SEOMANTIK
Log inGet the course
Why nowWhat it isWhat you buildThe six modulesThe content engineCoursesFAQ
The course · one-time purchase, no subscription

Build your SEO engine. Keep it.

Ten chapters, one build: an SEO engine of your own, made of Claude Code, live search data and the tested prompts that drive it. Each runs on your site. At the end, a domain and a seed keyword go in; a sourced plan comes out.

Get the course, from $149Course chapters
Rather not build it? The same method is ready-made as a subscription, data included.
claude ~/your-projectchapter 02 · a real run
›Crawl yourdomain.com and grade it against the topic universe
●22 pages read, 14 mapped to a subtopic. Graded against all 16 subtopics the topic spans:
│core head terms · thin, worth 321,970 searches a month
│brewing technology internals · nothing on the site
│+ 12 more graded in the file
✓Write data/entity-map.json
●1 subtopic is covered properly. 14 of the 16 are thin or missing, each named with the pages it needs.
Where it runsClaude Code, on your machineTerminal or desktop app. Nothing runs on your site.
How you drive itPlain-English promptsEvery prompt written and tested. Swap in your domain and run.
What a run costs$3 to $5 per runPay as you go. $50 minimum top-up, non-expiring, refundable.
What you keepEvery file, the engineEvery chapter writes a file. After the last one the engine is yours.
AI changed SEO

You rank for the question. The answer never mentions you.

The same question now gets answered twice: once in the ranked results, and once above them, where a few named sources settle it before anyone scrolls. Being a named source is the new job.

01AI answers sit above the results. A page they don't name goes unseen, however well it ranks.
02Search grades topics, not pages. Sites that cover the whole topic win over one perfect article.
03Few sites measure the AI answer layer. The engine checks your top questions against them.
⌕how fine should I grind for espresso?
AI answerYou are not here

Grind fine enough that a double shot runs in about 25 to 30 seconds, then adjust one step at a time.

a big competitora community foruman industry weeklyyour site
Organic resultsYou are here · #3
1Everything you need to knowA big competitor
2The complete guide, updatedAn industry weekly
3How fine should you grind for espresso?Your site
4Which grind size should you choose?A community forum
What the course is

A course you run, not one you watch.

Each chapter runs on your own site, and what it produces stays on your disk. After the last one the engine is yours: point it at another site, or another SEO job, and it runs that too.

01Every prompt comes written and tested. Swap in your domain and your seed, run.
02The chapter explains each part, so you can change any of it.
03Each step saves its work, and the next step picks it up.
Chapter 01 · The Universestep 3 of 7

Point the engine at your domain; the whole topic lands in one pass.

run this in Claude CodeMap the whole topic for yourdomain.com, seed "home espresso machine", en-US. Save the raw pulls, then synthesize data/universe.json.
Swap in your domain and your seed. The chapter explains each line of it.
→ data/universe.json16 subtopics · 104 entities
What you build, and keep

Build one engine. Six files come out of one seed.

You build each part and learn how they feed each other: the topic universe mapped from live data, your coverage graded against it, the questions ranked, the hubs, spokes and internal links laid out, the work sequenced into waves, and the first pages drafted with their sources.

↦Each part writes the file the next one readsuniverse → coverage → questions → architecture → waves → drafts
↻Run it again in a monthSame folder, same prompts, and a list of what changed.
⇄Point it at any siteUse it on every site you own, at no extra cost.
YOUR ENGINEClaude Code · live data
01Topic universe
universe.json
02Coverage, graded
entity-map.json
03Question queue
questions.json
04Architecture
architecture.json
05Build order
build-list.json
06Sourced drafts
output/
YOUR ENGINEClaude Code · live data · your prompts
01Topic universeuniverse.json
02Coverage, gradedentity-map.json
03Question queuequestions.json
04Architecturearchitecture.json
05Build orderbuild-list.json
06Sourced draftsoutput/
The modules

Six modules, one run, one folder. Each file feeds the next module.

Each module is one exchange in Claude Code and one file left behind, read by the next. A run costs $3 to $5 of data.

claude ~/your-projectModule 01
├Fetch related keywords for "home espresso machine"
●Saved the full response. 225 unique keywords came back. Top by volume:
│de'longhi espresso machine · 90,500
│home espresso machine · 49,500
│+ 222 more, 1,651,580 combined monthly searches
├Fetch the live results page for the seed
●Page 1: 9 organic results, 8 People Also Ask questions. No AI overview on this query.
├Ask the AI layer how it clusters the topic
●10 clusters. 3 appear nowhere in the results data: brewing technology, maintenance, troubleshooting. Those are the opportunity subtopics.
✓Write data/universe.json
●16 subtopics, 104 entities, 70 questions. Every item traces back to a record in data/raw/.
data/universe.jsonfeeds module 02
The content engine

The plan is only the start. The engine writes the pages too, and checks each one before it goes out.

