To a search engine, a page with no links is a page that barely exists. Every new page needs a path to it, in the HTML the server delivers, and links going out of it to the pillar page and its sibling pages. The link ladder is the habit that guarantees this: when you publish, edit two or three older pages to point to the new one.
Goal: link each new page to the pillar page and its siblings, and edit 2–3 older pages to point to it, in the same commit.
01 · Who links to whom
Click each step.
Click each block in the diagram to see what it does.
The links that count are the ones in the HTML delivered by the server, without depending on JavaScript.
02 · For every new page
Tap each step.
1. Outgoing links
In the body, 2 to 4 links with descriptive anchors. At the end, a "Read also" block with the pillar page and the siblings.
All links return 200 (the X-ray from course 1 checks this).
2. Ladder
Pick 2 or 3 old pages on the same topic where the link makes sense for the reader, and add one sentence with a link.
The topic's pillar page always gets the link to the new page.
3. Same commit
The new page and the ladder edits go up together. "I'll link it later" is how orphan pages are born.
4. Check the HTML
Run the X-ray on the new page and the old ones: the "internal links in HTML" line shows whether the links are in the server's HTML.
A catalog with in-browser search also needs links in the HTML: a collapsed index (details) solves it without changing the look.
At INEMAOn 09/30/2026, the courses page became "search only" and 310 listing pages lost the only internal link rendered on the server. The fix was a collapsed index, with no visual change.
03 · Link map, ladder block, and the check
Write down the map before you publish.
Link map for a new page
New page: <url>
Pillar page (parent): <url>
Siblings (2–4): <url> · <url>
Catalog (2): <url> · <url>
Ladder (2–3 old pages that will now link to the new one):
- <old url 1>: sentence "<anchor text>"
- <old url 2>: sentence "<anchor text>"
- pillar page: entry in the topic list
tools/auditoria.sh https://www.seusite.com /pagina-nova/ /antiga-1/ /antiga-2/
# "links internos no HTML" line for each page
# and, on the old one, look for the address of the new one:
curl -s https://www.seusite.com/antiga-1/ | grep -c 'href="/pagina-nova/"'
Example from the repository
$ tools/auditoria.sh https://www.inema.club / /cursos/
-- / [200]
links internos no HTML: 27
-- /cursos/ [200]
title: Cursos de inteligência artificial, agentes e automação | INEMA.club
links internos no HTML: 335
With the collapsed index, the courses page delivers the links in the server's HTML, even though it's a search in the browser.
"Click here": the anchor doesn't say what the page is about.
Links only through JavaScript: the crawler may not see them.
"Other sites in the network" block on every page: a link between sites is contextual and rare.
04 · Practice: link your page
Use the question page from lesson 4.
E1 · Link map
20 min
Fill in the link map for your new page: pillar page, siblings, catalog, and the 2 or 3 old pages for the ladder.
Done when: the map has a pillar page, at least 2 siblings, and 2 chosen old pages, each with the anchor sentence written out.
E2 · Ladder published
30 min
Publish the new page and the ladder edits in the same commit. Run the X-ray and the grep on the old pages.
Done when: the grep finds the new page's address in each old page of the ladder, and the X-ray shows internal links in the HTML of the new page.
05 · Test what you learned
The explanation shows up after your answer.
Question 1 of 4
What is the link ladder?skill
Correct
It's the habit that keeps the new page from being born an orphan.