Skip to content
BFRBUILT FOR RANK

Internal Linking for SEO: What Google Actually Uses It For

The two jobs internal links do according to Google's own documentation, the direction most publishing workflows get backwards, and a 30-minute audit you can run in Search Console.

SV
Stephen V

The Short Version

Internal linking is the only part of SEO where you control every variable, and it is the part most small business sites never touch after launch.

The reason it gets skipped is that the usual explanation for it — that links push authority around your site like water through pipes — is unverifiable. You cannot see it, measure it, or attribute a ranking change to it. So it sits on the checklist forever.

Google's own documentation makes a narrower and far more useful claim, and that claim is where the work should be aimed.

The Two Jobs Google Documents

Discovery. Google's SEO starter guide states that it "primarily finds pages through links from other pages it already crawled" and that "the vast majority of the new pages Google finds every day are through links." That is not a theory about authority. It is the mechanism by which a page you published becomes a page that exists in search.

Context. The same guide describes link text as "the text part of a link that you can see," and says the point of writing it well is that "users and search engines can easily understand what your linked pages contain before they visit." Every link you write is a short, voluntary description of the page on the other end.

Those two jobs are concrete, documented, and entirely within your control. The third job — ranking authority flowing between your own pages — is real in principle, and we are not going to pretend it can be audited on a forty-page business site. Build for discovery and context, and stop trying to measure the thing you cannot see.

The Direction Everyone Gets Backwards

Here is the part almost no guide mentions.

When you publish a new page, the links you add from that page do nothing to help it get found. Google has not crawled it yet. The link that matters is the one pointing at it, from a page Google already visits regularly.

Most publishing workflows only do the first half. The writer finishes the article, dutifully links out to three related pages, and publishes. Nothing links in. The page waits on the sitemap, which is not a path.

The content engine that publishes on this site treats the other direction as the actual deliverable. Every time it ships a page, it edits two or three existing pages to link to the new one, with anchor text written for the new page's subject, and records which pages it edited. That edit list is the work. The new page's own outbound links are a courtesy to readers.

If you take one thing from this guide: when you publish, go add the inbound links.

Google's documentation on crawlable links is blunt. "Google can only crawl your link if it's an <a> HTML element" with an href attribute, and it "can't reliably extract URLs from <a> elements that don't have an href attribute or other tags that perform as links because of script events."

The patterns it lists as not recommended are the ones modern website builders produce by accident:

  • A <span> or <div> with a click handler, styled to look like a link
  • An <a> tag with an onclick and no href
  • href="javascript:..." in place of a real path
  • Framework-specific link attributes that never render an href

In practice this shows up as clickable cards on a services grid, a location or filter menu built in JavaScript, and "load more" pagination. A visitor with JavaScript sees a working site. A crawler sees a page with no way out of it.

It is testable in about a minute per page: load the page, turn JavaScript off, and try to reach your service pages. Or open the URL Inspection tool in Search Console and read the rendered HTML. If those links are not <a href> elements, the structure you designed does not exist as far as search is concerned.

Anchor Text, by Google's Own Rules

Google's link documentation gives specific, quotable guidance here, which is rarer than it should be. Anchor text is "the visible text of a link," and "good anchor text is descriptive, reasonably concise, and relevant to the page that it's on."

Named as too generic: "Click here," "Read more," "website," "article." Named as too long: a whole sentence crammed into one link. The test Google suggests is the one worth keeping — "try reading only the anchor text (out of context) and check if it's specific enough to make sense by itself."

Two more rules from the same page that get broken constantly:

  • "Write as naturally as possible, and resist the urge to cram every keyword that's related to the page that you're linking to." The same exact-match anchor repeated across forty pages reads as what it is.
  • "Don't chain up links next to each other; it's harder for your readers to distinguish between links." Three links in one sentence are worse than one good one.

And when the link is an image, the alt attribute becomes the anchor text. An image link with an empty alt is a link with no description attached to it.

Your Sitemap Is Not a Substitute

This is the usual defense of an orphan page: it is in the sitemap, so Google knows about it.

