M01DONE
Baseline and crawl
Every URL, its status code, its canonical, and whether it is indexable. Plus a starting point for impressions and revenue.
Acceptance readyWebsite kit · six modules
How to Do SEO for Website: Six Modules, Six Proofs of Done
To do SEO for a website, break the job into six deliverable modules — baseline, keyword map, on-page templates, content, structured data and measurement — and give each one an output plus an acceptance test. Modules beat to-do lists because they tell you when a piece of work is finished, not merely started, and they let a small team ship in weeks rather than quarters.
Built for catalogues of 50 to 5,000 URLs · One owner per module
?
What is the most practical way to do SEO for a website?
Split the work into six modules that each produce something a non-SEO colleague can check: an inventory, a query-to-URL map, template rules, written pages, structured data and a reporting baseline. Every module ends with an acceptance test, so "done" is a fact rather than an opinion, and the next module starts from something solid.
01 / why modules
A to-do list hides progress. A module makes it visible.
Most SEO plans fail in the middle, not at the start. Someone writes "improve titles and content" on a slide, three people interpret it differently, and eight weeks later nobody can say what shipped. Modules fix that by tying each body of work to one artefact and one check. The artefact is reviewable; the check is fast.
There is a second benefit: sequence. The baseline tells you which URLs exist and which already earn impressions. The query map decides what each URL is for. Only then do templates and content make sense, because you finally know which page answers which question. Doing content first is the most common and most expensive mistake on a catalog site.
- One owner. A module with two owners is a module with none. Name a person, not a department.
- One output. A file, a sheet, a template change, a set of URLs. Something that exists after the week.
- One test. Checkable in ten minutes by someone who did not build it.
- One date. Modules ship weekly; scope flexes, deadlines do not.
02 / the kit
Six modules, six completion marks
Treat the grid as a build status, not a menu. You run all six, in order, and you stop each one when its test passes.
M02DONE
Query to URL map
One intent per URL, grouped into clusters. Overlaps are merged or redirected before any writing starts.
Acceptance readyM03DONE
On-page templates
Title, H1, heading order, internal link rules and media fields applied at template level, not page by page.
Acceptance readyM04DONE
Content briefs
Briefs that name the query, the sub-questions to answer, the proof to include and the word range.
Acceptance readyM05IN BUILD
Structured data
One JSON-LD type per template, filled from fields the business already maintains. Validated before release.
Validating on stagingM06QUEUED
Measurement
Search Console, analytics events and a monthly sheet that compares mapped clusters against the baseline.
Starts week 5
Kit progress
4 of 6 modules accepted · week 4 of 6
03 / output and proof
What each module leaves behind
If you cannot hand the output to a colleague and have them check it in ten minutes, the module is not defined tightly enough. This table is the contract.
| Module | Output | Acceptance test | Owner |
|---|---|---|---|
| M01 Baseline | URL inventory with status, canonical and indexability | Every live template is listed; zero unknown statuses | Analyst |
| M02 Query map | Cluster sheet mapping one intent to one URL | No two URLs target the same primary query | Strategist |
| M03 Templates | Title, heading and link rules per page type | Unique title and H1 on all sampled URLs | Developer |
| M04 Content | Briefs plus published pages for priority URLs | Every brief question is answered on the page | Writer |
| M05 Data | One JSON-LD type per template, validated | Markup matches visible page; no errors | Developer |
| M06 Measure | Monthly sheet with clusters against baseline | Four weeks of data; every cluster mapped | Analyst |
Template work is where the leverage sits. On-Page SEO Services applies these rules to every page type instead of editing pages one at a time, and SEO Content Writing Services fills the briefs with copy that answers the query rather than describing the company.
04 / schedule
The 30-day push, in four blocks
Four weeks is enough to get a mid-size catalogue from unknown to measured, provided the scope stays on the pages that already show demand. Extend only after the first pass is proven.
DAYS 1–7Baseline crawl, Search Console export, priority URLs chosen.
DAYS 8–14Query map, overlap cleanup, template rules agreed with dev.
DAYS 15–23Template release, first ten pages rewritten, internal links added.
DAYS 24–30Structured data release, measurement sheet live, next sprint scoped.
Baseline
Map
Build
Measure
What can slip without breaking the plan
Content volume can shrink. Structured data can move a week. Internal link passes can be partial. What cannot slip is the baseline, because without it month two has no comparison and the whole kit loses its evidence.
What the team sees each Friday
A one-page review: modules accepted, URLs shipped, queries mapped, and one open decision. That cadence matters more than any single tactic, because it keeps implementation and reporting on the same clock.
05 / traps
Four ways a module kit fails
- Writing before mappingContent produced before the query map usually duplicates an existing page, and the two then compete for the same intent.
- Testing by opinionIf the acceptance test is "looks good", it is not a test. Write the condition, then check it against live URLs.
- Module sprawlSix modules become twenty and nothing finishes. Keep the set closed; put new ideas in the next sprint, not the current kit.
- No measurement ownerReports get built but never read. Assign one person to compare clusters monthly, or the baseline quietly expires.
Run the kit with someone who has shipped it before
We can own two or three modules, hand over the rest, or run all six with your team.
06 / questions
Questions before you start the first module
How long does it take to do SEO for a website?
The build is usually four to six weeks for a focused site. First movement in search commonly appears between months two and four, depending on competition and how much copy needs rewriting.
Must every module finish before anything publishes?
No. Baseline and map come first because they define what exists and what each URL is for. On-page and content then ship page by page while later modules continue.
What counts as an acceptance test?
A short condition anyone can check: a unique title and H1 on every sampled URL, or no page more than three clicks from home. If it takes longer than ten minutes, tighten it.
How many pages should the first pass cover?
Start with ten to thirty URLs that already earn impressions or revenue. Prove the template there before rolling it across the catalogue, or you will not know what caused the change.
Can one person run this without an agency?
Yes, if a developer is available for templates and tracking. The risk is skipping the baseline, which leaves you unable to prove what worked or defend the next sprint.
How do I know it is working?
Watch impressions and average position for mapped clusters, plus conversions from organic landing pages. Compare monthly against the baseline and note what shipped in each period.