Flow Diagrams
Four maps. The diagnostic IS the product. The site is packaging. The generator is the scaling mechanism.
What does success look like?
Five binary gates. Lead quality >= 7/10 is THE gate.
OUTCOME MAP — Marketing Site Generator
═══════════════════════════════════════
┌─────────────────────────────────────────────────┐
│ SALES TEAM LEAD QUALITY >= 7/10 │
│ (The people receiving the leads rate them │
│ higher than traditional lead gen) │
└────────────────────────┬────────────────────────┘
│
┌──────────────┴──────────────┐
│ │
┌──────────▼──────────┐ ┌──────────▼──────────┐
│ DIAGNOSTIC │ │ SITE GENERATION │
│ COMPLETION >30% │ │ IS CONFIGURATION │
│ │ │ │
│ Genuine value │ │ Non-engineer can │
│ earns completion │ │ produce site from │
│ not form friction │ │ content brief + │
│ │ │ theme config │
└──────────┬──────────┘ └──────────┬──────────┘
│ │
└──────────────┬──────────────┘
│
┌────────────────────────▼────────────────────────┐
│ CONSENT BUILT IN │
│ Every CRM contact has a completed diagnostic │
│ as the consent artifact. Zero scraped data. │
└─────────────────────────────────────────────────┘Outcomes defined. Where does value accumulate in the pipeline?
Where does value accumulate?
Six-stage pipeline: brief → assembly → deploy → distribute → capture → handoff.
VALUE STREAM — Marketing Site Generator ═══════════════════════════════════════ BRIEF ASSEMBLY DEPLOY DISTRIBUTE CAPTURE HANDOFF ───── ──────── ────── ────────── ─────── ─────── ┌────────────┐ ┌────────────┐ ┌────────────┐ ┌────────────┐ ┌────────────┐ ┌────────────┐ │ Venture run │ │ 10 sections│ │ Build + │ │ Sneakers │ │ Diagnostic │ │ CRM entry │ │ 8-phase │ │ + diagnostic │ deploy + │ │ promotes │ │ completion │ │ with full │ │ hybrid │───▶│ + theme │───▶│ analytics │─▶│ via media │─▶│ self-qual │───▶│ context │ │ protocol │ │ from brief │ │ │ │ budget │ │ + consent │ │ + scores │ │ │ │ │ │ │ │ │ │ │ │ │ │ 1 day │ │ 2-3 days │ │ < 1 day │ │ ongoing │ │ per visit │ │ real-time │ └────────────┘ └────────────┘ └────────────┘ └────────────┘ └────────────┘ └────────────┘ TARGET: Brief-to-live in 5 business days (vs 3-4 weeks manual) REUSE ECONOMICS ═══════════════ ┌────────────────────────────┬───────────────────────────────┐ │ Built Once │ Per Site │ ├────────────────────────────┼───────────────────────────────┤ │ 25 marketing components │ Theme variant (colors, fonts) │ │ Nx generator (dual mode) │ Content brief (10 fields) │ │ SEO backfill templates │ Diagnostic questions + scores │ │ Prompt library assemblies │ CRM field mapping │ │ BUILD (hero→section→ │ Event names per site │ │ critique→refine) │ Domain + DNS │ │ AUDIT (4 critics + gate) │ │ │ Contact capture → CRM │ │ │ Analytics event library │ │ │ Deployment pipeline │ │ └────────────────────────────┴───────────────────────────────┘ Marginal cost of site N+1 drops toward zero. Generator + prompt assemblies = configuration, not engineering.
Value stream mapped. What must exist before what?
What depends on what?
Generator built. Bridge gap: venture data to content brief. Diagnostic engine and capture pipeline still needed.
