Why We Publish Our Evaluation Methodology

MarGen publishes comparative guides to GEO and AEO agencies. Those guides influence buying decisions, and that influence carries responsibility. We believe any organisation that compares and recommends agencies should be transparent about how it reaches its conclusions.

This page explains exactly how we evaluate GEO agencies — the criteria, the weighting, the scoring rubric, and the sources we use. If you disagree with our methodology, we welcome the conversation. If you are an agency that believes your score does not reflect your capabilities, we invite you to submit evidence and we will reassess.

Transparency builds trust. In a market full of opaque rankings and pay-to-play “best of” lists, we take a different approach.

The Evaluation Framework: Seven Pillars

MarGen’s GEO agency evaluation is built on seven pillars, each weighted according to its importance for UK regulated sector businesses — our primary audience.

PillarWeightWhat It Measures
1. Methodology Depth20%Rigour and sophistication of the agency’s GEO approach
2. Regulated Sector Expertise20%Experience with FCA, SRA, CQC, and other UK regulators
3. Metrics and Measurement15%Quality and transparency of performance reporting
4. Platform Breadth15%Coverage across AI platforms (not just Google)
5. Case Studies and Evidence15%Verifiable results from real client engagements
6. Team and Capability10%Depth of specialist GEO talent
7. Transparency and Ethics5%Pricing clarity, contract fairness, honest marketing

Total: 100%. Each pillar is scored 1-10, then weighted to produce a final score out of 10.

Pillar 1: Methodology Depth (20%)

This is the most important differentiator between genuine GEO agencies and those that have rebranded existing services.

What We Look For

Entity-first approach. Does the agency build its strategy around entity architecture — structured data, knowledge graph signals, individual and brand entity development? Or is it keyword-focused with a GEO label?

Prompt cluster methodology. Does the agency conduct systematic prompt cluster research, or does it rely on traditional keyword tools adapted for AI queries?

Content architecture. Does the agency design content systems (pillar pages, supporting content, internal linking) specifically for AI comprehension, or does it produce standalone articles?

Technical GEO capability. Schema markup, llms.txt implementation, AI crawler management, structured data validation — does the agency have genuine technical depth?

Scoring Rubric

ScoreDescription
9-10Proprietary methodology with documented process, entity-first approach, technical depth, and continuous innovation
7-8Strong methodology with clear GEO-specific processes, good technical capability
5-6Adequate methodology but heavily derived from SEO — limited GEO-specific innovation
3-4Generic digital marketing methodology with GEO terminology applied
1-2No discernible GEO-specific methodology

Pillar 2: Regulated Sector Expertise (20%)

For MarGen’s audience, this pillar carries equal weight to methodology. GEO for regulated businesses is fundamentally different from GEO for e-commerce or consumer brands.

What We Look For

Named regulatory knowledge. Can the agency reference specific regulations — FCA PRIN 2A, COBS 4, SRA Transparency Rules, CQC Fundamental Standards — and explain how they affect GEO strategy?

Compliance workflows. Does the agency have documented content approval processes that accommodate regulatory review?

Hallucination management. Does the agency actively monitor for and respond to non-compliant AI citations?

Regulated sector client history. Has the agency worked with FCA-authorised firms, SRA-regulated practices, CQC-registered providers, or equivalent?

Scoring Rubric

ScoreDescription
9-10Deep regulated sector experience across multiple regulators, documented compliance processes, named regulatory knowledge
7-8Meaningful regulated sector experience with at least one regulator, compliance-aware workflows
5-6Some regulated sector clients but limited compliance-specific processes
3-4Claims regulated sector capability but limited evidence
1-2No regulated sector experience or awareness

Pillar 3: Metrics and Measurement (15%)

A GEO agency that cannot measure its impact is guessing. The quality of an agency’s measurement framework reveals the depth of its understanding.

What We Look For

Citation-specific metrics. Does the agency track citation frequency, accuracy, sentiment, and platform coverage — or does it rely on proxy metrics like organic traffic?

Commercial attribution. Can the agency connect AI citations to website traffic, leads, and revenue?

Reporting transparency. Are reports clear, regular, and accessible to both marketing and leadership audiences?

