Phony Cloud - Pricing & Competitive Analysis
Purpose: Determine the optimal pricing strategy to maximize success probability
Last Updated: July 2026 - Revised to the connected-data-source model (local-first data plane)
Part 1: Competitive Pricing Analysis (Corrected)
1.1 Competitor Pricing Models (Actual)
┌─────────────────────────────────────────────────────────────────────────┐
│ COMPETITOR PRICING MAP │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ TONIC.AI │
│ ───────── │
│ │
│ STRUCTURAL (DB Sync): │
│ • $199/mo base + $19/table (up to 20 tables included) │
│ • Enterprise: Volume-based custom pricing │
│ │
│ FABRICATE (Schema-first, LLM-based): │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ Free: $10/mo credits (Trial access) │ │
│ │ Plus: $25/mo credits + pay-as-you-go │ │
│ │ Enterprise: Custom │ │
│ │ │ │
│ │ ★ CRITICAL INSIGHT: │ │
│ │ • LLM-based = expensive per record │ │
│ │ • $10 credits ≈ 100-1,000 records/month (varies by complexity) │ │
│ │ • Heavy users hit limits FAST │ │
│ │ • Cannot be truly unlimited (LLM cost floor) │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ GRETEL.AI (Acquired by NVIDIA, March 2025, $320M+) │
│ ────────── │
│ • Free: 15 credits/month (≈100K synthetic records) │
│ • Additional: $2/credit │
│ • Enterprise: Custom │
│ • Model: Usage-based credits │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ MOSTLY AI │
│ ───────── │
│ • Free: Limited generation │
│ • Pro: 100 credits/month + features │
│ • Enterprise: Custom │
│ • Model: Freemium + credits │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ FAKER (PHP/JS/Python) │
│ ───── │
│ • Price: $0 (OSS only, no cloud) │
│ • Limitation: Static lists, no learning, no cloud features │
│ │
└─────────────────────────────────────────────────────────────────────────┘1.2 The Fundamental Cost Difference
┌─────────────────────────────────────────────────────────────────────────┐
│ COST STRUCTURE COMPARISON │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ LLM-BASED COMPETITORS (Tonic Fabricate, Gretel, etc.) │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ │ │
│ │ Cost per record: $0.001 - $0.10 (depends on complexity) │ │
│ │ │ │
│ │ 100K records = $100 - $10,000 in LLM costs │ │
│ │ 1M records = $1,000 - $100,000 in LLM costs │ │
│ │ │ │
│ │ → They CANNOT offer truly unlimited generation │ │
│ │ → Credit-based model is REQUIRED for their business │ │
│ │ → Heavy users are expensive to serve │ │
│ │ │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ PHONY (Statistical N-gram) │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ │ │
│ │ Cost per record: ~$0.0000001 (CPU cycles only) │ │
│ │ │ │
│ │ 100K records = $0.01 in compute │ │
│ │ 1M records = $0.10 in compute │ │
│ │ 100M records = $10 in compute │ │
│ │ │ │
│ │ → We CAN offer truly unlimited generation │ │
│ │ → Generation is essentially FREE for us │ │
│ │ → Heavy users cost us almost nothing extra │ │
│ │ │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ ★ THIS IS OUR MOAT ★ │
│ We can do something competitors literally CANNOT: │
│ Offer truly unlimited synthetic data generation. │
│ │
└─────────────────────────────────────────────────────────────────────────┘Part 2: Strategic Positioning
2.1 The Winning Message
┌─────────────────────────────────────────────────────────────────────────┐
│ PHONY'S UNIQUE VALUE PROPOSITION │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ HEADLINE (at EVERY tier, including Free): │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ │ │
