SEO · Playbook

Internal linking: the free lift nobody ships

Two passes per quarter, no new content, measurable rank lifts within six weeks. The approach we run on every Growth account, the rules we follow, and a small planner tool that scores a proposed link against the rules before you add it.

Internal linking is the most reliably under-used ranking signal in the modern SEO toolkit. It costs nothing. It does not require new content. It does not require approval from legal or brand. It produces measurable rank movement in the six week window after it ships. And most content teams still do not run it as a deliberate process, because internal linking is nobody's job description and it does not show up on a content calendar.

Why it still works in 2026

Three reasons, briefly. First, internal links remain one of the clearest signals Google uses to distribute crawl priority and ranking authority within a domain. That has not changed across any of the algorithm shifts of the last five years. Second, most libraries have never had a deliberate pass, so the baseline is unusually low and the upside is correspondingly high. Third, LLMs that index sites follow internal links the same way Google does, which means an internal linking pass that surfaces your top 20 posts lifts both your classic rankings and your AI citation surface at the same time.

The two passes

We run two internal linking passes per quarter on every Boost account. They are deliberately small and deliberately different.

Pass one: the hub pass. Identify the 10 to 20 target posts you want to lift over the next quarter. For each, find every other post in the library that (a) is relevant at the topic level, (b) has higher or equal domain authority or backlink count than the target, and (c) does not already link to the target. Add a single contextual link from each source into the target, with anchor text that matches the target's head query (or a close variant). Stop at three to five incoming links per target in a single pass; more than that in one window can look artificial.

Pass two: the expansion pass. Six weeks after pass one, rerun the analysis. Add links from newly published posts into the targets. Add links between the targets themselves where the topical relationship is clean. Remove any links from the first pass that did not render well (anchor too generic, context awkward, paragraph rewritten since). The second pass is the maintenance loop that keeps the graph healthy.

The rule we follow: every link we add is a link a human reader would genuinely want to click in context. If the link does not pass that test, it does not get shipped, regardless of what the SEO case looks like.

Rules for anchor text

Anchor text is the single lever within an internal linking pass that moves rankings the most, and it is also the lever most teams get wrong. The rules we follow:

  1. Match the head query intent, not the exact head query. An anchor that reads "how mortgage pre-approval timing works" outperforms a brittle "mortgage pre-approval" across the target's long-tail queries, because it signals intent more specifically without over-optimizing on the exact phrase.
  2. Vary the anchors. Three incoming links with three different natural-language anchors are worth more than three incoming links with identical anchors. Google's anti-spam filters have rewarded anchor diversity for over a decade.
  3. Avoid "click here" and "learn more". These are wasted anchor slots. They transmit no topical signal. If a link is worth placing, it is worth writing a real anchor for.
  4. Stay descriptive, not promotional. "Our guide to X" is weaker than "a walkthrough of X". The second reads as editorial; the first reads as house-ad. LLMs and Google both prefer editorial framing.

Rules for placement

Internal links are not all equal in weight. Where on the page the link sits matters almost as much as what the anchor says.

  1. Body links beat sidebar links. A link in a paragraph carries measurably more ranking signal than the same link in a sidebar module or related-posts strip. We concentrate effort on in-body links.
  2. The first two paragraphs of the body are premium real estate. Links in the introduction carry more weight than links at the bottom of a post. This is counterintuitive to teams that treat links as a citation-style reference, but the data has been consistent on this for years.
  3. One link per paragraph, maximum. More than one link in a paragraph dilutes the signal and starts to look like a link farm pattern. If you have two links that want to live in the same paragraph, split the paragraph.

A small planner

Below is a planner we use internally to score a proposed link against the rules above before we ship it. Paste the source and target URLs, the anchor text, and the paragraph the link would sit in. It returns a score and the specific rules the proposal passes or fails.

> Internal link planner

Score a proposed link

Scores the proposal against anchor and placement rules. Fully client-side; no URLs are sent anywhere.

How to pick the targets

A useful mental model: the posts that benefit most from an internal linking pass are the ones that are close to a ranking threshold but not past it. Posts ranking at positions 5 to 12 have the most to gain from three extra internal links. Posts ranking at position 1 have almost nothing to gain; they are already at the ceiling for their current signal profile. Posts ranking outside the top 30 usually have a different problem (thin content, no backlinks, wrong intent) that internal linking cannot fix alone.

So the target list for a pass is: posts at positions 5 to 12 on their head query, with impressions trending up over the last 90 days, that are plausible candidates to move to positions 1 to 5 in a quarter. Pull this list from Search Console in 10 minutes. Everything else flows from there.

What to expect

On a healthy library with a deliberate pass shipped well, the typical 90 day outcome is: tracked target keywords move up by 2 to 6 positions on average, impressions on the target URLs rise 20 to 50 percent, click-through rates rise in line with the position gains, and the pass does not regress other posts because the sources were chosen to have enough authority to spare.

On a library where the pass is poorly anchored (generic "click here" style anchors) or over-concentrated (five links added to a single target in one day from five low-authority sources), the outcome is flat or mildly negative. The same rules that make internal linking free when done well make it costly when done poorly. It is worth running the pass deliberately or not at all.

The process, in one paragraph

Pull the target list from Search Console. For each target, find the three to five best source candidates using the rules above. Write the anchors. Find the paragraphs. Ship the edits in a single batch. Wait six weeks. Measure. Run the expansion pass. Repeat. That is the entire loop, it is free, it compounds, and it is the single most reliable lever in SEO that nobody on the content team is paid to run.