Methodology · The Content System · Productised AI content for B2B

How your content is built — and why it delivers results.

A transparent look at our systematic content production methodology: nine phases, measurable quality gates, and an architecture engineered for both classical search engines and AI answer systems.

Production phases
9
Strictly sequential, with quality gates
Data points / article
15–20+
Independently researched & validated
Custom images / article
6
Individually produced incl. infographic
Quality checks
23+
Automated & manually verified

The problem

Most teams produce content on instinct. It shows in the results.

A topic sounds interesting. A piece gets written. It gets published. Then everyone hopes for results. Keyword research is skipped because it “takes too long”. Sources don’t get verified. Images are generic stock or missing. Schema markup, internal linking, AEO structure — not on the checklist. Three months later the article ranks for nothing and converts no-one. The Content System eliminates these failure modes systematically. Every article moves through a nine-phase pipeline; every phase has a quality gate; nothing advances without passing. It pairs with our AI-for-marketing approach and the AI Marketing Workshop if your team wants the strategy frame first.

Our principles

System over instinct. Gate before publish. Data over opinion.

Four convictions shape every decision in the production process. They’re the reason it works regardless of topic, scope, or sector.

01

System over instinct

Good content doesn’t come from inspiration; it comes from a repeatable process. Every article moves through the same phases — regardless of topic or scope.

02

No phase without a gate

A defined quality gate sits between every phase. Until a gate is passed, the next phase doesn’t start. There are no exceptions.

03

Data over opinion

Keyword choice, topic angle, and FAQ questions come from real search data — not guesses. Every claim is backed by a verified source.

04

Learn from every miss

Every failure mode that has appeared in production has been analysed and converted into a systematic check. The process gets better with every article.

Nine phases

The production pipeline at a glance.

Strictly sequential, with quality gates between each. Only when a phase is fully complete and verified does the next phase begin.

Phase 01

Keyword research

Search-data validation, secondary keywords, SERP map, real questions for FAQ.

Phase 02

Deep research

15-20+ independent data points; cannibalisation check; exact-page citations.

Phase 03

Image production

Six custom images including a data-driven infographic, optimised + uploaded.

Phase 04

Article creation

Structure tuned for SEO and AEO. Internal links and visuals integrated as we write.

Phase 05

CMS setup

JSON-LD schema, meta tags, categorisation, hero image. CMS-ready before publish.

Phase 06

Publication

Automated pre-flight + immediate live verification with auto-correct loop.

Phase 07

Internal linking

4-6 incoming links from existing articles with varied anchor text. All republished.

Phase 08

Verification

23+ automated checks across content, images, metadata, schema, and links.

Phase 09

Documentation

Calendar updated, sitemap submitted, full traceability from keyword to publish.

Phases 1-3

Foundation, research, visuals.

The work that decides whether the article has a chance — before a single sentence is written.

01

Keyword research & validation

Search-volume validation, secondary-keyword identification, SERP-landscape analysis, extraction of real user questions from search data. No invented questions — only real queries surface in the FAQ.

02

Deep research & source validation

A research dossier with 15-20+ independent data points. Cannibalisation check against your existing rankings. Every external citation links to the exact page — never a homepage.

03

Professional image production

Six custom images: hero, two full-width, one data infographic, two accent images. Each optimised for load time and uploaded to the CMS before the writing phase begins.

Quality gate · Phases 1-3

Keyword research complete, deep research validated, cannibalisation cleared, all six images uploaded and confirmed. Only then is the writing phase released.

Phases 4-6

Creation, configuration, launch.

Where research becomes a published article — with every technical detail set before the publish button.

04

Article creation (SEO & AEO)

A proven structure optimised for classical search engines and AI answer systems. AEO — Answer Engine Optimisation — is what gets the content cited by ChatGPT, Google AI Overviews, and Perplexity. Internal linking, visuals, and citations integrated as part of the draft.

05

CMS configuration & metadata

JSON-LD schema markup, optimised meta description, correct categorisation, tag mapping, hero image. Every technical detail set before publication begins.