│ │ "Unlimited generation. Unlimited users. Unlimited AI agents." │ │
│ │ │ │
│ │ or │ │
│ │ │ │
│ │ "Your data never leaves your network. Pay per connected │ │
│ │ data source — nothing else is ever metered." │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ AGENT-ERA RATIONALE: │
│ In the agent era, coding agents multiply test runs, seeded │
│ databases, and ephemeral environments. Metering usage or seats │
│ punishes your best adopters. Phony meters ONE thing: the │
│ connected data source. What you generate with it — and how many │
│ people or agents do the generating — is never counted. │
│ │
│ COMPARISON TABLE (for marketing): │
│ ┌───────────────────┬──────────────────┬──────────────────────────┐ │
│ │ │ Tonic Fabricate │ Phony Cloud │ │
│ ├───────────────────┼──────────────────┼──────────────────────────┤ │
│ │ Free tier │ $10 credits/mo │ Unlimited generation │ │
│ │ 100K records │ Uses all credits │ Still unlimited │ │
│ │ 1M records │ $100+ extra │ Still unlimited │ │
│ │ Speed │ ~10 rec/sec │ 100,000+ rec/sec │ │
│ │ Deterministic │ No (LLM) │ Yes (same seed = same) │ │
│ │ Local training │ No │ Yes (MIT OSS) │ │
│ │ Data leaves env │ Yes (LLM cloud) │ Never (local data plane) │ │
│ └───────────────────┴──────────────────┴──────────────────────────┘ │
│ │
│ ATTACK ANGLE: │
│ "Tonic Fabricate's $10/month sounds free until you generate │
│ your first real dataset. Phony is actually unlimited — and │
│ your production data never touches our servers." │
│ │
└─────────────────────────────────────────────────────────────────────────┘2.2 What We Actually Sell
┌─────────────────────────────────────────────────────────────────────────┐
│ VALUE PROPOSITION CLARITY │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ Generation is our WEDGE, not our REVENUE: │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ │ │
│ │ PHONY OSS (Free forever) │ │
│ │ ├── CLI: train + generate (unlimited, local) │ │
│ │ ├── Language libraries + pre-trained models │ │
│ │ ├── Git-based package distribution (no hosted registry) │ │
│ │ ├── Local mock server │ │
│ │ └── MCP / agent surface │ │
│ │ │ │
│ │ Value: Replace Faker with something that learns │ │
│ │ Goal: Adoption, community, trust → Cloud funnel │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ │ │
│ │ PHONY CLOUD (Paid) │ │
│ │ ├── Generation, users, AI agents: UNLIMITED at every tier │ │
│ │ │ │ │
│ │ │ ★ THE KEY DIFFERENTIATOR ★ │ │
│ │ │ Competitors meter usage or seats. │ │
│ │ │ We include both free. Always. │ │
│ │ │ │ │
│ │ ├── WHAT WE CHARGE FOR (the value metric): │ │
│ │ │ • Connected data sources — the production databases │ │
│ │ │ that Phony syncs and anonymizes │ │
│ │ │ • Control-plane depth per tier: scheduling, PII │ │
│ │ │ inventory, compliance reports, SSO, audit retention │ │
│ │ │ │ │
│ │ ├── WHAT WE NEVER METER: │ │
│ │ │ • Generation, records, rows, tables, sync volume │ │
│ │ │ • Users, seats, AI agents, environments │ │
│ │ │ │ │
│ │ └── LOCAL-FIRST DATA PLANE: sync and anonymization run in │ │
│ │ YOUR infrastructure (the phony agent). Production data │ │
│ │ never reaches Phony servers. The cloud is a control │ │
│ │ plane only. │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ PHILOSOPHY: │
│ "Generation is free. The connected data source is the value │
│ metric." │
│ │
└─────────────────────────────────────────────────────────────────────────┘Part 3: Optimal Pricing Model
3.1 Design Principles
┌─────────────────────────────────────────────────────────────────────────┐
