The Synaptic Authority Engine is MarGen’s proprietary methodology for building AI citation authority. It is the system we use with every client, from mid-market SaaS companies to FCA-regulated financial services firms, to systematically move brands from AI invisibility to consistent citation across ChatGPT, Perplexity, Google AI Overviews, and Claude.

This page explains the methodology in full: what each step involves, why it matters, and how it connects to measurable outcomes.

Why We Built It

When MarGen began working in Generative Engine Optimisation, we encountered a problem: there was no established methodology. GEO was new enough that most agencies were applying traditional SEO frameworks to an AI citation problem, and the results were inconsistent.

AI language models do not rank pages. They synthesise answers from multiple sources and decide, in real time, which brands to cite. The signals that drive that decision overlap with SEO signals but are not identical to them. Building citation authority requires a different approach.

The Synaptic Authority Engine was built from the ground up for this specific problem. It draws on our work across regulated UK sectors: financial services, legal, healthcare, SaaS and e-commerce, where the stakes of AI misrepresentation are highest and the compliance constraints are tightest.

If you are comparing methodologies rather than reading ours, our page on the AI SEO framework sets out the named frameworks that exist, what the underlying research actually proves, and the questions worth asking of any of them, including this one.

We publish our own worked example of Step 1 rather than only describing it: the MarGen entity map lists every place our entity exists so a model can resolve them to one organisation, and /llm-info.html states the underlying facts in a form written for machines rather than readers. Both are the deliverables this methodology produces, on our own domain, which is the least we can do before charging for them.

The name reflects the methodology’s design principle: like synaptic connections in a neural network, each step strengthens the pathways that lead AI models to recognise, trust, and cite your brand.

The 13 Pillars

AI cites a brand when it can identify it, verify it, extract a clean answer from it, and find it where it looks. The 13 pillars sit in four authority domains that follow exactly that sequence.

Domain 1, Identity. Can AI confidently identify you?

  1. Entity Definition, a clear, distinct, machine-readable entity: Organisation schema, canonical naming, an on-site entity hub.
  2. Knowledge Graph Presence, representation in Wikidata, the Google Knowledge Panel, and consistent sameAs connections.
  3. Identity Consistency, name, descriptors and key facts consistent everywhere the brand appears.

Domain 2, Credibility. Can AI verify and trust you?

  1. Third-Party Authority, independent credible sources citing the brand through digital PR, earned mentions and data-led PR.
  2. Source Presence, presence on the specific source domains AI already cites for your queries.
  3. Review & Reputation Signals, third-party reviews and reputation markers a model reads as trust.
  4. Authority Relationships, co-occurrence and partnerships with entities already trusted.

Domain 3, Extractability. Can AI lift a clean answer from you?

  1. Answer Structure, claim-first writing, direct answers, clean question and answer structure.
  2. Structured Data & Schema, FAQ, Article, HowTo, Product, Organisation, Speakable.
  3. Citation-Magnet Content, assets engineered to be the quotable answer: original data, frameworks, sourced statistics.
  4. Technical Accessibility, crawler access, clean rendering, structure and speed.

Domain 4, Dominance. Do you win, defend and own the answer?

  1. Comparison & Consideration Presence, winning the high-intent “best X”, alternative and comparison queries.
  2. Competitive Citation Share & Defence, measuring share against competitors, displacing rivals, defending won positions.

Which pillars matter most

Identity and Credibility, and the published research keeps pointing the same way.

Branded web mentions correlate with AI visibility at 0.664, against 0.218 for backlinks (Ahrefs, December 2025). Roughly 85% of brand mentions in AI outputs come from third-party sources rather than the brand’s own domain (AirOps, 2025). Entity presence across four or more third-party platforms is associated with a 2.8x citation lift (Digital Bloom AI Visibility Report).

Domain Authority runs the other way. Across the studies we reviewed it correlates negatively with AI visibility, between -0.08 and -0.21. A strong traditional SEO profile does not transfer, and that is the single most expensive assumption we see brands make.

How it is scored

Each pillar is scored 1 to 5 against stated criteria, and the 13 combine into one GEO Maturity Score.

ScoreBandWhat it means
4.5 to 5.0GEO LeaderScoring 4 or 5 on nearly every pillar. The work shifts to holding the position.
3.5 to 4.4GEO CompetitiveStrong presence with specific gaps. Refinement compounds from here.
2.5 to 3.4GEO DevelopingFoundations in place, significant gaps remain. Targeted investment pays best.
1.5 to 2.4GEO EmergingEarly stage. Needs concentrated foundation work across several domains.
1.0 to 1.4GEO InvisibleLittle or no AI visibility. Full programme.