Benchmarking. Does the agency benchmark performance against competitors and industry standards?

Scoring Rubric

ScoreDescription
9-10Comprehensive citation metrics with commercial attribution, real-time dashboards, competitor benchmarking, board-ready reporting
7-8Strong citation metrics with good reporting, some commercial attribution
5-6Basic citation tracking with monthly reporting, limited attribution
3-4Metrics focused on traditional SEO proxies with GEO terminology
1-2Minimal or no GEO-specific measurement capability

Pillar 4: Platform Breadth (15%)

AI visibility is not a single-platform challenge. An agency that only optimises for Google AI Overviews is addressing perhaps 40% of the AI search landscape.

What We Look For

PlatformWeight in AssessmentRationale
Google AI OverviewsHighLargest UK search market share
ChatGPTHighFastest-growing AI platform in UK
PerplexityMediumTransparent citation model, growing B2B usage
ClaudeMediumStrong in professional and regulated contexts
GeminiMediumGoogle ecosystem integration
Microsoft CopilotLow-MediumEnterprise relevance growing

An agency scoring 9-10 on platform breadth tracks and optimises for all six. A score of 5-6 indicates coverage of two to three platforms.

Pillar 5: Case Studies and Evidence (15%)

Claims without evidence are marketing. We weight verifiable results heavily.

What We Look For

Named or detailed anonymous case studies. We prefer named clients but accept detailed anonymised studies if the metrics are specific and the methodology is clear.

Before/after data. Citation frequency before engagement versus after. Platform coverage improvement. Commercial impact (traffic, leads, revenue) where available.

Time-specific results. “Improved AI visibility” is not a case study. “Increased citation frequency from 4.2 to 18.7 per 100 target queries over 120 days across ChatGPT, Perplexity, and Google AI Overviews” is.

Regulated sector case studies. For our evaluations, case studies from regulated businesses carry additional weight.

Scoring Rubric

ScoreDescription
9-10Multiple detailed case studies with specific metrics, named clients or richly detailed anonymised studies, regulated sector representation
7-8Good case studies with clear metrics, at least some regulated sector examples
5-6General case studies with limited metrics specificity
3-4Testimonials rather than case studies, no specific metrics
1-2No verifiable case studies or evidence of results

Pillar 6: Team and Capability (10%)

GEO requires a specific mix of skills — technical SEO, content strategy, entity architecture, data analysis, and (for regulated sectors) compliance awareness. We assess whether the agency has the depth of talent to deliver.

What We Look For

Pillar 7: Transparency and Ethics (5%)

A small but important pillar. Agencies that are transparent about pricing, honest about timelines, and fair in their contracts deserve recognition.

What We Look For

How We Gather Information

Our evaluations draw from multiple sources:

SourceHow We Use It
Agency website and published materialsMethodology claims, case studies, team profiles
Direct agency briefingsAgencies can request an evaluation session
Client referencesWhere provided, we contact references with standardised questions
Industry reputationPeer assessment from other specialists in the market
Platform and tool testingWe test claimed capabilities where possible
MarGen’s own experienceCompetitive analysis from our client work

We do not accept payment for evaluations. We do not offer “premium listings” or “featured agency” placements. Every agency is assessed against the same criteria.

How Agencies Can Improve Their Score

We actively encourage agencies to engage with our evaluation process. If you are a GEO agency and want to improve your MarGen evaluation score:

  1. Submit evidence. Send us case studies, methodology documentation, and client references.
  2. Request a briefing. We offer 60-minute evaluation sessions where agencies can present their capabilities.
  3. Address gaps. Our published evaluations identify specific areas for improvement. Address them and we will re-evaluate.
  4. Be transparent. Publish your pricing, your methodology, and your results. Transparency is scored.

Using Our Evaluations

Our agency comparison guides are designed for UK businesses evaluating GEO agencies — particularly those in regulated sectors. They are a starting point, not a final decision.

We recommend using our evaluations alongside your own due diligence: direct agency conversations, reference checks, and pilot engagements where appropriate.

If you want help interpreting our evaluations for your specific situation, or if you want to understand your AI visibility position before approaching any agency, request a free audit. We will provide an independent assessment with no obligation.