GOMAX ULTIMATE 5.53.0 adds an internal linking engine that reads your published content, works out which posts are most topically related using local TF-IDF similarity, and suggests exactly which internal links to add — plus it finds "orphan" pages that nothing links to — all computed on your own server with no external service and no AI key. Internal links are one of the most under-used SEO levers, and this makes them a two-minute job instead of a manual audit.
Search engines follow internal links to discover and rank pages, and links pass authority between related content. Yet most sites link almost at random. This engine turns your own content into a map and tells you where the missing links are.
Key takeaways
Relevant link suggestions. For any post, suggest_links returns the most topically related published posts to link to, ranked by a similarity score.
Smart anchors, no duplicates. Each suggestion includes a suggested anchor (the target's title) and skips targets you already link to.
Orphan-page finder. retrieve_orphans lists published pages that no internal link points to — the pages search engines tend to overlook.
Local and private. It uses TF-IDF content similarity computed on your own server; nothing is sent to an external service, and no AI key is required.
Foundation for more. The same statistical model underpins GOMAX's semantic recall for its AI features.
Why internal links matter so much
When a page has no internal links pointing to it, search engines struggle to find it and rarely rank it — it's effectively invisible no matter how good it is. And when related posts don't link to each other, the ranking authority they could share stays trapped. Fixing internal linking is one of the highest-return, lowest-effort SEO tasks, but doing it by hand means re-reading your whole site. The engine does that reading for you.
How it works
It builds a keyword profile of each published post and page — the words that make it distinctive, weighted by how rare they are across your site (that's TF-IDF). To suggest links for a post, it compares that post's profile to every other and ranks the closest matches by cosine similarity. Because it looks at topics, not just shared keywords, the suggestions are genuinely related, not coincidental.
For orphans, it checks which pages are referenced by internal links anywhere in your content and lists the ones that aren't — so you know exactly which pages need a link pointing to them.
A simple workflow
- Run retrieve_orphans to see which pages are invisible to internal linking. 2. For each orphan (or any post), run suggest_links to find the best related posts. 3. Add a link from those related posts to the target, using the suggested anchor or a natural phrase.
An AI assistant can do this end to end: find the orphans, get the suggestions, and edit the posts to add the links — turning a tedious audit into an automated task.
Frequently asked questions
Does it use an external service or an AI key? No. The similarity is computed locally with TF-IDF on your own content. There's no external API and no AI key required — it works entirely on your server.
What is an orphan page and why is it bad? An orphan page is one that no internal link points to. Search engines discover pages by following links, so orphans are hard to find and rarely rank, even if the content is excellent. Adding a few internal links fixes that.
Will it add the links for me automatically? It suggests, it doesn't auto-edit — so you stay in control. It gives you the target, the suggested anchor and the similarity, and you (or an AI assistant) add the link where it reads naturally.
How does it pick the anchor text? It suggests the target post's title as a starting anchor. You can use that or write a more natural phrase in the sentence where the link fits — descriptive anchors are best for SEO and readers.
Does it handle large sites? It analyzes your most recently updated published posts and pages up to a sensible cap, so it stays fast. For very large sites it focuses on the freshest content, which is usually where linking matters most.
Top comments (0)