06

Controlled publication

Automated pre-flight check before publish. Immediate live verification with an auto-correct loop if the live version diverges from the draft. HTTP 200 confirmed before sign-off.

Quality gate · Phases 4-6

Article structure, visual components, CMS configuration, schema validation, and live-body integrity individually verified before phase 7 begins.

Phases 7-9

Architecture, verification, closure.

The work after publish that turns a single article into a strategic asset for the whole site.

07

Strategic internal linking

Systematic scan of existing articles for linking candidates. 4-6 incoming links with varied anchor text. Builds topical authority, distributes link equity, and accelerates indexation. All linked pages republished.

08

Post-publish verification

23+ automated checks: content integrity, image rendering, metadata, schema, link functionality, CMS configuration. On failure: immediate correction and re-check.

09

Documentation & indexing

Content calendar updated, production progress logged, sitemap submitted to Google. Full traceability from the keyword decision to the publication date.

Built for the AI era

Your content isn’t just found. It’s cited.

AI answer systems — ChatGPT, Google AI Overviews, Perplexity — synthesise information from multiple sources into a single answer. Whoever gets cited wins. Visitors arriving via AI-engine citations are 4.4× more valuable than classical search visitors.

AEO visitor value

4.4×

more valuable than classical search visitors, per 2026 industry benchmarks.

How AEO works

Three levers, every article.

The Content System bakes all three into every piece — not as a bolt-on, but as part of the structure.

Technical

AI-crawler accessibility

GPTBot, Google-Extended, Claude-Web, and PerplexityBot must reach the content. We configure robots.txt, firewall whitelisting, and confirm no JavaScript rendering blocks access.

Structural

Citable content structure

AI engines parse content in chunks. Each section is a self-contained, citable answer that works independently of the rest of the page.

Semantic

Schema & authority signals

JSON-LD for machine-readable structure, FAQ schema for answer boxes, author schema for trust signals. 6+ external citations and 12-18 strategic internal links signal topical depth.

Quality assurance

The checks every article must pass — without exception.

Nine mandatory check areas, combined automated and manual verification. None of these are optional.

Check areaWhat is checkedMethodStatus
Keyword validationSearch volume, secondary keywords, FAQ questions sourced from real search dataAutomatedRequired
CannibalisationNo new article competes with existing rankingsAutomatedRequired
Source integrityEvery external citation links to the exact page — no homepagesManual + autoRequired
Image auditSix images generated, uploaded, correctly embedded, optimisedAutomatedRequired
Visual componentsAll design elements render correctly, no display errorsAutomatedRequired
Schema markupJSON-LD valid, complete, error-freeAutomatedRequired
Pre-publish checkPre-flight check across every CMS field before publishAutomatedRequired
Live verificationHTTP 200, images, metadata, canonical URL, body integrityAutomatedRequired
Internal linking8-12 outgoing + 4-6 incoming with varied anchor text. All republished.Manual + autoRequired

The difference

Systematic vs. ad-hoc — line by line.

How The Content System differs from conventional content production.

CriterionConventional approachThe Content System
Keyword researchSkipped or superficialVolume validation, secondary keywords, SERP analysis, AEO discovery
ResearchGoogle search, first results page15-20+ independent, validated data points
Source standardHomepage links or no sourcesExact-page URLs, every source individually verified
Images1-2 generic stock photosSix custom images including infographic
AI optimisationNot presentAEO structure, schema markup, AI-crawler accessibility
Internal linkingAd hoc, 1-2 links8-12 outgoing + 4-6 incoming with varied anchor text
Quality assuranceProofread, then ship23+ automated checks plus manual verification
After publishPublished = doneLive verification, sitemap, internal linking, documentation

Investment

Graduated pricing. The first four pieces protect quality; volume rewards commitment.

Per-article pricing in GBP plus VAT. New articles run 800-1,200 words. Refresh / rewrite covers updates to existing pieces. The first four pieces in any monthly volume bill at the base rate; each additional tier reduces marginal cost. EUR equivalents available on request.

