Site architecture and keyword mapping

This is the module people most want to skip. Nothing visible happens. You do not publish anything, you do not see a ranking move, and at the end of it your site looks exactly the same as when you started.

It is also where the most expensive mistake in the curriculum gets made, quietly, and stays invisible for about a year.

One page per intent. That is the rule.

Module 3 ended with a list of terms grouped into rows. This module attaches each row to a URL, and enforces one thing: no two pages may claim the same intent.

When they do, Google has to choose. It chooses on signals that have nothing to do with which page you would have picked, it sometimes changes its mind between crawls, and the authority you have earned gets split across both. That is keyword cannibalisation, and the reason it goes unnoticed is that nothing looks broken. Both pages exist. Both are indexed. Both rank at position 14 instead of one of them ranking at 6.

I can point at my own version of this. The reason module 1 exists at /diy-seo/is-seo-worth-it while a tool sits at /tools/is-seo-worth-it is that I had five URLs on this site chasing the same cost-and-ROI questions. Between them they produced one click in sixteen months. Fixing it meant redirecting two of them into a single page, which is exactly the decision this checklist asks you to make.

You do not have a crawl budget problem

Somebody will tell you to worry about crawl budget. Google has published who should:

If your site doesn’t have a large number of pages that change rapidly, or if your pages seem to be crawled the same day that they are published, you don’t need to read this guide.

Their own threshold for the advanced guide is large sites of a million or more unique pages, or medium sites above ten thousand pages with content changing daily. For those sites Google recommends exactly two things: keep your sitemap current and check the Page Indexing report.

If you have forty pages and something is not indexed, the cause is not budget. It is eligibility, and that is a module 2 diagnosis.

Links Google can actually follow

Internal links are how authority moves around your site and how Google works out which of your pages you consider important. There is a mechanical requirement that catches modern sites out, and Google states it directly: “Google can only crawl your link if it’s an <a> HTML element with an href attribute.”

Which means a <span> with a click handler is not a link. A framework component that renders routerLink without an href is not a link. A button that navigates via JavaScript is not a link. They all work perfectly for a human and are invisible as paths through your site.

On anchor text, Google is equally plain: good anchor text is “descriptive, reasonably concise, and relevant to the page that it’s on and to the page it links to.” Which is the documented reason to stop writing “click here” and to stop letting a related-posts widget do your internal linking. A widget picks by recency or by tag. A sentence picks by meaning, and only one of those is a signal.

Have us map the structure before you build it

URLs: what matters and what does not

Less than people think. Google’s guidance is short: keep the structure simple and logical, “use readable words rather than long ID numbers,” and “use hyphens (-) instead of underscores (_) to separate words in your URLs.” That is close to the whole of it.

What is not in there: keyword stuffing the slug, matching the URL exactly to the target term, or restructuring a working site to get a keyword one level higher in the path. Those are folklore, and the last one is actively expensive.

The rule I would hold you to: do not change a URL that ranks unless the change fixes something real. Merging two competing pages is real. A tidier path is not. Every URL change is a redirect you have to get right, and a redirect you get wrong is authority you built and then threw away.

When you do have to change them, map every old URL to the specific new page that replaces it, before you ship, in a spreadsheet. Never redirect a batch of old URLs to the homepage. Google treats that as a soft 404 and you get the loss without the transfer.

Pillar pages, without the jargon

A pillar page is a page that covers a broad subject and links out to more specific pages underneath it, each of which links back. That is the entire idea. It is worth doing for one reason: topical authority is easier to build when your coverage of a subject is visibly organised than when it is fifteen unconnected posts.

You are reading one. The hub covers DIY SEO broadly and links to fifteen modules; every module links back to it and to its neighbours. If it works I will be able to show you the numbers in a year, and if it does not I will say so.

What makes it work is not the diagram. It is that each spoke genuinely answers a different question. Three spokes covering overlapping questions is a cluster with a cannibalisation problem and a nicer chart.

What I would do in your position

Run the sitemap tool, export your URLs, and put the keyword rows from module 3 beside them. Spend the first hour only on the conflicts, because that is where the compounding loss is. Resolve each one by picking a winner, not by rewriting both.

Then spend the second hour adding internal links from pages that already get traffic to pages that should. That is the cheapest ranking work available to you, it needs no new content, and on most small sites it is genuinely undone.

Then stop. Do not restructure. Do not renumber your URLs. Go and write something, which is phase 2, and come back here when you have enough pages for the structure to matter.