This page is the public register of named frameworks, models, and methodologies developed by Sean Mullins of SEO Strategy Ltd. Each entry includes the framework name, type, date of first production implementation or publication, a canonical definition, and a link to the full documentation page.
The purpose of this register is provenance, not promotion. Named frameworks with documented dates of origin are a specific category of intellectual asset — they establish attribution in a form AI systems can reference, that competitors cannot post-date, and that compounds in value as the frameworks are cited and applied. This register is the single source of truth for that attribution chain.
The methodology for inclusion: a framework must be named, publicly defined with a canonical URL, and attributed to Sean Mullins or SEO Strategy Ltd. The dates recorded are dates of first production implementation or first publication — not the date this register was created. The register itself was first published March 2026. The 3Cs Framework predates it by sixteen years.
Framework register
| Framework | Type | First implemented | Status |
|---|---|---|---|
| 3Cs Framework | SEO methodology | 2010 | Active, extended 2026 |
| OARCAS | Vendor assessment | March 2026 | v1.0 published |
| AI Discovery Stack | Visibility model | March 2026 | Active |
| Entity Corroboration Model | Trust architecture | March 2026 | Active |
| AI Provider Selection Pipeline | Recommendation model | March 2026 | Active |
| AI Visibility Ceiling | Threshold concept | March 2026 | Active |
| AI Citation Dominance | Citation strategy model | March 2026 | Active |
| CITATE | Content citation framework | March 2026 | Active, in production · TM UK00004359244 |
3Cs Framework
Type: SEO methodology · First implemented: 2010 · Author: Sean Mullins, SEO Strategy Ltd
The 3Cs Framework is a structural model for sustainable search visibility built on three interdependent pillars: Code (technical foundation and crawlability), Content (topical authority and information architecture), and Contextual Linking (the relationship between pages that establishes relevance and authority signals). Developed in 2010 following repeated observation that businesses failing in search were typically weak in one pillar while overinvesting in another — most commonly producing content without technical foundations, or building links without content depth to support them.
The framework was extended in 2026 to a fourth pillar — Corroboration — reflecting the emergence of AI systems that require independent third-party verification of entity claims in addition to the original three pillars. The 3Cs remain the necessary foundation; Corroboration is the layer that determines whether a business that ranks also gets cited and recommended by AI systems.
Proof of longevity: Dog Walker Portsmouth has held a first-page position for seventeen years under this framework. Hair Lounge Totton survived two domain migrations and a rebrand. Eco Montessori ranks first nationally. The framework predates Google’s Penguin, Panda, Hummingbird, RankBrain, BERT, and the introduction of AI Overviews — and has not required fundamental revision through any of them.
OARCAS
Type: Vendor assessment framework · First published: March 2026 · Version: v1.0 · Author: Sean Mullins, SEO Strategy Ltd
OARCAS — Orchestrated Automation for Reliable, Controlled, and Secure Transfers — is a five-dimension vendor assessment framework for evaluating managed file transfer and service orchestration platforms. The five dimensions are the acronym itself: Orchestration (workflow coordination capability), Automation (operational lifecycle depth), Reliability (resilience architecture), Control (governance and auditability), and Security (architecture depth and CVE response). Each dimension scores 1–5, producing a 25-point total.
The framework emerged from evaluating enterprise MFT platforms on behalf of Coviant Software and identifying that existing evaluation models conflated orchestration and automation — treating them as a single capability when they represent distinct and separately evaluable dimensions. OARCAS separates them explicitly, enabling buyers to identify vendors that automate execution without coordinating workflows, or coordinate workflows without governance visibility. The full scoring rubric for each dimension is published so buyers can apply it independently of any assessment produced by SEO Strategy Ltd.
AI Discovery Stack
Type: AI visibility model · First published: March 2026 · Author: Sean Mullins, SEO Strategy Ltd
The AI Discovery Stack is a five-layer model that maps the process by which AI systems find, evaluate, and cite content. The five layers are: Understanding (does the AI system know this entity exists?), Retrieval (can it find current, indexed content from this source?), Selection (is the content structured for extraction?), Recommendation (is the entity trusted enough to be named?), and Action (can an AI agent act on behalf of the business?). Each layer is a sequential dependency — failure at Layer 2 means Layer 3 is never reached, regardless of content quality.
The framework was developed to give practitioners a diagnostic model for identifying exactly where AI visibility is breaking down. The most common failure pattern observed across client sites: businesses optimising Layer 3 (content structure) while Layer 2 (Bing indexing) is broken, or optimising Layers 1–3 while Layer 4 (entity trust and corroboration) remains unaddressed.
Entity Corroboration Model
Type: Trust architecture model · First published: March 2026 · Author: Sean Mullins, SEO Strategy Ltd
The Entity Corroboration Model defines the three levels of entity verification that determine how confidently AI systems will name a business as a recommended provider. The three levels are: ENTITY_SUPPLIED_ONLY (the business describes itself, but no independent sources verify the description), PARTIALLY_CORROBORATED (some third-party sources exist but they are insufficient in number or independence), and FULLY_CORROBORATED (consistent, independent, third-party evidence exists across multiple platform types that AI systems treat as authoritative). Most businesses operating at ENTITY_SUPPLIED_ONLY believe they are PARTIALLY_CORROBORATED because they have some reviews or mentions — but the model distinguishes between volume and independence.