Volume (pieces / month)Per new piece (800-1,200 words)Per refresh / rewrite
1-4£250 + VAT£200 + VAT
5-9£150 + VAT£100 + VAT
10-19£125 + VAT£90 + VAT
20+£90 + VAT£75 + VAT

Total monthly investment by volume

Worked examples for typical commitments. Graduated — first four pieces at base rate, then progressive.

QuantityNew contentRefreshesHow it’s calculated
1£250£2001 × base rate
4£1,000£8004 × £250
5£1,150£900(4 × £250) + (1 × £150)
8£1,600£1,200(4 × £250) + (4 × £150)
10£1,875£1,390(4 × £250) + (5 × £150) + (1 × £125)
15£2,500£1,840(4 × £250) + (5 × £150) + (6 × £125)
20£3,090£2,275(4 × £250) + (5 × £150) + (10 × £125) + (1 × £90)
25£3,540£2,650(4 × £250) + (5 × £150) + (10 × £125) + (6 × £90)

AI transparency

How AI shows up in the Content System.

  • Drafting + research. Article drafts and research dossiers are AI-assisted with frontier large language models, then human-edited. Every claim and statistic is human-verified before publication.
  • Image production. Hero, infographic, and accent images are AI-generated to brand-aligned prompts and reviewed by hand before they ship.
  • Author attribution. Every article carries a named human author who reviewed and approved the final text. Authorship signals trust to both readers and AI engines.
  • Compliance posture. Helium42 follows the EU AI Act (Article 50 transparency obligations, effective 2 August 2026) and the UK ICO 2025 guidance on AI in marketing. We do not train external models on client data.

Questions before booking

What marketing leaders ask before they commit.

Six straight answers. Anything else, ask us on the demo call.

What is The Content System?

The Content System is Helium42's productised AI content production methodology. Each article moves through nine sequential phases with a quality gate between every phase. Built around AEO so the content gets cited by AI engines, not just ranked on classical SERPs.

What are the nine phases?

Keyword research, deep research, image production, article creation (SEO + AEO), CMS configuration, controlled publication, strategic internal linking, post-publish verification, documentation. The pipeline diagram above shows the sequence.

How does AEO-optimised content rank in AI search?

Three levers: AI-crawler accessibility (GPTBot, Google-Extended, Claude-Web, PerplexityBot reach the page), citable content structure (each section is a self-contained answer), and schema + authority signals (JSON-LD, FAQ schema, 6+ external citations, 12-18 internal links). The Content System bakes all three into every article.

What does The Content System cost?

Graduated pricing from GBP 90 to GBP 250 per new article (800-1,200 words) plus VAT, depending on monthly volume. The first four pieces of any commitment bill at the GBP 250 base rate; each additional tier reduces marginal cost. The full rate card and worked totals are above.

Why founder-led and not just a tool?

Most AI content services are tools. The Content System is a methodology refined across 100+ articles with both Helium42 founders involved at the design level. Clwyd Probert and Peter Vogel built the quality gates around the failure modes they actually saw in production.

How does it ensure quality?

23+ checks per article, automated and manual: keyword validation against real search data, cannibalisation against existing rankings, source integrity (every external link to the exact page), image audit, schema validation, pre-publish preflight, post-publish live verification, internal-linking audit. Failures block the article from advancing — there are no exceptions.

Book a demo

See The Content System on your own topic.

Tell us about your content goals and your monthly volume target. We respond within one working day to schedule a 30-minute demo — with a live walkthrough of one of your priority topics through the pipeline.

  • Founders reply. No SDR loop. The first response comes from Clwyd or Peter.
  • One working day. Every enquiry is answered within 24 working hours.
  • Real topic demo. The walkthrough uses one of your priority keywords, not a canned showcase.
  • No obligation. The demo doesn’t commit you. If we’re not the right fit, we say so.

Or reach us directly

Email info@helium42.com
Phone +44 (0) 208 834 4795

Want training first?

Book the AI Marketing Workshop instead.

If your team needs the strategy frame before the deployment, the four-hour AI Marketing Workshop is the right starting point. Both founders run every cohort.