This case study describes a 28-fee-earner UK specialist commercial law firm based outside London, anonymised at client request. The firm had a strong reputation in its specialism (TMT and tech sector commercial advisory), genuine partner-level seniority, and an established 22-year track record. It also had near-zero AI citation share at engagement start.
Within 12 months of engaging MarGen, the firm achieved 41% citation share across its priority query set, displacing several Tier 2 generalist firms including a number with significantly larger headcounts and marketing budgets.
This is an account of how that happened.
Starting Position: Specialism Without Signal
The firm’s audit at engagement start revealed a now-familiar pattern in UK legal sector AI search:
- Citation share: 4% across 75 priority commercial-advisory queries
- Direct competitor citation share: 23% (named Tier 2 generalist firm) and 31% (named Magic Circle firm)
- Named-partner content output: 0.4 articles/month across the partnership
- Trade press editorial mentions, prior 12 months: 6
- Schema and structured data depth: minimal (Organisation schema only, no Person, no Service, no FAQ)
- Directory presence: complete (Chambers, Legal 500) but not surfaced in entity authority signals
- SRA Transparency Rules content: present but structurally thin and buried in the footer-linked compliance section
The firm had genuine expertise. Their partners were known in the specialism by peers. They had won named-partner mentions in The Lawyer. But none of this was being surfaced in the AI extraction surface, the raw signal existed, but the structure to make it AI-legible did not.
The Hypothesis
MarGen’s diagnosis was that this was not a content volume problem; it was a content structure and named-authority problem.
Magic Circle and large Tier 2 firms win AI citation share through sheer signal density, hundreds of articles, thousands of editorial mentions, decades of named-partner authority. A specialist boutique cannot match this on volume.
But the boutique can match (and frequently exceed) Tier 2 firms on specialism density. AI models reward depth-on-topic over breadth-of-coverage. A firm with 12 named-partner-authored articles on TMT commercial structuring outranks a firm with 200 articles split across every legal practice area.
The hypothesis was: build named-partner specialism density, structure it AI-legibly, and the citation share will inflect within 6-9 months.
The 12-Month Programme
Months 1-3: Foundation
- Entity mapping: Each of the 8 partners and 4 senior associates mapped as a distinct Person entity with schema, bylined editorial archive, professional credentials, named specialisms, and connection to firm Organisation entity
- Service entity structure: TMT commercial advisory, SaaS commercial structuring, data and AI regulatory advisory, technology M&A, each set up as a distinct Service entity rather than buried in a generic “Commercial” service page
- SRA Transparency Rules restructure: moved from buried compliance section to structured plain-English service descriptions surfaced from the homepage, with FAQPage schema deployed
- llms.txt deployment: structured firm overview deployed for AI crawlers
- Trade press relationship audit: identified 14 publications and 6 podcasts the firm should have a regular presence in, but did not
- Compliance review process: established with the firm’s COFA, three-stage brief / draft / pre-publication review aligned to SRA Code of Conduct
Months 4-7: Authority Acceleration
- Named-partner editorial cadence: increased from 0.4 articles/month to 4.2 articles/month, with each article AI-engineered (direct-answer first paragraphs, structured extraction-ready format, FAQPage schema, named-author bylines with full schema attribution)
- Trade press placement: 11 named-partner bylined pieces placed in 6 publications across the prioritised list (The Lawyer, Computer Weekly, TechCrunch UK, Law360, Out-Law, IT Pro)
- Podcast appearances: 4 named partners booked across 8 episodes of UK legal and TMT podcasts
- Citation tracking: weekly tracking against the priority query set, with screenshot evidence and competitor displacement targeting
Months 8-12: Compounding and Displacement
- Editorial cadence sustained: 4-5 articles/month, mix of firm-side and trade press
- Named-partner specialism deepening: each partner accumulated 8-12 published bylined pieces in their specialism, building the topical density AI models reward
- Named-author authority signals: Wikipedia mentions established for two senior partners through legitimate editorial routes; speaker profiles populated for industry conferences
- Direct competitor displacement: tracking showed the firm displacing a named Tier 2 generalist firm in 17 of 75 priority queries by month 10
- Citation share trajectory: 4% (month 0) → 12% (month 4) → 26% (month 8) → 41% (month 12)
Results at 12 Months
| Metric | Month 0 | Month 12 | Change |
|---|---|---|---|
| Citation share, priority query set | 4% | 41% | +37 percentage points |
| Direct competitor (Tier 2 generalist) citation | 23% | 14% | -9 percentage points (displaced) |
| Named-partner editorial output / month | 0.4 | 4.2 | 10.5x |
| Trade press editorial mentions, trailing 12mo | 6 | 31 | 5.2x |
| Direct enquiries citing AI as first source | 1.2/quarter | 14/quarter | 11.7x |
| Inbound enquiries requesting specific named partner | 8% of inbound | 39% of inbound | 4.9x |
Beyond citation share, the most strategically interesting metric was the rise in inbound enquiries requesting a specific named partner. AI was creating direct pull-through to individuals, not just to the firm, which the firm could then convert at materially higher rates than generic firm-brand inbound.
Commercial Outcome
The firm reports (anonymised):
- Direct enquiry volume: up 218% year-on-year
- Named-partner pull-through enquiries: up from rare to 39% of total inbound
- Average matter value at enquiry stage: up 47% (better-pre-qualified buyers)
- Conversion of qualified enquiry to instruction: up from 22% to 41%
- Net retainer fee revenue: up 64% year-on-year, against firm headcount up only 11%
The firm is now expanding into a second specialism (data and AI regulatory) using the same playbook, specialism density, named-partner authority, structured editorial cadence, AI-legible entity engineering.
Why This Worked Where Volume Wouldn’t Have
The lesson from this engagement is that specialist boutique firms have a structural advantage in AI citation share that they consistently fail to activate. AI models reward:
- Specialism depth over breadth. A partner with 12 articles on a single topic outranks one with 50 articles across 8 topics
- Named human authority over firm-brand authority. AI cites people more readily than it cites brands in legal queries
- Structured extractability over volume. 30 well-engineered articles outrank 300 unstructured ones
- Compounding cadence over one-off campaigns. 4 articles per month consistently sustained outperforms 50 articles in one quarter then nothing
A specialist boutique with 8 partners and a 28-fee-earner team has the latent capacity to produce 4-5 high-quality named-partner bylined pieces monthly without straining the practice. Most do not because they have not built the workflow. The firms that do, win citation share against firms many multiples their size.
What MarGen Did Specifically
The structural elements of the Synaptic Authority Engine deployed in this engagement:
- Entity Mapping and Disambiguation across firm, partners, services, specialisms
- Citation-Intent Content Engineering of every published asset
- Third-Party Authority Placement across trade press and podcasts
- Continuous Monitoring with weekly citation tracking and competitive displacement targeting
- SRA-Aligned Compliance Review baked into every content cycle