Google's sitemap documentation says a sitemap "doesn't guarantee that all the items in your sitemap will be crawled and indexed." The same page describes when you can skip a sitemap altogether — if your site runs "about 500 pages or fewer" and is "comprehensively linked internally," because then "Googlebot can find all the important pages on your site by following links starting from the home page."

Read that the other way around and you have the standard worth building to. The sitemap is the backup. The link structure is the system.

The Search Console help for pages Google has found but not indexed puts it in one line: if parts of your site are missing, "make sure they are linked properly." Worth knowing, too, that "crawled — currently not indexed" is in Google's words "a common message for many sites" that "doesn't necessarily indicate an issue that needs attention." Not every unindexed page is a linking failure.

The 30-Minute Audit

You need Search Console and a browser.

  1. Open Search Console, then Links, then Internal links. That section covers "links within your own site to your own site" and shows "which of your pages is linked the most from within your own site."
  2. Sort to the bottom, not the top. The interesting pages are the ones with one internal link, or none. Google notes that "this report isn't a comprehensive list of every link on your site," so treat a low count as a flag to verify rather than a verdict.
  3. Write down every page that matters commercially — services, locations, pricing — and check each one appears. A service page that brings in money and has a single inbound link from the footer is the highest-value fix on your list.
  4. Reach your most important page from your homepage using only links. Count the clicks. If it takes four, your structure is telling visitors and crawlers that the page is minor.
  5. Check the anchor text you actually use. Search Console will not help here: its Top linking text table covers "backlinks to your site" from other sites, not your own links. Use your site's search, a crawler, or search your content files for the target URL and read the text around each hit.
  6. Fix the top five, not all forty. Add a contextual, descriptive link from the most closely related existing page. One real sentence inside the body copy beats a block of related links nobody reads.

Hub and Spoke, Using This Site as the Example

The structure that makes this maintainable is boring: a hub page per topic, spokes linking up to the hub, and spokes linking across to their closest siblings.

This site runs on it. More than 130 content pages sit under five hubs — services, industries, resources, blog and portfolio — and no hub list is maintained by hand. The resources, blog and portfolio hubs are generated from the content files themselves, so a page is listed the moment it exists. Services and industries come from one shared registry that also feeds the main navigation and the footer, so a page is registered once and appears in all three places or none.

That is the detail worth copying, not the design. A hub list maintained by hand goes stale the first time somebody publishes in a hurry, and that is how most orphan pages are born.

Two honest limits, since this is the part guides oversell.

Navigation links count, but they are not the good ones. The footer here lists eight services and eight industries out of several dozen such pages, chosen deliberately. A mega-menu that links everything from everywhere is not a structure, it is a flat list. The links that carry a description you chose are the ones inside the body text.

Breadcrumbs are free links, and most sites skip them. Every interior page here has a breadcrumb trail built from real <a href> links up to its hub, which is also the markup that can earn the visible breadcrumb trail in search results instead of a raw URL.

What Internal Linking Cannot Do

It will not rank a thin page. If the page does not answer the query, more links pointing at it produce a slightly better-crawled page that still does not rank.

It is also not a crawl budget exercise, whatever your tooling implies. Google's crawl budget guide opens by saying it is "an advanced guide intended primarily for" sites with a million or more pages changing weekly, or ten thousand or more changing daily, and that if your site does not have a large number of rapidly changing pages, "you don't need to read this guide." A small business site does not have a crawl budget problem. It has a discovery problem and a context problem, which are the two jobs above.

The Workflow That Keeps It From Rotting

Internal linking is not a project you finish. Every page published without inbound links is a new orphan, and one redesign can strip a year of them in an afternoon.

What works is making it part of publishing rather than a quarterly cleanup:

  • Hubs generated from content, never hand-maintained. A new page is listed the moment it exists.
  • Two or three inbound links added at publish time, written into the body of the most closely related existing pages.
  • A record of what got linked. We keep a ledger per site; a spreadsheet does the same job.
  • One link check after any redesign or URL change. The website migration checklist covers pointing internal links at final URLs instead of leaving them to redirect, which is where most of the damage happens.

