🔍 Internal linking: how to turn links within a site into a working system

Internal links look boring next to external promotion: you don't have to beg for them, nobody pays for them, and they already sit inside your site. That's exactly why editorial teams often hand their traffic to neighbors, even though it's easier to start with your own structure and add external distribution later through a guest posting marketplace. In this piece, we'll treat internal linking not as a one-time task before a release, but as a workflow: with an audit, priorities, anchors, and a clear criterion for when to stop.

How to build an internal link map in one evening

💡 Quick overview:

• Step 1: Export the full list of site URLs from your sitemap and from the indexing report, then combine them into one spreadsheet.
• Step 2: Highlight the pages that bring in money or leads in a separate color: these are your priority targets.
• Step 3: Crawl the site and count incoming internal links for each URL.
• Step 4: Put everything with zero incoming links and everything deeper than three clicks from the homepage into a separate list.
• Step 5: For each priority page, find three topically relevant donor pages and add links with descriptive anchors.
• Step 6: After a month, compare crawl coverage and rankings, and keep only what actually moved.

Repeat this cycle: it requires no development access, no budget, and no approvals. All you need is a spreadsheet and half an hour of discipline per week.

Three jobs one link does

An internal link rarely does just one thing. It solves three different tasks at once, and confusion between them causes most of the contentious decisions in teams.

The first job is navigational: the reader finishes a paragraph, a question comes up, and the link answers it without going back to the menu. The second job is semantic: the anchor and surrounding text tell the search engine what the page on the other end is about. The third job is distributive: links determine which sections of the site the crawler treats as central and which as peripheral.

Note an important detail: the three tasks almost never coincide at the same point. A footer link helps navigation but says almost nothing about meaning. A link in the middle of a long article, by contrast, conveys context well, but a smaller share of readers will see it. A good structure combines both types rather than choosing between them.

The practical takeaway is simple: before adding a link, answer which of the three jobs you're assigning to it. If none, the link is unnecessary, and removing it makes the page cleaner.

Orphan pages: where half your work quietly disappears

A piece of content without a single incoming internal link lives in a strange status. Formally it's published, physically accessible, sometimes even included in the sitemap, but for the search engine it hangs apart from the main fabric of the site. Ahrefs in its guide on how to find and fix orphan pages describes a typical scenario: pages get lost during a template change, when an old category is closed, or after a blog moves to a new address.

The scale of the problem is easy to underestimate. According to Ahrefs' December 2023 study covering about 14 billion URLs, 96.55% of pages get no traffic from Google at all, and another 1.94% get between one and ten visits per month. A significant share of that mass consists of pages that nobody inside the site connected to the rest of the content.

How to find orphans without expensive tools: export URLs from the sitemap, separately export the list of pages the crawler found by following links from the homepage, and subtract the second from the first. The difference is your candidates. After that, decide manually: return live content to the structure, merge outdated content with a stronger piece, and remove completely empty pages while setting up a redirect.

A rule that saves time: don't rescue everything. A page deserves incoming links if it has its own query, its own value, and a place in the overall topic of the section. Otherwise you're just spreading the crawler's attention across content nobody needs anyway.

The anchor says more than the link position

Google in its guide to crawlable links, updated in December 2025, states the requirement directly: the system only follows an element with an href attribute, and the anchor should be descriptive, concise, and relevant on both pages. A separate point recommends avoiding phrases like "click here" and not stuffing anchors with keywords.

In practice this translates into a few simple habits. Write the anchor so it makes sense out of context. Don't repeat the same link text across a dozen different URLs: that confuses both the reader and the algorithm. Spread links throughout the text instead of stacking a block of five in a row, because each one then gets its own semantic context.

A separate trap: trying to manually control weight by blocking links from indexing. For internal structure this is almost always wasted effort: it's simpler to remove the link itself if it's not needed than to keep it and hope an attribute will handle it.

Click depth and crawl budget

The deeper a page sits, the less often both people and crawlers reach it. Semrush in its breakdown of common internal linking mistakes from August 19, 2025, offers a guideline worth keeping in mind: important pages should be no more than three clicks from the homepage.

The topic of crawl budget deserves a reality check: it doesn't apply to most sites. Google in its guide to managing crawl budget, updated July 22, 2026, explicitly defines the document's audience: large sites with over 1 million unique pages updated at least once a week, and medium sites with over 10 thousand pages with daily changes. If your blog is smaller, saving crawl budget is pointless, but cleaning up depth and links is still useful.