│ PRICING DESIGN PRINCIPLES │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ 1. NEVER METER USAGE │
│ • Generation, records, rows, tables, sync volume: unlimited │
│ • Users, seats, AI agents, environments: unlimited │
│ • At EVERY tier, including Free — this is the moat │
│ • In the agent era, metering usage or seats punishes the │
│ best adopters │
│ │
│ 2. VALUE METRIC = THE CONNECTED DATA SOURCE │
│ • A production database that Phony syncs and anonymizes │
│ • Scales with value delivered, not with usage │
│ • Trivially countable; impossible to hit by accident │
│ • No overage anxiety, ever │
│ │
│ 3. LOCAL-FIRST DATA PLANE │
│ • Sync/anonymization compute runs in the CUSTOMER's infra │
│ (the phony binary/agent) │
│ • Production data never reaches Phony servers │
│ • Cloud = control plane only (orchestration, scheduling, │
│ PII inventory, audit, compliance reports) │
│ • Result: ~92-93% gross margin (Enterprise ~85%) │
│ │
│ 4. SIMPLE & PREDICTABLE │
│ • Customer knows exactly what they'll pay │
│ • Flat per data source — tables never counted (vs Tonic) │
│ • No surprises at month end │
│ │
│ 5. GENEROUS FREE TIER │
│ • All OSS free forever: CLI train + generate, language │
│ libraries, models, git-based packages, local mock server, │
│ MCP/agent surface │
│ • Free cloud tier: 1 preview data source, zero hosted-infra │
│ COGS to us │
│ │
│ 6. CLEAR UPGRADE TRIGGERS │
│ • "Need another data source" → Upgrade │
│ • "Need scheduled / CI-triggered sync" → Upgrade │
│ • "Need compliance reports, SSO, audit" → Upgrade │
│ • NOT "ran out of credits" — that trigger does not exist │
│ │
└─────────────────────────────────────────────────────────────────────────┘3.2 Final Pricing Structure
┌─────────────────────────────────────────────────────────────────────────┐
│ │
│ PHONY CLOUD PRICING │
│ │
│ "Unlimited generation · unlimited users · unlimited AI agents" │
│ — at every tier. Pay per connected data source. │
│ │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ FREE $0 — no credit card │ │
│ ├───────────────────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ Generation / users / AI agents: ★ UNLIMITED ★ │ │
│ │ Data sources: 1 (preview) │ │
│ │ Sync: manual │ │
│ │ PII detection: basic heuristics │ │
│ │ History: 7 days │ │
│ │ Hosted infra: none (zero COGS to serve) │ │
│ │ │ │
│ │ Plus all OSS, free forever: CLI train + generate, language │ │
│ │ libraries, models, git-based packages, local mock server, │ │
│ │ MCP/agent surface │ │
│ │ │ │
│ │ Perfect for: evaluation, side projects, OSS users │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ STARTER $49/month │ │
│ ├───────────────────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ Generation / users / AI agents: ★ UNLIMITED ★ │ │
│ │ Data sources: 1 │ │
│ │ Sync: daily, scheduled │ │
│ │ PII detection: full │ │
│ │ Hosted mock endpoints: 5 │ │
│ │ Tracked snapshots: 10 │ │
│ │ │ │
│ │ Perfect for: individual developers, small products │ │
│ │ │ │
│ │ vs Tonic Structural: their base fee alone is $199/mo — │ │
│ │ before the first per-table charge │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ TEAM $199/month │ │
│ ├───────────────────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ Generation / users / AI agents: ★ UNLIMITED ★ │ │
│ │ Data sources: 3 │ │
│ │ Sync: hourly / cron + CI/CD triggers │ │
│ │ PII inventory report │ │
│ │ Hosted mock endpoints: 25 │ │
│ │ Audit log: 30 days │ │
│ │ │ │
│ │ Perfect for: small teams, multiple services │ │
