Skip to content
ROI Synth

How to run an internal link audit, starting from the pages that matter

By ROI SynthPublished 6 min read

An internal link audit checks whether the links between your own pages help people and search engines reach the pages that matter most. Start by ranking your pages by business value, then crawl the site and compare: do your most important pages receive links from relevant content, are any pages orphaned or buried deep, and do links point at broken or redirected URLs? Fix in that order, using descriptive anchor text, and re-crawl to confirm.

On this page
  1. What internal links actually do
  2. Rank your pages before you crawl anything
  3. Crawl the site and separate content links from navigation
  4. The five problems worth fixing, in order
  5. A worked example (hypothetical)
  6. Verify the fixes and keep it from drifting
  7. Questions
  8. Sources

Most internal link audits start with a crawler export and end with a spreadsheet of thousands of rows that nobody acts on. The problem is not the tool. It is that the audit never decided which pages deserve links in the first place, so every finding looks equally urgent.

This method reverses the order. It starts with a short list of priority pages and measures everything against it. It works with any crawler, and on most small and mid-sized sites the whole audit fits in an afternoon.

What internal links actually do

Internal links do three jobs at once. They let visitors move from a general page to the specific one they need. They let search engines discover pages and understand how they relate. And their anchor text tells both what the destination is about.

Google's documentation on crawlable links is precise about the mechanics: it can reliably follow links that are HTML anchor elements with an href attribute. Links created only by JavaScript click handlers, or buttons that change the page without a real URL, may not be followed. On JavaScript-heavy sites it is worth checking this first, because a page reachable only through such controls can be found late or not at all.

Rank your pages before you crawl anything

Write down the 5 to 20 pages that matter most commercially: the service or product pages that lead to enquiries, the pricing page, and the guides that bring in qualified visitors. Give each one a priority from 1 to 3. This list is the yardstick for every later finding.

If you have Search Console, add two columns: impressions over the last three months and the queries each page appears for. A priority page with impressions but few clicks is often under-linked or competing with another page for the same queries.

Crawl the site and separate content links from navigation

Use any crawler that exports internal links with their source page, destination, anchor text and status code. Then make one adjustment most audits skip: separate links in the main content from links in the header, footer and menus.

Navigation links appear on every page, so they inflate the count for whatever sits in the menu and say little about relevance. A link from a paragraph on a closely related page is a much stronger signal. Counting both together makes a page look well supported when nothing in the content actually points to it.

What to record for each priority page
MeasureHow to get itWhat a problem looks like
Links from contentCount inbound links whose source sits in the main content areaZero or one, for a page you want to rank
Click depthShortest number of clicks from the homepageFour or more for a priority page
In the sitemap?Compare crawled URLs with your XML sitemapIn the sitemap but not reached by the crawl (orphan)
Anchor textList the anchors pointing to the pageMostly "click here", "read more" or the bare URL
Status of inbound linksStatus code of each linked URLLinks pointing to a 404 or a redirect instead of the final page

The five problems worth fixing, in order

Work through these in sequence. Each one is cheaper to fix than the next and usually affects more pages.

  • 1. Broken internal links

    Links to pages that return 404. Point them at the correct live page, or remove them if nothing equivalent exists.

  • 2. Links through redirects

    Links to an old URL that redirects. They still work, but each one adds a hop. Update the link to the final destination.

  • 3. Orphaned pages

    Pages in the sitemap that no crawled page links to. Either link to them from the most relevant content, or ask whether they should exist at all.

  • 4. Under-linked priority pages

    Important pages with no links from related content. Add links from the pages that discuss the same topic, where a reader would genuinely want to go next.

  • 5. Vague anchor text

    Replace generic anchors with words that describe the destination, in natural language. "Our technical SEO audit" tells the reader and the crawler more than "learn more".

Resist two temptations: adding a block of links to every page, and repeating the exact same keyword-rich anchor dozens of times. Both make pages worse for readers and add little information.

A worked example (hypothetical)

A firm with a 60-page site lists eight priority pages. The crawl shows that its most valuable service page has 52 inbound links, which looks healthy, but all 52 come from the header menu. Not one paragraph on the site links to it, including the three guides that explain the problem the service solves.

The fix takes an hour: one contextual link from each of those three guides, with an anchor that names the service, and a link from the pricing page where the service is listed. Two older blog posts that linked to a retired URL are updated to the live page. The re-crawl shows four content links instead of none, and Search Console's links report reflects them over the following weeks.

Verify the fixes and keep it from drifting

Re-crawl after the changes and compare the five measures with your first pass. Check a few fixed pages with URL Inspection in Search Console to confirm that Google sees the links in the rendered HTML, which matters on sites that build pages with JavaScript.

Then make it routine. A quarterly crawl is enough for most sites, with a quick check for broken links after any redesign, migration or bulk content change. On a coded website the most durable fix is in the templates: related-page links generated from your content data, so new pages are linked correctly the day they go live.

If you want this done properly on a coded site, our internal linking service covers the audit and the template changes, and the free website audit checks the links on any single page.

Questions people also ask

How often should I audit internal links?

Quarterly for most sites, plus a check after any redesign, migration or large content change. Sites that publish daily benefit from a lighter monthly check for broken links and orphans.

How many internal links should a page have?

There is no fixed number. What matters is that important pages receive links from relevant content and that every link helps a reader. A page with three well-placed contextual links is better supported than one with fifty menu links.

Do footer and menu links count?

They help discovery and navigation, but because they repeat on every page they say little about how pages relate. That is why this method counts content links separately.

What is an orphan page?

A page no other page on the site links to. Search engines may still find it through the sitemap, but it receives no context from related content, and visitors cannot reach it by browsing.

Should internal links use exact-match keywords?

Use descriptive, natural anchors. Naming the destination clearly is helpful; repeating the identical keyword phrase on every link is not, and it reads badly.

Where ROI Synth can help

Sources

Search engines change their documentation and features. Facts were checked against these sources on 14 November 2026.

Talk to ROI Synth about your website

Share where your site stands and what you want more of. We will tell you plainly which work is worth doing first, and what it would cost.