PILLAR TWOOF THE AUTHORITY ENGINE

Semantic SEO. Written for meaning, not for keywords.

Search engines no longer match words to pages. They understand meaning, relationships between concepts, and what a person is actually trying to accomplish. We build every page so a machine can read it that way.

Five layers behind every page

Query semantics

Lexical semantics

Frame semantics

Entity extraction

Answer engineering

Keyword density stopped working a long time ago.

For years, SEO content meant choosing a keyword, using it enough times, and hoping.

Search engines now build a model of meaning. They identify the entities on a page, the attributes attached to them, how those entities relate to each other, and whether the page answers what the person actually wanted. Then they compare that against every other page competing for the same intent.

A page can contain the right words and still fail, because it does not contain the right meaning, in the right structure, in the right order.

This is why so much content is published, ranks briefly, and then disappears. It was written for a keyword rather than for a query.

What a search engine actually sees

Entity

+ attributes

+ relations

Nobody buys software from a single search.

A property manager does not search once and buy. They search across weeks.

They start with a problem they cannot yet name. Guest parking is out of control and residents are complaining. Then they learn the category exists. Then they compare options. Then they search a specific product name alongside the word "alternatives" because someone recommended one and they want to check.

Each of those stages is a different query, with different intent, requiring a different kind of page.

Semantic SEO is how your product appears at every stage of that path, and appears as the answer rather than as one of ten links.

From problem to purchase

Searches

guest parking is out of control

READS a blog post on the problem

Searches

guest parking management software

VISITS a category page, sees 4 options

Searches

[competitor] alternatives

COMPARES 2 products side by side

Requests a demo, buys

Every page we publish is engineered before it is written.

The Topical Blueprint decides what to publish. The Semantic Brief decides exactly how that one page must be built in order to rank. No writer begins without one.

Semantic brief research and specification items
Research
Keyword researchfor the specific topic, not for the category.
Competitor researchfor that exact query set, not for the domain overall.
SERP analysiscovering the format, angle, depth and freshness of what currently ranks. If the top results are all comparison tables, a long essay will not win.
Query semantics.What the person actually wants, as opposed to what they typed.
Lexical semantics.The words the subject genuinely uses, including the ones your competitors are missing.
Frame semantics.The mental frame the query sits inside, which determines what a satisfying answer contains.
N-gram analysisof the ranking corpus, showing the phrase patterns that consistently appear in content that performs.
Entity extractionand the attributes each entity must carry on this page.
Search intent and sub intentdefinition, because most queries contain more than one.
Contextual positionof this page within the Blueprint, meaning what it must connect to and what it must not duplicate.
Specification
Macro context.What the page is fundamentally about. One thing, stated clearly, so the page does not drift.
Micro context per section.What each section covers and why it exists.
Main content and supplementary content,separated, so the writer knows what carries the page and what supports it.
Heading vectorwith the correct HTML tag on every heading, ordered to match how the subject is actually structured rather than how it looks tidy.
Contextual bridgesbetween sections, so the page reads as one argument rather than a list of blocks.
Answer first requirementfor every section. The answer appears in the first sentence, then the explanation follows.
Writing methodologyfor the writer, specific to this page.
Internal linking targetsin both directions. Which pages this page must link to, and which existing pages must link to it.
Schema requirement.
Visual asset requirement.
Word count guidancederived from the SERP rather than from a template. Some pages need 600 words. Some need 3,000. The winning results tell us which.

Scroll to see more →

There is no lighter version for supporting content.

This is deliberate and we get asked about it often.

Most agencies producing content at volume use a full research process for money pages and something much lighter for supporting articles. It is the obvious way to make the numbers work.

We do not do it, because it removes the mechanism.

Outer section pages are what build the topical authority that makes your revenue pages rank. A map covered with shallow supporting content does not establish a site as a source on its subject. It establishes a site with a lot of pages, which is a different thing and worth considerably less.

When volume rises, we add brief specialists. We do not reduce the specification.

One specialist produces fifteen to twenty full briefs a month. Fifty briefs a month means three specialists. Sixty five means four. The cost sits in the price, and the method stays intact.

If the sprint falls behind, the sprint moves. The brief does not.

Same recipe, any batch size

Full Semantic Brief

10 research + 11 spec items
Query semanticsEntity extractionHeading vector+18 more

×15

Core

×50

Coverage

×65

Authority

Writing to specification.

This is not article writing. It is content built against an engineering spec, where structure, sequence and phrasing were decided before anyone started typing.

  • Answer first construction in every section. The reader gets the answer immediately, then the reasoning. This also happens to be exactly what AI systems extract.

