← back to portfolio
case study try live ↗
// case study · 2025 — 2026

Nautilida
marketing site, end-to-end.

A complete B2B marketing presence — designed, written, and shipped in Claude Code as a co-founder. Optimized for SEO, AEO, and GEO so AI answer engines cite the product by name.

Role · Co-founder & CEO · marketing lead Scope · Marketing site + microsite + content system Stack · Claude Code · HTML/CSS/JS · llms.txt · schema.org Timeline · Site live in 3 weeks
// 01 — site map

What lives at nautilida.ai.

A snapshot of every surface I designed and shipped — main site, comparison pages, use cases, glossary, blog, and the 7-chapter literacy microsite.

nautilida.ai/

Home

The category positioning: AI Team Lead for Sales. Hero, problem framing, product surfaces, social proof, CTA.

View breakdown →
/compare/{gong,clari,gainsight,...}

Comparison pages (×12)

Head-to-head explainers built for AI-search citation — Gong, Clari, Gainsight, and nine more.

View breakdown →
/use-cases/{manufacturing,catering}

Use cases

Industry-specific framing for recurring-account B2B teams — manufacturing and catering.

View breakdown →
/glossary

Glossary

Plain-English definitions of category terms — AEO-friendly definition pages with schema.

View breakdown →
/blog

Blog

Long-form pieces: the weekly loop manifesto, category notes, comparison narratives.

View breakdown →
/literacy

Literacy microsite

7-chapter category primer for VCs and B2B buyers. Long-form narrative, not a brochure.

View breakdown →
// 02 — page breakdowns

Where what is, and why.

The category landing.

First-time visitors land here and either get the category in 10 seconds or bounce. Everything is paced for that.

The home page does three jobs at once: positions Nautilida as the AI Team Lead for Sales category (not another assistant, not another dashboard), frames the problem in operator language (revenue coordination, not "AI productivity"), and gets a qualified visitor into either a comparison page or the contact flow.

Structured data (Organization, Product, FAQ) is embedded inline; llms.txt at the root tells AI crawlers exactly which pages to cite for which intents.

  • 01Hero — Headline + sub + dual CTA (book a call, see how it works).
  • 02Problem framing — What revenue teams actually run on, and where the breaks are.
  • 03The weekly loop — Monday brief, mid-week priorities, Friday recap, compounding memory.
  • 04Product surfaces — Three concrete capabilities with visual examples.
  • 05Comparison teaser — Quick-glance differentiators vs Gong, Clari, Gainsight.
  • 06Footer — Quick links, contact, llms.txt reference, schema.

Twelve comparison pages, each built for AI-search.

When a buyer asks "Nautilida vs Gong?" in ChatGPT, Perplexity, or Claude — there's a page sitting there ready to be cited.

Each comparison page follows a strict editorial pattern: a short answer at the top (extractable by AI), a differentiator table in the middle, category clarification for buyers who think the two tools are alternatives (often they're complementary), and FAQs with schema at the bottom.

Examples shipped:

  • 01/compare/gong — Revenue intelligence vs coordination layer.
  • 02/compare/clari — Forecasting platform vs operating rhythm.
  • 03/compare/gainsight — CS platform vs cross-functional revenue coordination.
  • 04/compare/{outreach,salesloft,hubspot,...} — Plus nine more, same pattern, different category.

Where the category clicks: industries with recurring revenue.

Vertical pages exist because the buyer's "is this for me?" question is faster to answer with their own industry than with abstract category language.

  • 01/use-cases/manufacturing — Named accounts, predictable reorder cadence, AM-coordinated revenue.
  • 02/use-cases/catering — B2B kitchens with repeat customers, office and event programs, deep account memory.

Each page is structured the same: the industry's recurring-revenue pattern, three concrete coordination problems, how Nautilida's weekly loop addresses each, and a section-specific contact CTA.

The category dictionary.

When a buyer or a search engine asks "what is an AI team lead for sales?" — there is a clean, schema-marked definition page ready to answer.

Every glossary entry follows the same anatomy: short answer (1–2 sentences, AI-extractable), longer explanation, related terms, and FAQPage / DefinedTerm schema embedded inline. Definitions cover the entire category surface — coordination layer, weekly loop, AI team lead, revenue book memory, and more.

Long-form, written to be cited.

The blog isn't a feed of news. Each piece exists to anchor a specific narrative in the buyer's head and to give AI engines something substantive to quote.

Shipped pieces:

  • 01The weekly revenue coordination loop, explained. Manifesto for the operating rhythm. 9 min read.
  • 02What is an AI team lead for sales? Definition + category note. 7 min read.
  • 03Revenue intelligence vs Nautilida. Comparison narrative, complementary positioning. 6 min read.

Each post has the AEO pattern: short answer at the top, narrative body, key takeaways, FAQs with schema.

The Literacy microsite — 7-chapter category primer.

A separate microsite that exists to explain the AI-native sales category to VCs and B2B buyers before any product pitch happens. Long-form, navigable, designed to be opened in fundraising conversations.

  • 01Why AI-Native Sales — Why the category exists now and why it didn't before.
  • 02The Old Stack Breaks — Where the existing revenue tooling falls short.
  • 03Agents, Not Seats — From software-as-tool to software-as-team-member.
  • 04Data as Destiny — Why cross-stack memory is the real moat.
  • 05Trust at Scale — How AI earns the right to act on revenue data.
  • 06The New Metrics — What you measure when coordination is the moat.
  • 07What Comes Next — The trajectory of the category, and what to bet on.

Written and designed solo. Used in live fundraising conversations with VCs and design-partner intros.

// 03 — design system

A small, opinionated system.

Tokens chosen for legibility, AEO-cited copy, and a recognizable look across surfaces.

// palette

Violet / lime / ink / bone

Editorial cream canvas with one assertive accent (violet) and one playful accent (lime). High-contrast text on warm bg, generous white space.

// type

Inter + Instrument Serif + JetBrains Mono

Sans for body, italic serif for poetic emphasis (like this), mono for kickers, URLs, code, and AEO machine-readable structure.

// structure

AEO-first content pattern

Every page has a short answer at the top (extractable), narrative middle, FAQ + schema at the bottom. Built to be cited by Perplexity, ChatGPT, and Claude as a default answer source.

// schema

Structured data + llms.txt

Organization, Product, FAQPage, DefinedTerm, Article schema embedded inline. llms.txt at root routes AI crawlers to the right source pages per intent.

// motion

Reveal on scroll, restraint elsewhere

Fade-up on scroll for narrative sections, brief micro-interactions on CTAs, no auto-playing video or distracting parallax. Marketing reads, doesn't perform.

// tooling

Claude Code, end-to-end

Site designed and shipped entirely in Claude Code. No agency, no developer handoff — copy, layout, schema, and deploy all inside the same loop.

// 04 — outcomes

What this shipped.

3 wks
main site live
12
comparison pages
7
literacy chapters
1st
LOIs + design partner
// 05 — design data

JSON, for rebuild and reference.

All the structure on this page lives as JSON, so a designer or another model can rebuild the entire case study from it.


  

Also embedded as <script type="application/json" id="nautilida-project-data"> below — copy-paste friendly.