Learn How to Do SEO by Finishing Six Units, Not Six Courses

Why courses fail and units work

Most SEO courses are tours. You watch someone else's screen, recognise the interface, and finish with the feeling of competence and no change to any site. Recognition fades in about a week, which is why people re-watch the same modules every year and still cannot explain why one page outranks another.

A unit is different. It has one idea, one task and one pass test. You learn what a canonical tag does, then you find three broken ones on your own site, then you prove it by re-crawling. The idea is attached to an action and a number, and that combination is what survives.

The second difference is order. Units are sequenced the way search actually processes a page: find it, understand the query, match the content, connect it to the site, keep it reachable, then measure. Learning out of order is the main reason beginners drown in technical jargon before they have written a single title.

The track: six units, one task each

Every unit below has the same three parts. Learn is what you read or watch, about an hour. Practice is what you ship on your site, about an hour. Done when is the standard that stops you from moving on too early. The tasks are deliberately small; the point is finishing, not impressing anyone.

1. 1

   How search finds and ranks a page

   Unit 1 · 2 hours · crawl, render, index, rank

   Learn

   What happens between a URL and a ranking, why a page can be indexed and still never rank, and what sitemaps and canonical tags actually change.

   Practice

   Take twenty URLs from your site. Log for each one: indexed or not, which canonical is used, and which title appears in the results.

   Done when

   You can show a twenty-row table with an index status and a reason for every exclusion.

2. 2

   Keyword research that maps to pages

   Unit 2 · 2 hours · seed, cluster, intent, page

   Learn

   How to cluster a seed list, how intent types change the page you need, and how to tell when a query deserves its own URL instead of a section.

   Practice

   Build a thirty-query map for one category: query, intent, target URL, and a flag for queries with no page yet.

   Done when

   Every query in the map points at one URL, and no URL is trying to serve two intents.

3. 3

   On-page: turning a query into a page

   Unit 3 · 2 hours · title, H1, opening, proof

   Learn

   The five positions on a page, the length and shape rules for each, and how a proof block turns a claim into something a buyer can check.

   Practice

   Rewrite titles and H1s on five pages, rewrite one opening to forty to seventy words, and add one spec or comparison block.

   Done when

   Five pages pass a five-line checklist, and a colleague can name the query each one targets.

4. 4

   Internal links and site shape

   Unit 4 · 2 hours · depth, anchors, orphans

   Learn

   How link equity moves through a site, why depth matters more than total link count, and how to write anchor text that describes a destination.

   Practice

   Add three contextual links to five pages, remove one dead-end link per page, and list every orphan URL a crawl finds.

   Done when

   No revenue page sits more than three clicks from the homepage.

5. 5

   Technical health you can check yourself

   Unit 5 · 2 hours · status codes, canonicals, speed

   Learn

   Status codes, redirect chains, canonical conflicts, parameter duplicates, and the three speed metrics that actually affect a page.

   Practice

   Crawl two hundred URLs from your own site, fix five errors that block indexing, then re-crawl and compare the count.

   Done when

   The error count drops and you can explain each fix in one sentence without notes.

6. 6

   Measurement that survives a review

   Unit 6 · 2 hours · freeze, segment, attribute

   Learn

   Impressions, clicks and position, why site averages hide results, and how to attribute a change to a specific edit without guessing.

   Practice

   Freeze twenty queries, build a one-page before and after report, and write one sentence naming the change behind the difference.

   Done when

   The report shows movement on the same twenty queries and states which edit caused it.

The unit table: what you keep from each

If you only remember one row per unit, remember the artefact. Each unit should leave a file behind, because the file is the evidence that you did the work and the reason the next unit is easier.

The three-layer self-test

Before you call a unit finished, test yourself in this order. Each layer is harder than the last, and skipping to the third without the first two produces people who can run a tool and cannot diagnose a site.

If you can explain but not judge, you have vocabulary. If you can judge but not do, you have taste. Only the third layer changes a site, and it is the one most courses never test.

Pacing that fits around a job

One unit per week is the pace that works for most people, with a two-hour block for practice and a short review the following Monday. Six weeks in total, and the site has changed six times. If a week gets lost, repeat the unit rather than skipping it, because units four to six assume the artefacts from the earlier ones.

1. Weeks one and two: understand the ground. Crawl and keywords. These are the slowest units and the ones people try to skip. Do not.
2. Weeks three and four: change pages. On-page and internal links. This is where the site visibly improves and where most of the practice hours go.
3. Week five: remove blockers. Technical health. Run it after links, because you will have created new URLs and new paths to check.
4. Week six: prove it. Measurement. Without this unit you cannot tell whether any of the previous five worked, and you will repeat the same mistakes next quarter.

Four ways learners stall

• Watching without shipping. Four hours of video and not one edit. The unit is not finished until a page changes.
• Choosing a site they cannot edit. If the CMS blocks title fields, the learner cannot practise. Use a small site you control, even a staging copy.
• Skipping measurement. Without a frozen query set, every later claim is a feeling. Unit six is the one that makes the other five cumulative.
• Learning tools instead of tasks. Knowing which button exports a report is not SEO. Knowing which five pages to fix this week is.

Where we come in

Some units are easier with a reviewer. We run the track with in-house teams, mark the artefacts, and take over the parts that need a specialist, such as a full keyword map or a template-level fix. You keep the files and the reasoning either way.

1. Unit reviews. You ship the task, we mark it against the pass test and return written notes, usually within two working days.
2. The keyword map, done with you. For catalogues with hundreds of categories, our Keyword Research Services team builds the map in unit two so units three and four have a target.
3. Template and copy support. When the fix belongs in the template rather than in a page, On-Page SEO Services handles it while your team continues the track.
4. Handover. At the end you own the crawl baseline, the query set and the reporting template. No dependency on us to keep doing it.

What one learner shipped in six weeks

A two-person export team took the track with no prior SEO experience. Week one produced a twenty-URL status table that showed eleven pages excluded by a leftover noindex from a staging migration. Nobody had noticed for seven months, because the pages still appeared in the internal search.

Weeks two to four produced a thirty-query map, forty rewritten titles and H1s, and 120 contextual internal links across six categories. Week five found eight redirect chains and two canonical conflicts, both introduced by a plugin update. Week six produced a one-page report on twenty frozen queries.

The numbers moved modestly in the first quarter, which is normal. The bigger change was internal: the team could now say which pages were eligible, which query each page targeted, and what had changed since last month. That is what finishing the units buys you, and it is the part that keeps working after the course ends.