The brief
Crystal Tax is an international corporate-consulting firm founded in 2014, specializing in business structuring for e-commerce and digital companies. The firm operates across 50+ jurisdictions and works with founders relocating their operations between Cyprus, Estonia, the UAE, Singapore, Portugal, and similar destinations.
The original site was a brand placeholder. There was no lead-generation funnel, no service-area landing pages, no schema markup, and no presence in AI-generated search results — at exactly the moment those answers started driving real consultation requests.
What we did
Full site rebuild on Next.js + Cloudflare Workers
We rebuilt the site from scratch on a static-first stack — sub-second LCP across every page, full server-side rendering, proper hreflang for English and Russian, and a content architecture organized around buyer intent (by jurisdiction, by industry, by stage of business).
Service-area landing pages, 12 in total
One per jurisdiction Crystal Tax actively services, each written by a senior copywriter and reviewed by the firm's managing partner. Real numbers in every page: tax rates, formation timelines, banking requirements, total cost ranges.
Schema markup for AI citation
LegalService and Organization schema, FAQPage on every service page, DefinedTerm for tax concepts, Person schema for the managing partner. The goal was specific: appear in LLM-generated answers for queries like "best jurisdiction for e-commerce holding."
LLM-readable content architecture
Content rewritten in the format LLMs cite well — single-claim H1s, evidence in H2s, definitions, FAQs. Crystal Tax was one of the first sites we built explicitly for AI-search visibility as a primary channel.
Lead-capture and consultation booking
Calendly integration with deal-size qualifying questions, Stripe for paid initial consultations, automated CRM sync into HubSpot.
What we learned
Crystal Tax was a useful proving ground for our LLM-optimization practice. Two takeaways that shaped how we build for other consulting clients now:
- Specific numbers in content drive specific citations. Pages that listed exact tax rates, exact formation timelines, exact banking requirements got cited far more than pages that described the same services in marketing language.
- Founder bylines on long-form content materially improved both classic SEO and AI citation rates. Editorial-style author cards with credentials and links outperformed anonymous "Crystal Tax team" attribution by a margin we did not expect.
Related
The full playbook is at /promotion-in-ai-search-and-llm-systems. For similar professional-services builds: /development-of-a-website-for-a-law-firm. For multi-locale builds: /seo-promotion.
Project lead: Marketing MIX. Client: Crystal Tax (crystal.tax). Year: 2023–2026.