Plotted on a 13-axis radar chart, the shape says more than the number. A spiky profile means strategic imbalance. A strong outline with a hollow centre means the visible work has been done and the infrastructure underneath it has not.

Fix the weakest pillar, not the strongest

This is the part most people get backwards. The pillars feed each other rather than adding up:

Improving thisAlso improvesBecause
Identity (1 to 3)Third-Party Authority, Source PresenceA clear entity makes third-party data about you accurate, and accurate data is what gets corroborated
Technical Accessibility (11)Answer Structure, Structured DataCrawlers can only extract well-structured content if they can reach and render it
Third-Party Authority (4)Comparison Presence, Citation ShareIndependent corroboration is what turns a mention into a recommendation
Citation-Magnet Content (10)Answer Structure, Source PresenceOriginal data gets quoted, and being quoted puts you on the source domains models already read

So moving a 2 to a 3 on the weakest pillar usually returns more than moving a 4 to a 5 on the strongest. The ceiling is set by the floor.

Two rules we hold ourselves to. Every pillar score is evidenced from an audit output, never estimated, because “I think we are about a 3” is not a measurement. And the assessment is re-run every 90 days, because the score is not static when the platforms are not static.

Which pillars each tier works is set out on packages: Foundation covers the Identity and Extractability foundations, Authority runs all 13, and Dominance runs all 13 multi-front.

The Six Steps

The 13 angles are what we measure. The six steps are how the work gets done. The scoring happens inside Step 1, and what it finds is what Steps 2 to 6 act on.

Step 1: AI Visibility Audit

What it is: A scored assessment of your brand’s current citation footprint across every major AI platform.

What we do:

Why it matters: You cannot improve what you have not measured. The audit establishes a baseline against which every subsequent action is measured. Most clients are surprised by the results: brands that dominate traditional search often have minimal AI citation presence, while smaller competitors with better-structured content may already be earning citations.

Output: A detailed AI Visibility Report with citation gap analysis, competitor mapping, and prioritised recommendations.

Step 2: Prompt Cluster Research

What it is: The identification and mapping of every prompt pattern that potential customers use when asking AI platforms about your products, services, or sector.

What we do:

Why it matters: Traditional keyword research tells you what people type into Google. Prompt cluster research tells you what people ask AI. The queries overlap, but the language, structure, and intent are different. AI prompts tend to be longer, more specific, and more conversational than search keywords.

Output: A Prompt Cluster Map, your strategic roadmap for content creation and citation targeting.

Read more: Prompt Cluster Research: How to Find the Questions AI Is Answering in Your Market

Step 3: Entity Authority Building

What it is: The systematic strengthening of your brand’s entity signals so that AI models recognise your brand as an authority in your sector.

What we do:

Why it matters: AI models decide which brands to cite based partly on entity recognition. If a model does not have strong entity signals for your brand, it will default to citing brands it does recognise, even if your content is objectively better. Entity authority building ensures the AI model knows who you are before it evaluates your content.

Read more: Building Your Brand’s Entity Signals for AI and Search Visibility

Step 4: Citation-Optimised Content

What it is: The creation of content specifically structured to earn AI citations, content that AI models can parse, trust, and reference in their generated answers.

What we do:

Why it matters: Most content is written for human readers or search engine crawlers. Citation-optimised content is written for both, and for AI language models as well. The structural requirements are specific: AI models favour content that provides a clear, quotable answer in the first 150 words, supports it with evidence, and attributes it to a credible source.

Read more: How to Write Content That AI Models Actually Cite

Step 5: Technical Infrastructure

What it is: The technical optimisation of your website to ensure AI platforms can efficiently crawl, parse, and cite your content.

What we do:

Why it matters: The best content in the world will not earn citations if AI platforms cannot access it, parse it, or understand its structure. Technical infrastructure is the foundation that makes everything else work.

Read more: Schema Markup for SEO and AEO: The Complete Practical Guide

Step 6: Measurement and Iteration

What it is: Ongoing monitoring of your AI citation performance, with systematic iteration based on what the data shows.

What we do:

Why it matters: AI citation is not a one-time achievement. Models update their training data, competitors publish new content, and platform algorithms evolve. Ongoing measurement ensures your citation authority grows over time rather than eroding.

How the Six Steps Connect

Each step reinforces the others, which is why the order matters:

This interconnection is by design. AI citation authority builds through the systematic accumulation of signals across every dimension AI models evaluate.

Who It Is For

The Synaptic Authority Engine is designed for UK businesses in sectors where trust, expertise, and compliance matter:

Get Started

Every engagement begins with Step 1, the AI Visibility Audit. It is free, takes one week, and gives you a clear picture of where your brand stands across all major AI platforms.

Request your free AI Visibility Audit or email [email protected].