L0product

Marketing Site Generator

Hybrid marketing sites with trojan horse diagnostics that capture qualified contacts through consent, not scraping.

1,200

Priority Score

Pain × Demand × Edge × Trend × Conversion

Customer Journey

Why should I care?

Five cards that sell the dream

1Why

Chasing kills. Berley compounds.

What if the lead told you their problem before you called?

The friction: Every lead gen campaign Sneakers has run produces poor quality. Sales teams complain. CPL doesn't justify contact time. Scraping + cold contact is the old game.

The desire: Diagnostic templates that earn trust by giving genuine value. The completed template IS the qualified lead — with pain, gaps, conviction, and consent.

The proof: AI Strategy Review built as first trojan horse. Pattern works. Now extract it.

Diagnostic stories

Picture

A fishing boat casting berley into dark water — fish gathering below the surface, each glowing with a diagnostic score. Cinematic, 16:9

2Evidence

Configuration beats engineering.

Can a non-engineer produce a site variant by changing a content brief and a theme?

Each new venture site is a custom project — weeks of engineering. The hybrid protocol exists as documentation but not as an executable pipeline.

Current	Target
3-4 weeks brief-to-live	5 business days
Custom engineering per site	Content brief + theme config
New components per site	Zero new components

Value stream compression

Picture

A content brief document transforming into a live website — typed data objects flowing into rendered sections. Cinematic, 16:9

3Platform

Consent beats scraping.

What data does a completed diagnostic give you that a scraped email never could?

Scraped lead: Name + email. No context. No consent. Legal exposure.

Diagnostic lead: Name + email + self-reported pain + gap scores + conviction levels + consent artifact.

NZBN for company targeting (public record). Diagnostics for personal contact (consent given). Zero scraped personal data.

Contact capture stories

Picture

Two contact cards side by side — left shows only name+email (scraped), right shows name+email+pain+gaps+scores+consent (diagnostic). Cinematic, 16:9

4Loop

Trust compounds. Ads don't.

If you stopped spending on ads today, would your diagnostic still attract leads in six months?

Without diagnostic	With diagnostic
Campaign ends, leads stop	Diagnostic lives permanently
Traffic = cost	Each completion = proof it works
No learning between campaigns	Each diagnostic improves the next

Pipeline economics

Picture

A feedback spiral — diagnostic completions creating social proof, attracting more visitors, producing more completions. Crimson on dark. Cinematic, 16:9

5People

Sales team rates it, not us.

What happens to your pipeline if every lead already told you their problem?

Who	Their job
Business owner	Qualified leads who already know their problem
Media agency (Sneakers)	Lead gen product measured by sales team satisfaction
Person filling diagnostic	Genuine insight — not be sold to
Platform team	Generate production site in one command

Kill signal: Lead quality <5/10 after 30 leads. The metric is their satisfaction, not our funnel.

Who we test with

Picture

A sales team reviewing diagnostic results on a screen — each lead card showing pain scores, gaps, and a quality rating badge. Cinematic, 16:9

1 / 5

Same five positions. Different seat.

The business owner sees qualified leads. The agency sees a product to sell. The visitor sees genuine insight. The platform team sees configuration over engineering.

Feature Dev Journey

How does this get built?

Five cards that sell the process

1Job

The job: berley that qualifies.

What if the diagnostic IS the product and the site is just packaging?

Situation: Lead gen campaigns produce volume, not quality. Sales teams waste time on unqualified contacts.

Intention: Sites with diagnostic templates that earn trust, capture context, and hand off pre-qualified leads.

Obstacle: Zero proof the model works. No diagnostic engine. Pattern exists in one hand-built instance.

Intent contract

Picture

A diagnostic form transforming into a CRM contact card — pain scores flowing from one to the other. Cinematic, 16:9

2Stories

Seven stories, four jobs.

If the THEN doesn't name a data source and threshold, what are you actually testing?

Story	THEN	Forbidden
S1: Site from brief	10 sections, responsive, no placeholders	Placeholder text in required fields
S2: Diagnostic	Scores + recommendations, >30% completion	Empty scores, no recommendations
S3: Contact capture	CRM record with scores + consent	No consent flag, no source field
S5: Generator	Zero new components, builds first run	Manual edits required

Full story contracts

Picture

A test dashboard — seven story IDs with status indicators. Cinematic, 16:9

3Build

60% reuse, 40% new.

What already exists that you haven't wired yet?