What actually helps at any scale: short paths to revenue. Make a list of pages that bring in leads and count how many clicks from the homepage each one takes. Anything beyond the third level should be lifted up with links from stronger content. This is cheaper than redoing the menu and noticeably faster.

Real case: two similar projects, different results

Semrush provides a telling example in the same article about nine internal linking mistakes: two startups with comparable domain authority worked in the same niche. The first had a well-built internal linking structure, while the second had accumulated thousands of link problems. The first project held 8% of all its target keywords in the top of search results and collected about 8.6 thousand organic visits per month, while the second showed 6.3% of keywords in the top and roughly 1.9 thousand visits.

The traffic gap turned out to be more than fourfold with nearly identical starting conditions. The difference accumulated not from one heroic link, but from a sum of small things: some content remained without incoming connections, priority pages sat too deep, anchors repeated themselves and conveyed nothing specific.

From this case it's easy to extract a working priority logic. First fix what isn't connected to the site at all. Then lift important pages closer to the homepage. Then rewrite anchors on priority topics. Only after that does it make sense to argue about the finer points of weight distribution.

Mistakes that repeat from project to project

• Links for the sake of quantity. A quota like "five links in every article" turns posts into a maze and hurts readability.
• One anchor for different URLs. The reader stops understanding where the link leads, and the search engine gets a contradictory signal.
• A "read also" block instead of context. An automated list at the bottom is almost always weaker than one relevant link in the middle of a paragraph.
• Internal linking only for fresh content. Old articles accumulate authority for years and serve as the best donors, but people forget to open them.
• Links to pages with redirects. Redirect chains appear after migrations and quietly eat away part of the effect.
• No review process. Structure breaks with any template change, so audits need to happen on a schedule, not on a whim.

Separately, note a habit that looks harmless: adding a link right after publication and never returning to it. The topic changes, new content appears, and the old URL stops being the best answer. Reviewing links in key articles once a quarter costs little and keeps the structure alive.

⁉️🤔 Frequently asked questions

How many internal links should I put in an article?

There's no universal number, and quotas here do more harm than good. Follow the logic of the text: if a reader has a question in a specific paragraph that another piece of your content answers, the link is appropriate. If there's no such question, the link will get in the way. A long article naturally produces more connections than a short note.

What matters more: a link at the beginning of the text or at the end?

Position matters less than the context around the anchor. A link in the middle of a substantive paragraph where the topic is developed works better than a formal block at the bottom of the article. That said, links in the upper part of the text get more clicks simply because a smaller share of readers makes it to the end.

Should I block some internal links from indexing?

For a typical site, almost never. If a link isn't needed by the search engine, it usually shouldn't be shown to the reader either, which means it's simpler to remove the link itself. Attributes make sense for user-generated content and paid placements, but not as a tool for manually distributing weight within a site.

How quickly will internal linking show results?

The timeline depends on how often the crawler visits your site. On active projects, changes in crawling become visible within a few weeks; on rarely updated ones, the process stretches out. A reasonable evaluation horizon: a month for changes in crawl coverage and a quarter for rankings, provided the page content actually answers the query.

Does internal linking help if the content is weak?

No, and this is an important limitation. Links redistribute attention but don't create value. If a page doesn't answer the query better than competitors, an influx of internal links will lift it briefly and without lasting results. Content first, then the link structure around it.

What should I do with old articles nobody reads anymore?

Split them into three groups. Update content with unique value and bring it back into the structure with links from fresh articles. Merge pieces that duplicate each other into one strong article and set up a redirect. Remove content that has completely lost its purpose: empty pages dilute the site's topic and distract the crawler.

Summary

Internal linking wins not through clever schemes, but through consistency. Three things deliver almost the entire result: no important page without incoming links, no priority page deeper than three clicks, and no anchor that says nothing. Everything else is fine-tuning, which you should move to once the basics are in order.

Start small: take one section of the site, run it through the checklist steps at the beginning of this article, and write down what you find. Most likely you'll discover a few orphans and a couple of priority pages at the fourth level of depth. Fix this in one section, evaluate the result after a month, and only then scale the approach to the whole site. If you want to expand reach beyond your own audience, the next step is to arrange publications on external platforms, but the foundation is still laid inside your own link structure.