The plan says what to write. The engine gathers the evidence first, writes from it, and holds back any page whose claims it cannot trace to a source.

1Assemble the evidenceEvery fact is matched to its source before the writing starts.
2Write from the bundleEvery paragraph written from the evidence you gathered.
3Check the draft, and list what it could not checkA script checks the draft: every source opens, every number points to where it came from, and anything it could not check is listed for you.
4Humanize against real proseRewritten to sound like your site, then checked again so every source survives the rewrite.
5Publish, or hold it backIf a claim has no source behind it, the page waits for your decision.
claude ~/your-project
├Assemble the evidence for "How fine should I grind for espresso?"
●Every fact is stored with the source it came from.
✓Write content/…grind-for-espresso.bundle.json
├Draft from the bundle
├Check traceability, by script
│every source id resolves
│every price, rate, date and duration cites something
│a competitor's claim is credited to that competitor
│5 blind spots printed, each labelled NOT CHECKED
├Humanize against the prose the site already writes
├Check again, the same script, after the rewrite
✓Write content/…grind-for-espresso.final.md
✓Write content/…grind-for-espresso.sources.json
●7 claims, each linked to the sentence it supports and the source behind it. A draft it cannot trace waits for you.
One engine, four kinds of site

Pick your kind of site. Every course teaches the same engine, applied to it.

$149 each, or all four for $399. One payment, no renewal, 14-day refund, every future update included. No monthly tool bill: a run costs $3 to $5 of data.

01
BlogThe whole engine for content sites: topic map, coverage audit, site architecture, drafted articles.
$149one time
Get course
02
SaaSFeature, comparison and use-case pages, planned around how buyers search for software.
$149one time
Get course
03
EcommerceCategory, collection and buying-guide pages, sequenced by buying intent.
$149one time
Get course
04
Local & serviceService and area pages for businesses that live on local search and the map pack.
$149one time
Get course
The bundle
All four$596$399save $197

The same engine on every kind of site you will ever run, so a new project never needs a new course. One payment, no renewal, 14-day refund, updates to all four.

  • ✓All four courses, every chapter
  • ✓Every prompt tested, ready to paste in
  • ✓Yours to keep, updates for life
Get the bundleCard, PayPal or USDT.
The curriculum

Six phases, the content engine, and two extras that keep your work.

Every phase ends with a file on your disk. The last two keep the work safe: one pushes the plan to WordPress, one logs the session so any run resumes cleanly.

00Strategy & SetupClaude Code installed, the data account opened, your project folder set. Everything ready for the first run.setup
01The UniverseThe whole topic mapped from live search data: every subtopic, entity and question search ties to your niche, in one file on your disk.universe.json
02Entity MapYour site crawled and graded against that map: every gap, orphan and off-topic page, named.entity-map.json
03Question Clusters + the answer-layer passEvery question worth answering, ranked, with the ones where AI answers quote a competitor instead of you marked first.questions.json
04ArchitectureThe map turned into a blueprint: hubs, spokes, and the exact internal links that hold the cluster together.architecture.json
05PrioritizationThe whole build sequenced into waves, so the next step is always named: quick wins first, big clusters after.build-list.json
06Content GenerationPer-page briefs built from every upstream file, and your first cluster drafted, sourced, ready to publish.output/
⚙The Content EngineIt gathers the evidence first, writes the page from it, then checks that every claim leads back to a source.sources.json
+Bonus · WordPress automationThe finished plan pushed straight into WordPress: drafts, meta, internal links and redirects, one session.automation
§Annex · Session loggingA short project log so any session picks up cleanly and recovery is one prompt.project log
FAQ

Questions, answered.

Anything else: contact us.

No. You talk to Claude Code in plain English; it runs the tools and reads the data for you. You give it sentences, never code or API calls.

Stop guessing what to write. Run the engine.

A domain and a seed keyword in. A complete, sequenced, sourced site plan out, on your own machine, $3 to $5 per run.

Get the courseAccess the course
From $149, one time, 14-day refund.
Newsletter

Field notes, one email a month.

What moved in search, what the engine does about it, and every course update. One email a month; unsubscribe in one click.

A confirmation link comes first. See the Privacy Policy.