The Starting Point: Invisible to AI
A UK-based regulated professional services firm came to MarGen with a problem that is becoming increasingly common: strong traditional search rankings, a solid reputation built over 15 years, but zero presence in AI-generated answers.
When their target buyers asked ChatGPT, Perplexity, or Google AI Overviews questions directly related to their services, the firm didn’t appear. Competitors (some with weaker credentials) were being cited and recommended instead.
The firm had invested heavily in SEO. They ranked on page one for several key terms. But AI search was creating a parallel discovery channel where none of that investment was paying off.
The Diagnosis: Strong SEO, Weak Entity Signals
MarGen’s initial AI Visibility Audit revealed a pattern we see repeatedly in regulated sectors:
- Citation frequency: 0%. The firm appeared in none of the top 30 AI-relevant queries for their sector
- Entity signals: fragmented. Inconsistent brand mentions across directories, no schema markup, no structured authorship
- Content format: SEO-optimised but not AI-extractable. Long-form pages written for keyword ranking, not for answer extraction
- Competitor citation share: 3 firms dominated. The same three competitors appeared in 80%+ of relevant AI answers
The firm had the authority. They had the expertise. But AI systems couldn’t see it because the signals weren’t structured in a way that language models could parse and cite.
The 90-Day Synaptic Authority Engine Implementation
Days 1-15: Entity Mapping
We started with the foundation: making the firm legible to AI.
- Comprehensive schema markup deployed across all key pages (Organisation, Person, Service, FAQ)
- Google Business Profile fully optimised with service categories, descriptions, and attributes
- Directory audit: identified 40+ directories where the firm had inconsistent or missing listings, corrected all of them within two weeks
- Authorship attribution added to all content, individual practitioners linked to their professional credentials
- llms.txt file created and deployed, providing AI systems with a structured overview of the firm
Days 15-45: Trust Trident Activation
Three simultaneous workstreams:
Directory Consensus. Beyond basic listings, we ensured the firm was prominently listed in the sector-specific directories that AI models weight heavily: professional body registers, industry accreditation databases, and specialist comparison platforms.
PR Citation Stack. Secured two bylined articles in sector trade publications and one podcast appearance for the managing partner. Each piece was structured to contain extractable statements that AI models could cite.
Community Presence. Built systematic presence in the professional forums and LinkedIn discussions where sector queries originate. Not promotional, genuinely helpful answers that established expertise.
Days 45-75: Answer-First Content Architecture
This is where the content strategy shifted from SEO to GEO:
- Restructured the firm’s top 10 service pages with extractable answer blocks, self-contained sentences at the top of each section that AI models can lift verbatim
- Created 8 new FAQ pages targeting the exact prompt clusters their buyers use in AI searches
- Built 3 comparison pages directly addressing “best [service] firms in [location]” queries
- Added structured data (FAQ schema, HowTo schema) to every relevant page
Days 75-90: Citation Monitoring and Competitive Displacement
- Set up real-time monitoring across ChatGPT, Perplexity, and Google AI Overviews for 30 priority queries
- Identified which competitor citations were weakest (least authority backing them) and targeted those positions first
- Deployed additional content specifically designed to displace the two most-cited competitors on high-value queries
The Results at 90 Days
| Metric | Day 0 | Day 90 |
|---|---|---|
| AI citation frequency (top 30 queries) | 0% | 37% |
| Queries where firm appears in top 3 citations | 0 | 11 |
| Competitor displacement (queries where firm replaced a competitor) | 0 | 7 |
| Inbound enquiries mentioning AI research | 0/month | 8/month |
| Schema markup coverage | 0 pages | All key pages |
| Directory consistency score | 34% | 96% |
The firm went from completely invisible in AI search to appearing in over a third of the queries their buyers were asking, in 90 days.
What Made the Difference
Three factors were decisive:
1. Entity signals came first. Most agencies start with content. We started with making the firm legible to AI. Without consistent entity signals, even the best content gets ignored by language models that can’t verify who wrote it.
2. Extractable formatting, not more content. The firm already had strong content. What it lacked was content formatted for AI extraction. Restructuring existing pages delivered faster results than creating new ones.
3. Regulated sector expertise. Every piece of content had to work within regulatory guidelines. Generic GEO tactics that work for SaaS companies would have created compliance risk for this firm. Our sector-specific playbook meant zero compliance issues.
What Happened After 90 Days
The firm continued with MarGen on an ongoing engagement. At the six-month mark:
- AI citation frequency reached 58% across monitored queries
- The firm became the #1 cited source for 4 of their highest-value query clusters
- AI-attributed inbound enquiries grew to 15+ per month
- Two competitors who had previously dominated AI citations dropped to second and third position