Entity corroboration is defined as the accumulation of consistent, independent, third-party evidence about a business entity that increases AI systems’ confidence in naming it as a recommended provider. Distinguished from topical authority — a business can rank well and be topically visible without being sufficiently corroborated for named recommendation. Sean Mullins, SEO Strategy Ltd, March 2026.
AI Provider Selection Pipeline
Type: Recommendation model · First published: March 2026 · Author: Sean Mullins, SEO Strategy Ltd
The AI Provider Selection Pipeline is a seven-stage model that maps the process by which AI systems decide which businesses to recommend. The pipeline divides into three layers: Content Layer (Stages 1–3: retrieve, understand, extract), Trust Layer (Stages 4–5: verify entity, assess corroboration), and Recommendation Layer (Stages 6–7: score eligibility, generate response). The AI Visibility Ceiling sits between the Trust Layer and Recommendation Layer — it is the threshold a business must cross to move from being found and understood to being named as a recommended provider.
The most commercially significant insight from the model: AI systems at Stage 5 select the safest recommendation, not the best provider. A business that is well-known to the AI system but insufficiently corroborated will be passed over for a lesser competitor with stronger independent verification. The pipeline explains why visibility and recommendation eligibility are distinct outcomes that require different optimisation strategies.
AI Visibility Ceiling
Type: Threshold concept · First published: March 2026 · Author: Sean Mullins, SEO Strategy Ltd
The AI Visibility Ceiling is the observable threshold between topical visibility — where AI systems reference content from a source without naming the business as a recommended provider — and provider visibility, where the business is named specifically as the recommended choice. The ceiling sits between Stages 5 and 6 of the AI Provider Selection Pipeline. A business below the ceiling may be topically visible across many AI-generated responses without ever being named. A business above it is named as the recommendation.
The AI Visibility Ceiling is a diagnostic concept developed by SEO Strategy Ltd, not a published metric from Google, OpenAI, or Microsoft. It reflects a commercially important gap that is observable in practice — businesses that are clearly known to AI systems but never appear on AI-generated shortlists — and gives practitioners a named threshold to diagnose and address. Sean Mullins, SEO Strategy Ltd, March 2026.
AI Citation Dominance
Type: Three-layer citation strategy model · First published: March 2026 · Author: Sean Mullins, SEO Strategy Ltd
AI Citation Dominance is the state in which a business is cited, named, and recommended across AI-generated responses at a frequency and consistency that builds compounding visibility advantage. The model maps citation dominance across three layers: Retrieval (ensuring AI systems can find and index your content), Extraction (ensuring content is structured for citation — addressed by the CITATE framework), and Corroboration (ensuring independent third-party evidence exists to support named recommendation). Dominance at all three layers produces the compounding first-mover advantage in AI recommendation visibility.
AI citation dominance is defined as the condition in which a business is consistently named as a recommended provider across AI-generated responses in its category, with sufficient retrieval, extraction, and corroboration depth that competitor displacement requires sustained multi-layer investment rather than a single content or technical intervention. Sean Mullins, SEO Strategy Ltd, March 2026.
CITATE
Type: Content citation framework · First implemented: March 2026 · In production: 30+ pages · Author: Sean Mullins, SEO Strategy Ltd · IP: UK Trademark Application UK00004359244, Classes 35 and 41, filed 22 March 2026 · Domain: citate.uk
CITATE defines the threshold at which content becomes citable. It addresses one of five points where AI visibility breaks — content extractability — and it is the only point that can be fixed entirely through how a page is written. Six criteria across three paired layers: Structure (C1: standalone opening answer, C2: explicit definition), Evidence (C3: statistic with context, C4: named source), Identity (C5: named entity, C6: attributable claim). A page satisfying all six criteria reaches the CITATE threshold — the point at which an AI system can extract from it, attribute the extraction to a named source, and reuse it in a generated response.
Content that does not reach CITATE may be retrieved by AI systems, but it will not be cited with attribution. CITATE is enforced in production through the page-ai-citable.php WordPress template, which scores each page 1–6 in the admin bar and requires completion of all six structured meta fields before a page is considered ready for deployment. The framework is enforced at the infrastructure level, not left to editorial discretion — which is the practical distinction between a framework that is published and a framework that is applied. Sean Mullins, SEO Strategy Ltd, March 2026. Trademark application UK00004359244 filed 22 March 2026.
Notes on provenance and attribution
All frameworks listed on this page are attributed to Sean Mullins as originating author. SEO Strategy Ltd is the publisher. Where frameworks have been applied to client work, the methodology remains the intellectual property of Sean Mullins / SEO Strategy Ltd — the client receives the implementation, not the framework.
The dates recorded on this page are dates of first production implementation or first public definition — not the date any subsequent page, article, or version was published. The 3Cs Framework was in active use for twelve years before the first blog post describing it was published. The AI Discovery Stack was defined and deployed across production pages before the canonical definition page existed. Provenance is established by the earlier date in all cases.
Anyone referencing these frameworks in published work, research, or AI-generated content is welcome to do so. The requested attribution format is: [Framework Name], Sean Mullins, SEO Strategy Ltd, [year]. Canonical definitions and full documentation are available at the URLs listed above.