CRITICAL PATH (updated — generator exists)
══════════════════════════════════════════
Marketing components (EXIST, 25 components)
+ Auth (L3, EXIST)
+ CRM contacts (L3, EXIST — companies BLOCKED)
+ Nx generator (BUILT — 40/40 tests, dual mode)
+ Prompt library assemblies (BUILT — build + audit pipelines)
+ SEO backfill (BUILT — robots, sitemap, og, JSON-LD)
+ Theme system (BUILT — primaryColor, accentColor in schema)
───────────────────────────────────────────
→ Venture→ContentBrief transform (BRIDGE GAP)
→ Assembly→content-brief wiring (BRIDGE GAP)
→ Diagnostic template engine (BUILD)
→ Contact capture pipeline (BUILD)
→ First venture site dry run
→ ══════════════════════════════════
→ VALIDATE: Lead quality >= 7/10?
→ ══════════════════════════════════
→ Commission with real venture
BLOCKERS
════════
┌────────────────────────────────┬──────────┬─────────────────────────────────────┐
│ Blocker │ Impact │ Mitigation │
├────────────────────────────────┼──────────┼─────────────────────────────────────┤
│ Venture→ContentBrief transform │ High │ Pure function in agency lib. │
│ (no automated bridge) │ │ Maps venture JSON → 10-field brief. │
├────────────────────────────────┼──────────┼─────────────────────────────────────┤
│ CRM company view BLOCKED │ Medium │ Contacts land without company. │
│ (Issue #20) │ │ Enrichment follows. │
├────────────────────────────────┼──────────┼─────────────────────────────────────┤
│ No diagnostic template engine │ High │ First 2-3 diagnostics hand-built │
│ │ │ to validate pattern, then extract. │
├────────────────────────────────┼──────────┼─────────────────────────────────────┤
│ Prompt assemblies unwired │ Medium │ Assemblies exist but content-brief │
│ │ │ still hand-filled after generation. │
└────────────────────────────────┴──────────┴─────────────────────────────────────┘Dependencies mapped. What gets built in what order?
What gets built in what order?
Build by hand → Extract generator → Commission with generator. Do NOT extract until berleytrails proves the pattern.
PHASE 1: Build berleytrails by hand ⬚ PENDING
════════════════════════════════════
Sprint 0 Sprint 1 Sprint 2 Sprint 3
──────── ──────── ──────── ────────
#1 Site from #4 Diagnostic #7 Analytics #6 Lead scoring
brief engine event wiring + filtering
#2 Sign-up #5 Contact #8 Funnel viz
form capture
#3 Theme system Bridge: venture
→ ContentBrief
transform
PROVES: Pattern works. Diagnostic creates value. Leads convert.
═══════════════════════════════════
GATE: Lead quality >= 7/10 on 30 leads?
If NO → Revisit diagnostic. Do NOT proceed.
If YES ↓
═══════════════════════════════════
PHASE 2: Extract generator ✅ DONE
════════════════════════
#3 Theme variant system → primaryColor + accentColor in schema
#9 SEO backfill → robots.ts, sitemap.ts, opengraph-image.tsx, JSON-LD
#10 Nx generator: marketing-site
• Dual mode: marketing (lead-gen) | app (authenticated)
• 16 schema params, 6 hero variants
• Configurable scroll sections (hook,proof,mechanism,offer,...)
• Blog generation (withBlog), newsletter (withNewsletter)
• Diagnostic route (withDiagnostic), contact route (withContact)
• 40/40 tests passing
PROVED: Patterns extractable. Builds on first run.
Generator exceeds original spec (dual mode, blog, newsletter).
PHASE 3: Commission — dry run with real venture ⬚ NEXT
═══════════════════════════════════════════════
#11 Per-site domain config (Partial — Nx app independent, Vercel pending)
#12 First venture site generated + deployed
PROVES: Zero new components. Independent lifecycle.
Venture JSON → ContentBrief → Generator → Deployed site.Questions
Which map surprised you — and is that where the real insight lives?
- If the diagnostic IS the product, why does most of the build order focus on the site shell instead of the diagnostic engine?
- If brief-to-live drops from 3 weeks to 5 days, does that matter if media distribution takes 4 weeks to produce completions?
- What dependency would kill the whole pipeline if it failed — and do we have a fallback?