Exists	Build
25 marketing components	Diagnostic template engine
5 Nx generators	Contact capture pipeline
Auth (Clerk, L3)	Analytics event wiring
CRM contacts (L3)	Generator extraction
Vercel deployment	Theme variant system

Dependency map

Picture

A component library grid — 25 components lit up, 4 dark gaps where new ones need building. Cinematic, 16:9

4Sprints

Build, extract, commission.

Why build by hand before extracting the generator?

Phase	Ships	Proves
1 (S0-S3)	berleytrails by hand	Pattern works, leads convert
2 (S4)	Extract Nx generator	Patterns are extractable
3 (S5)	Generate stackmates	Zero new components

Build order

Picture

Three phases on a timeline — Build berleytrails, Extract generator, Commission stackmates. Cinematic, 16:9

5Validation

Lead quality >= 7/10 or we stop.

What's the smallest experiment that proves the berley model works?

Kill signal: Lead quality <5/10 after 30 leads from berleytrails pilot. Do NOT extract the generator until berleytrails proves the pattern works.

Gate	Evidence
L1	berleytrails renders all sections from brief
L2	First 10 diagnostic submissions in CRM
L3	Sales team rates quality >= 7/10 on 30 leads
L4	stackmates generated via nx g, zero new components

Commissioning gates

Picture

A kill switch — lead quality gauge showing 7/10 threshold. Below: 30 lead cards, each with a quality rating. Cinematic, 16:9

1 / 5

The pitch is the shape. The flow diagrams prove the thinking. The VV stories validate the value.

Flow Diagrams VV Stories

Problem

Situation

Every lead gen campaign Sneakers has run produces poor quality. Sales teams complain. CPL doesn't justify contact time. Scraped leads have a name but no context. Each new venture site is a custom engineering project.

Intention

A generator that produces marketing sites from content briefs. Diagnostic templates earn trust by giving genuine value. The completed diagnostic IS the qualified lead — with pain, gaps, conviction, and consent.

Obstacle

Zero revenue. No case studies. Diagnostic template engine doesn't exist yet. CRM company view is blocked. Each diagnostic (AI Strategy Review) was hand-built as custom React.

Hardest Thing

Proving the berley model converts. Clear path: generator → site → diagnostic → CRM → revenue. But zero proof that diagnostics produce leads sales teams prefer over traditional lead gen.

Scorecard

Priority (5P)

5/5

Pain

4/5

Demand

4/5

Edge

5/5

Trend

3/5

Convert

Readiness (5R)

Principles4 / 5

Performance1 / 5

Platform3 / 5

Process3 / 5

Players4 / 5

What Exists

Component	State	Gap
Marketing components (25)	Working	HeroSimple, HeroSplit, FeatureGrid, PricingTable, FAQSection, CTABanner, StatsCounter, LogoWall etc. No diagnostic template component.
Nx generators (5)	Working	crud-pages, app-workflow-scaffold, etc. No marketing-site generator. Pattern exists — extraction follows proven flow.
Auth (Clerk)	Working	L3. Need diagnostic submission as optional auth trigger.
CRM contacts	Partial	L3 but company view BLOCKED. Contacts land without company association.
Analytics	Partial	L2 — Sales Dev dashboard. Need per-site event wiring with site_id.
Deployment pipeline	Working	Vercel pipeline exists. Need per-site domain config.
AI Strategy Review	Working	First trojan horse diagnostic — hand-built as custom React. Pattern exists but not extracted.

Relationships

PRD	Contributes
Sales CRM	CRM-001 contact management consumed. Diagnostic leads flow into CRM pipeline.
ETL Data Pipelines	NZBN pipeline for business targeting. Distribution without data = manual targeting.
Sales Dev Agent	Downstream consumer. Agent works qualified leads from diagnostics.
BerleyTrails	First venture implementing this pattern. Reference site for generator extraction.

Kill Signal

HRV pilot lead quality rating <5/10 after 30 leads. If the diagnostic doesn't produce leads that sales teams prefer over traditional lead gen, the model is wrong — not the execution. Do not extract the generator until berleytrails proves the pattern works.

Questions

What happens when the diagnostic gives someone genuine insight and they solve their problem without buying anything — is that failure or proof the berley works?

Which of the five test clients has the highest diagnostic completion probability — and is that the same one with the highest lead value?
At what scale does the generator pattern break — 10 sites? 50? What forces customization back in?
If Sneakers can sell this to any advertiser, what prevents the diagnostic quality from degrading as it scales?