│ │ │ │
│ │ vs Tonic Structural at 20 tables: $579/mo — and every │ │
│ │ schema migration can raise their bill │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ BUSINESS $599/month │ │
│ ├───────────────────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ Generation / users / AI agents: ★ UNLIMITED ★ │ │
│ │ Data sources: 10 (+$39/mo each additional) │ │
│ │ Compliance: KVKK/GDPR report pack │ │
│ │ Security: SSO/SAML + RBAC │ │
│ │ Audit log: 1 year │ │
│ │ Sync: CDC (change data capture), PITR │ │
│ │ Hosted mock endpoints: unlimited │ │
│ │ │ │
│ │ Perfect for: companies with compliance requirements │ │
│ │ │ │
│ │ vs Tonic Structural at 50 tables: $1,149/mo │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ ENTERPRISE from $15K/year │ │
│ ├───────────────────────────────────────────────────────────────────┤ │
│ │ │ │
│ │ Custom (typical band $15-40K/yr, blended ~$20K/yr) │ │
│ │ │ │
│ │ Everything in Business, plus: │ │
│ │ • Unlimited data sources │ │
│ │ • Self-hosted / air-gapped control plane │ │
│ │ • SCIM provisioning, SIEM export │ │
│ │ • SLA, DPA/BAA │ │
│ │ • Dedicated support │ │
│ │ │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
│ Annual billing: 2 months free. │
│ Startup / non-profit: 50% off Starter–Business. │
│ │
└─────────────────────────────────────────────────────────────────────────┘The canonical tier matrices live at /spec/product/pricing and /spec/business/model; this document is the rationale behind them.
3.3 Pricing Comparison: Why We Win
┌─────────────────────────────────────────────────────────────────────────┐
│ COMPETITIVE PRICING ANALYSIS │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ SCENARIO: Team anonymizes one production DB for staging + CI │
│ │
│ TONIC STRUCTURAL ($199/mo + $19/table): │
│ • 20 tables: $579/month │
│ • 50 tables: $1,149/month │
│ • 100 tables: $2,099/month │
│ • Cost scales with schema size — a migration raises the bill │
│ │
│ PHONY CLOUD: │
│ • Starter $49/mo (1 data source) or Team $199/mo (3 sources) │
│ • Tables NEVER counted: 20, 50, or 500 tables — same price │
│ • Unlimited generation, users, and AI agents on top │
│ │
│ WINNER: Phony — flat per data source; 3-10x cheaper at any │
│ realistic schema size, and the gap grows with the schema │
│ │
│ ────────────────────────────────────────────────────────────────────── │
│ │
│ SCENARIO: Agent-era usage explosion (LLM-based generation) │
│ │
│ TONIC FABRICATE / GRETEL (credit-based): │
│ • 1M records at ~$0.01/record = $10,000+/month │
│ • Every agent-triggered test run burns credits │
│ • CANNOT offer unlimited (LLM cost floor) │
│ │
│ PHONY CLOUD: │
│ • 1M, 10M, 100M records: included, flat │
│ • 1 developer or 100 AI agents: same price │
│ │
│ WINNER: Phony by 10-100x — only possible because the │
│ statistical engine makes generation essentially free │
│ │
│ ────────────────────────────────────────────────────────────────────── │
│ │
│ TRUST ANGLE (no price tag; often the deal-closer): │
│ • Tonic and Gretel route production data through their cloud │
│ • Phony's data plane runs inside YOUR network — the control │
│ plane never sees production data │
│ • The easiest possible security review │
│ │
│ ────────────────────────────────────────────────────────────────────── │
│ │
│ REVENUE MODEL (at the ~$600K ARR target): │
│ • Starter: 200 × $49 = $9,800/mo │
│ • Team: 90 × $199 = $17,910/mo │
│ • Business: 18 × $599 = $10,782/mo │
│ • Enterprise: 6 × ~$1,665 = $9,990/mo (blended ~$20K/yr) │
