar every time, because the surrounding text tells Google and the reader what the target is about.

Navigation and footer links are still legitimate. Asked about four identical sitewide links to one page, Mueller answered: "Having 4 identical links on a page to another page seems fine & common to me, I wouldn't worry about that." The older warning people quote was about a site that links from all pages to all pages, where Google loses context of how pages sit within the site. Related-content modules are likewise an ordinary internal-link type. Never disallow footer URLs in robots.txt to "save" crawl budget; when an owner asked about exactly that, Mueller answered, "If you do this, future-you will be annoyed by the problems current-you is creating."

One more piece of evidence shapes placement. That Zyppy study found clicks rising with incoming-link count up to about 40 to 44 links, then reversing: "After a URL receives about 45-50 internal links, the effect reverses. Google traffic begins to decline as the number of internal links rises." The author's explanation is that pages with 50-plus incoming links are usually sitting in sitewide navigation, while the winners below that line earn unique body links. Read it as support for this lesson's order: body links first, navigation as backup, never a footer dump as the whole strategy.

How many links per page

Google refuses to give you a number, and you should respect that. Its full statement: "There's no magical ideal number of links a given page should contain. However, if you think it's too much, then it probably is." Any vendor number below is that vendor's judgment, not Google's.

Know the three you will meet. Moz describes a rough crawl heuristic of 150 links per page before search engines may stop following further links, flexible for important pages up to 200 or even 250, and advises staying under 150 in general practice. Ahrefs' author suggests three to five contextual links per article as a starting point, alongside normal navigation. Zyppy's correlation peaked at 40 to 44 incoming links per URL before reversing, which counts links pointing at a page, not links placed on one, so never recite it as a per-page cap.

The defensible rule we recommend is this: every page you care about gets at least one contextual body link from a relevant page, important pages sit within three clicks of the homepage, and you stop adding links to a page when a reader would call it cluttered. That rule cites Google's one-link bar, the widely shared three-click depth flag, and Google's own too-much test, in that order.

Lesson 6: mistakes that waste your month

Each mistake below comes with the Google statement that settles it. Learn the statement and you will never pay for the myth again.

Nofollow sculpting fails. The theory says marking internal links nofollow keeps ranking credit for the pages you prefer. Mueller's answer: "You don't hoard anything when you make links nofollow. It's a common SEO myth." Asked about nofollowing links to a terms page, he added that it means work you maintain forever, yet "doesn't change anything for your website" — wasted work, in his word. Google's own rule reserves nofollow for links you do not trust: "Use nofollow only when you don't trust the source, and not for every external link on your site." Links carrying these rel attributes will generally not be followed, which makes nofollow the wrong tool for shaping your own site; to stop Google fetching one of your own URLs, the documented control is a robots.txt disallow. Audit your internal links for nofollow attributes and remove them except on user-generated content, where your system should add nofollow or a similar annotation automatically.

Buttons are not links. If a "link" is a <button> element with a script behind it, Googlebot will not click it: "Googlebot won't click on these buttons to see what happens. So we would not see that there's a link associated to another URL within your website." The fix is to "use normal HTML links and just style it with CSS to make it look like a button rather than to use button elements in HTML". Google's crawlable-link rule agrees: "Generally, Google can only crawl your link if it's an <a> HTML element (also known as anchor element) with an href attribute." Check pagination, faceted navigation, and call-to-action buttons for <span href>, onclick-without-href, and javascript: hrefs, and convert every one that matters into a real anchor. Paginated sequences need <a href> links from each page to the next, with each page carrying its own canonical rather than all of them canonicalising to page one.

Chasing a first-link rule wastes an afternoon. When a page links twice to the same URL, someone will tell you only the first anchor counts — but Google's documentation states no such rule, and the idea traces to a 2008 practitioner post. When an owner proposed converting a homepage button into a non-link so a lower FAQ anchor would "win", Mueller answered in July 2026: "I suspect you're overthinking it, Google has practice dealing with lots of websites so I wouldn't expect you to see any visible change there." Leave both links as normal anchors and improve the wording instead.

Hidden links are spam. Placing links where visitors cannot easily see them — white text on white, CSS off-screen, zero opacity, a single linked hyphen — violates Google's hidden-text-and-links policy. Accordions and tabs the user can open are explicitly not a violation, so collapsible content your readers use is safe. Keep every internal link visible to a human reader.

