Internal linking prompts only work if you paste in your real page list — titles, URLs, or a site structure export. A model with no knowledge of your site can’t suggest real links; a model given your actual page titles and hierarchy can suggest genuinely relevant ones.
Linking a New Article
- “Given this new article on [topic] and this list of my existing page titles [paste list], which 5 pages should it link to, and with what anchor text?”
- “Given this list of site pages [paste titles/URLs], which existing pages should link TO this new article on [topic]?”
- “Suggest 3 places in this pillar page [paste its card/section list] where a new article on [topic] should be added.”
Anchor Text
- “Does this article [paste] use generic anchor text like ‘click here’ or ‘read more’ that should be replaced with something descriptive?”
- “Suggest 3 different anchor text variations for linking to [target page title] from different contexts across the site.”
- “Is this anchor text [paste] over-optimized (exact-match keyword stuffed) in a way that looks unnatural?”
Finding Orphan Pages
- “Given this list of all my pages [paste] and this list of pages with at least one internal link pointing to them [paste], which pages appear to be orphaned?”
- “Given this page [describe/paste], what 3 existing articles on related topics could reasonably link to it?”
Pillar-Cluster Structure
- “Given these related article titles [paste list], suggest a logical pillar-to-cluster linking structure — which should be the hub, which are supporting articles?”
- “Given these two related articles [A] and [B], where should each link to the other, and with what anchor text?”
- “Suggest a ‘related articles’ set of 4 links for this article on [topic], prioritizing genuine topical relevance over recency.”
Guardrails
Never let the model invent a page title or URL that doesn’t exist on your site — it will confidently suggest a plausible-sounding link target that was never published. Always paste your real page list into the prompt and ask the model to choose only from it, not to generate suggestions from general knowledge of what a site “probably” has.
- Paste your real page titles/URLs into every internal-linking prompt — never let the model invent a link target.
- Ask explicitly for the model to choose only from the pages you’ve listed, not to suggest plausible-sounding pages that don’t exist.
- Fix generic anchor text (“click here”) in older articles as a quick, low-effort optimization pass.
Frequently Asked Questions
Will AI invent a link to a page that doesn’t exist?
Yes, if you don’t constrain it — always paste your real page list and instruct the model to choose only from it, then verify every suggested URL actually resolves before publishing.
How many internal links should one article have?
There’s no fixed number, but very high counts (upward of 100 links from a single page) can dilute link equity and hurt user experience — prioritize genuine relevance over quantity.
Can this replace a proper internal-linking audit tool?
No — a crawl-based tool gives you the actual current link graph; use these prompts on top of that data to decide what to change, not as a substitute for measuring it.
Further reading & sources
- Link best practices — Google Search Central
See how your site actually shows up in AI search. An AI visibility audit maps where you’re cited, where you’re invisible, and what to fix first — in plain English.
Get your AI visibility auditTry the free SEO tools →
Prefer self-serve? The interactive checklists turn guides like this one into a working to-do list.
Keep reading in AI SEO
Get one email when something genuinely changes
AI search moves fast and most of it is noise. We send one short email when a real shift is worth your time. Unsubscribe anytime.
Published by Plain Intelligence — practical AI SEO, GEO, and technical SEO, documented in plain English. About Plain Intelligence →
↑ Back to AI SEO · Explore all articles · Free tools & resources · Glossary