On our plans this sits inside the monthly SEO content updates rather than being a separate line item — when we publish a page for a client, adding the inbound links from their existing pages is part of that work. What is included at each tier is on the pricing page.

If You Only Do One Thing Today

Open Search Console, find your single most valuable commercial page, and check how many of your own pages link to it. Then add one link to it from the page your visitors actually read, with anchor text that describes what they will get.

That is twenty minutes, it costs nothing, and it is the only SEO work where no competitor and no algorithm update can get in your way.

The links you don't control are a different and far more cynical market — we wrote up what Google actually calls link spam, and what two of our builds earned without buying a single link. For the full technical picture, the SEO audit checklist puts internal links in priority order against everything else. And if you would rather have someone map your structure and tell you where the orphans are, that is a free consultation.

Frequently Asked Questions

Google publishes no number, and any guide quoting one is quoting a tool's default. The useful question is the inverse: how many links point at the page, and from where. A page with zero inbound internal links is a real problem, because links are how Google finds pages. A page with four or five contextual links from closely related pages is in good shape. Beyond that you are adding links for their own sake, and Google's own link-text guidance warns about the result: "Don't chain up links next to each other; it's harder for your readers to distinguish between links." Write the page, link where a reader would genuinely want the other page, and stop.

Two of the three things people mean by that are documented by Google. First, discovery: the SEO starter guide says Google "primarily finds pages through links from other pages it already crawled," and that "the vast majority of the new pages Google finds every day are through links." Second, context: link text is described as the visible text whose purpose is that "users and search engines can easily understand what your linked pages contain before they visit." The third claim — that internal links pass measurable ranking authority between your own pages — is real in principle but not something you can audit, measure or attribute on a small business site. Build for the two documented jobs and the third takes care of itself.

A page on your site that no other page links to. It usually happens one of three ways: it was published outside the normal template so the hub page never picked it up, it was unlinked during a redesign and nobody noticed, or it only ever existed in the XML sitemap. The last case is the one people argue about. Google's sitemap documentation says a sitemap "doesn't guarantee that all the items in your sitemap will be crawled and indexed," and the Search Console help for pages Google has found but not indexed tells you to "make sure they are linked properly." A sitemap entry is a suggestion. A link is a path.

Partly, and the limits matter. The Links report has an internal links section — "These are links within your own site to your own site" — and a Top linked pages table that shows "which of your pages is linked the most from within your own site." That is enough to find the pages sitting at or near zero. Two caveats: Google says plainly that "this report isn't a comprehensive list of every link on your site," and the Top linking text table covers links from other sites rather than yours, so Search Console will not tell you what anchor text your own pages use. For that you need a crawl of your own site, or a search through your own content files.

Not for search purposes. Google's crawlable-links documentation says "Google can only crawl your link if it's an `<a>` HTML element" with an `href` attribute, and that it "can't reliably extract URLs from `<a>` elements that don't have an `href` attribute or other tags that perform as links because of script events." The patterns it lists as not recommended are exactly the ones modern site builders produce by accident: a `<span>` carrying an href, an `<a>` with an onclick handler and no href, and `href="javascript:..."`. Clickable cards, filter menus and load-more buttons are the usual offenders. They work for a visitor with JavaScript and are invisible as a path for a crawler.

No. A page you genuinely do not want in search results needs a noindex tag, which is the instruction built for that job; nofollow on your own internal links mostly just makes your site harder to crawl. The pages people usually have in mind — a thank-you page, a thin terms page — do no harm sitting in the index. And the authority-shaping theory behind the tactic rests on exactly the unmeasurable third job above. Spend the effort on giving your important pages more real inbound links instead of trying to starve the unimportant ones.

Want a website that actually ranks?

Get a free consultation: we'll look at your competition and recommend a plan to help you rank.

Results from sites we built: Mixed Bag Sportsman +19,700% Google impressions · Revival Coffee +15,889% Google impressions · Smoky Valley Distillery +10,377% Google impressions

Built For Rank is a Type 5 Marketing LLC company in Richardson, Texas. info@builtforrank.com