SEO Roadmap
Six steps, in sequence. Each one is a full guide, free and ungated, and each assumes the one before it. If you are starting from nothing, start at step one and work down.
The sequence
Most SEO reading is a pile of tactics with no stated prerequisites, which is why people learn keyword research before they can read a crawl report, and learn how Google ranks pages before checking whether their buyers use Google. These are ordered by what each one needs you to already know. Skipping ahead is allowed; it is just slower.
Where search actually happens: the five reasons people search, the six categories of surface that serve those reasons differently, and how to pick the two or three that are genuinely yours.
Why here: First, because everything after it assumes you know which surfaces your buyers use. Learn how Google ranks pages before you know whether your buyers use Google and you will optimize one surface out of six.
How search works from zero: crawling and indexing, how pages are chosen and ranked, technical foundations, AI search, and what a realistic timeline looks like.
Why here: Second, because web search is the largest surface on the map you just drew, and it feeds the AI layer. Everything after this assumes you can read a crawl and recognize a result that will take six months.
How to find what people actually search for and decide what is worth targeting: six stages from demand through to a map, the three metrics everybody misreads, clustering, and business value as the axis no tool has a column for.
Why here: Third, because research is the first step that produces a decision, and a decision needs the two steps above it. Do it earlier and you will target keywords the site cannot yet win, on surfaces you had not chosen.
What would actually satisfy the demand you just found: reading a results page to decide what kind of page deserves to exist, in what format, at what angle, and whether you could realistically be the best answer on it.
Why here: Fourth, because keyword research tells you what people search for and this tells you what would satisfy them. Skip it and you get the beginner failure the whole roadmap exists to prevent: keyword found, article written, nobody served.
Where the pages step four says should exist actually live: how many URLs a cluster really needs, what the parent is, what links to it, and how prominently the site treats it. Six decisions, and the deliverable is a site map with a verb on every row.
Why here: Fifth, because intent tells you what kind of page deserves to exist and this decides where it goes. Skip it and every page gets its parent assigned after it ships, which is how a redirect file gets written.
What actually goes on the pages step five said should exist: whether each one deserves to be written at all, what it will contain that a model could not have guessed, and how to keep it alive afterwards. The deliverable is a brief that can say no.
Why here: Sixth, because architecture decides where a page lives and this decides what is on it. It is also the last step that can still cancel a page cheaply, which is the whole reason it comes before writing rather than during it.
Prefer to hear the whole pass before you read? The free 5h 44m SEO course is the same roadmap taught in one continuous session, with each chapter deep-linked and mapped back to the step it belongs to. Watch the pass, then come back here for the depth.
Once the roadmap runs out, the writing continues in the blog: traditional SEO for the mechanics, AI SEO for visibility inside AI answers. The working files behind all of it are in free SEO resources.