Linking every page to every page destroys the structure you are building. Mueller's older warning was about exactly this pattern: a site that links from all pages to all pages loses the context of how pages sit within the site. Automated widgets that spray keyword-rich links everywhere belong in the same bin.

Pointing hrefs at the wrong URL wastes every link you just built. Link to the canonical URL rather than a duplicate, parameter variant, or redirecting address, because consistent canonical linking helps Google consolidate signals onto your preferred URL. Lesson 7 works through this repair.

Lesson 7: fix broken and redirected internal links

Open the same reports from Lesson 1 and work them top to bottom: broken-link and redirect reports with source URLs, Ahrefs' Internal pages for 4XX targets, Semrush's broken internal links, internal redirects, and chains and loops. Every row names a source page whose href needs editing, so export the list before you touch anything.

Apply the same repair everywhere: point the href at the canonical live URL. If the target redirects permanently, replace the href with the redirect target, since the permanent redirect marks that target as canonical. If a chain passes through several hops, update the source to the final destination. If the target is a duplicate or parameter variant, link to the canonical version your site declares. If the target is genuinely dead with no equivalent, restore the page, redirect it to the closest live equivalent and update the href, or remove the link. Check image links in the same pass, since their anchors come from alt text your editors may have left empty.

Then decide the fate of each orphan from Lesson 2. An orphan that should exist gets at least one body link from a relevant page, per Google's one-link bar. An orphan that should not exist — a retired campaign, a test page, a duplicate — gets deleted, redirected to its closest live equivalent, or noindexed with crawling allowed, and its sitemap entry removed.

Recrawl when the repairs are in. We find that teams who skip this confirmation step rediscover the same broken hrefs a year later, so confirm zero internal 4XX targets, zero internal redirect chains, and no remaining blank crawl depths except pages you deliberately removed.

Lesson 8: measure what moved

Wait before you judge anything. Google's own timescale for assessing any change: "Every change you make will take some time to be reflected on Google's end. Some changes might take effect in a few hours, others could take several months. In general, you likely want to wait a few weeks to assess whether your work had beneficial effects in Google Search results." There is no internal-links-specific timetable on record, so plan on a few weeks minimum and promise yourself no verdict before then. No internal-link change guarantees ranking movement, and this module promises none.

Measure in three places. First, recrawl with the same crawler and compare like with like: Unique Inlinks per target, Link Score or Page Rating movement, crawl-depth changes, and a clean Internal Linking issue list. Screaming Frog's crawl comparison, which needs the paid licence, exists for exactly this diff. Second, open Search Console's Links report and check the Top internally-linked pages table: confirm your targets actually rose in internal-link count inside Google's sample, and select individual URLs to see which of your pages Google credits as linking to them. Remember the report's limits: a sample grouped by canonical, capped tables, possibly stale links, no nofollow column. Third, watch the Performance report page by page for your targets over the weeks that follow.

File the before-and-after exports with dates. That folder is the evidence base for the next audit, and it stops the next round of work from starting with opinions.

Self-check: your site this month

Answer these about your own site before you close the module. Every "no" is next week's task list.

1. Does every money page have at least one body link from a relevant page, with a descriptive anchor?
2. Does any href on the site point at a redirect, a parameter variant, or a duplicate instead of the canonical live URL?
3. Is any internal link nofollowed on purpose, and can you defend the reason without using the word "sculpting"?
4. Does any important page sit more than three clicks from the homepage, or carry only one incoming internal link?
5. Does every button, paginated sequence, and filtered view a customer uses have a real <a href> behind it?
6. Has Search Console's Links sample moved for your target pages since you made the changes, after waiting a few weeks?

If all six answers are yes, the module is done. If any answer is no, you know exactly which lesson to reopen.

Where this sits next to bought links

Internal linking costs time and no money, which is why it comes first. None of the 70 providers ranked on the M1.3 Provider Buyer Score sells internal linking as a service, and that absence is the point: nobody can sell you what you already own. If the foundations of link building itself are new to you, read that guide first; when you are ready to earn outside citations, the complete backlink method is the next module. When the site has done this work and still needs external authority, the ranking of link building services shows how we score the firms that sell it, and how we score providers documents the criteria. Buy links only after your own house is linked.