  • Entity placement and optimisation. The entities the brief specified, positioned where they carry meaning rather than scattered for density.

  • NLP friendly sentence construction. Clear subject, verb, object. Low ambiguity. Short sentences. A machine parsing the page should never have to guess what a pronoun refers to.

  • Term and co-occurrence balancing. The words that genuinely belong together in this subject, present in the proportions that ranking content uses.

  • Table, list and comparison formatting built for machine extraction rather than for visual effect.

  • Contextual bridging between sections so the page holds together.

  • Editorial review against brief compliance. Every piece is checked against its brief before it goes anywhere. Not checked for style. Checked for whether it did what the specification required.

  • Fact and source verification. Every claim traced. Every statistic sourced.

The production pipeline

  • Answer first construction

  • Entity placement

  • NLP sentence construction

  • Term balancing

  • Table and list formatting

  • Contextual bridging

  • Editorial review

  • Fact verification

Published, brief compliant.

Applied at three points, not bolted on at the end.

Most agencies treat on page SEO as a checklist run after writing. We apply it at three stages, so by the time a page goes live it is already finished.

Stage 1

At brief

Heading vector, entity requirements, linking spec, schema requirement

Stage 2

At writing

Title and heading optimisation, answer engineering, entity placement, term balancing

Stage 3

At publishing

Meta, slug, schema, internal links, Open Graph, indexing confirmation

There is no separate optimisation phase afterwards, because there is nothing left to optimise.

The fastest results usually come from pages you already have.

Improving existing pages is almost always the quickest source of measurable movement in the first ninety days.

Those pages already have history, links and some ranking position. Moving a page from position twelve to position four produces results far faster than a new page can, because the hard part, being indexed and trusted at all, is already done.

Existing page

  • Backlinks14
  • Age2.3 yrs
  • IndexedYes
  • Position12

New page

  • Backlinks0
  • Age0 days
  • IndexedNot yet
  • Position

What we do

  • Decay analysis across all existing content, identifying what used to perform and no longer does.

  • Striking distance identification, meaning pages sitting between positions five and twenty where a modest improvement produces a large traffic change.

  • Intent mismatch diagnosis, where a page targets a query it was never structured to answer.

  • Gap analysis against the pages currently outranking it.

  • Cannibalisation consolidation, where two of your own pages compete for the same query and both lose.

  • Rewrite, restructure, merge or redirect decisions, made per page rather than applied as a policy.

  • A full Semantic Brief for every revived page. Republish, reindex and movement tracking.

Why this matters commercially

Our ninety day performance guarantee is only defensible because content revival is included. New content rarely moves inside ninety days. Revived pages move in weeks.

Documents and published pages.

Every month

  1. 1

    A complete Semantic Brief for every page, delivered before writing begins. You can read them. Most clients do, at least at first, and it is usually the point at which the difference becomes obvious.

  2. 2

    Published pages, fully optimised, entity optimised, internally linked, schema applied, confirmed indexed.

  3. 3

    Revived pages, with before and after documentation.

Delivered every month

Semantic Brief

Before writing begins

Published pages

Fully optimised, indexed

Revived pages

Before/after documented

Volumes by plan

Volumes by plan
CORE25 pages per month during Coverage Sprint, 10 during Authority Sustain, plus 4 revivals per month.
COVERAGE50 during Sprint, 18 during Sustain, plus 8 revivals per month.
AUTHORITY65 during Sprint, 22 during Sustain, plus 12 revivals per month.

Questions

Can we see a sample brief?

Yes. Ask on a call and we will send one, with the client details removed.

Do you write in our tone of voice?

Yes. We take your existing material, style guidance and product language during onboarding. The brief governs structure and meaning. Tone sits on top of it.

Can our own writers work from your briefs?

Yes, and some clients prefer that. We produce the briefs and your team writes. It costs less and works well when you already have writers who know the product.

How do you handle technical product detail?

The brief flags where subject matter input is needed, and we ask specific questions rather than sending you a draft to correct. Your time goes into the parts only you can answer.

Is AI used to write the content?

Research is partly tool assisted, as it should be at this volume. The brief is built by a specialist. The writing is done by writers working to that brief, then reviewed against it. A page that reads as though nobody thought about it will not rank, regardless of how it was produced.

What if a page does not perform?

It goes into the revival queue. Diagnosed, rebriefed, rewritten and republished. Included, not billed again.

No sales sequence. Just the results.

See what your content is missing.

The AI visibility check shows where your product appears when a buyer asks an AI about software in your category. It is usually the fastest way to see whether your content is doing the job you think it is.

Free. Three working days. No call required.