│ • Total: $48,482/mo ≈ $582K ARR with 314 paid customers │
│ (fewer, higher-value customers than a usage-metered model) │
│ • Blended ARPU ~$154/mo · free→paid 5-8% · CAC ~$250 │
│ • LTV (24 mo) ~$3,700 → LTV:CAC ~15x · target churn <4% │
│ │
└─────────────────────────────────────────────────────────────────────────┘Part 4: Revised Timeline & Milestones
4.1 Phased Approach with Validation Gates
┌─────────────────────────────────────────────────────────────────────────┐
│ EXECUTION TIMELINE │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ PHASE 1: FOUNDATION + DEMAND VALIDATION (Month 1-3) │
│ ════════════════════════════════════════════════════ │
│ │
│ Month 1-2: Core Engine │
│ • N-gram engine (train, generate, save, load) │
│ • 5 pre-trained models (names, emails, addresses, phones, companies) │
│ • Basic CLI │
│ │
│ Month 2-3: Laravel + Launch │
│ • Laravel integration (facades, config, artisan commands) │
│ • Documentation (README, examples, migration guide) │
│ • OSS launch (GitHub, Packagist) │
│ • ★ CLOUD WAITLIST PAGE (phony.cloud) ★ │
│ │
│ Month 3: Validation │
│ • Waitlist interviews ("What would you pay for?") │
│ • Founding member pre-sales (50% discount, lifetime lock) │
│ │
│ KPIs: │
│ • 300+ GitHub stars │
│ • 100+ weekly Packagist downloads │
│ • 200+ Cloud waitlist signups │
│ • 5+ founding member pre-sales │
│ │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ GATE 1 DECISION (End of Month 3): │ │
│ │ │ │
│ │ IF waitlist < 50 AND pre-sales = 0: │ │
│ │ → Pause Cloud development │ │
│ │ → Focus on OSS community building │ │
│ │ → Re-evaluate value proposition │ │
│ │ │ │
│ │ IF waitlist 50-200 OR pre-sales 1-4: │ │
│ │ → Cautious proceed │ │
│ │ → Build MVP with minimal scope │ │
│ │ → Interview waitlist for feature priority │ │
│ │ │ │
│ │ IF waitlist > 200 AND pre-sales > 5: │ │
│ │ → Full speed ahead │ │
│ │ → Expand MVP scope │ │
│ │ → Consider founding member launch │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ PHASE 2: MVP CLOUD (Month 4-6) │
│ ═════════════════════════════ │
│ │
│ MVP Scope (Minimal): │
│ • MySQL connection (source + target) │
│ • Schema introspection │
│ • Basic anonymization (Replace, Mask, Phony Model) │
│ • Full table sync │
│ • Minimal web UI (connect, configure, sync) │
│ • Stripe billing │
│ │
│ NOT in MVP (Later): │
│ • PostgreSQL, SQLite (Month 7+) │
│ • Mock API (Month 9+, if demanded) │
│ • Subset sync (Month 7+) │
│ • Snapshots (Month 7+) │
│ • Team features (Month 9+) │
│ │
│ Month 4-5: Build │
│ • Core sync engine │
│ • Web UI │
│ • Internal dogfooding │
│ │
│ Month 6: Beta │
│ • Invite 10 beta users from waitlist │
│ • Fix critical issues │
│ • Collect feedback │
│ │
│ KPIs: │
│ • MVP working on our own data daily │
│ • 5+ beta users actively testing │
│ • 0 data loss incidents │
│ • Clear feedback on what to build next │
│ │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ GATE 2 DECISION (End of Month 6): │ │
│ │ │ │
│ │ IF beta users not engaged OR major issues: │ │
│ │ → Extend beta, fix problems │ │
│ │ → Do not launch publicly │ │
│ │ │ │
│ │ IF beta users happy AND willing to pay: │ │
│ │ → Proceed to public launch │ │
│ │ → Convert founding members │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ PHASE 3: FIRST REVENUE (Month 7-9) │
│ ══════════════════════════════════ │
│ │
│ Month 7: Launch │
│ • Convert founding members to paid │
│ • Public launch (Product Hunt, HN, Laravel News) │
│ • First real customers │
│ │
│ Month 8-9: Iterate & Expand │
│ • PostgreSQL support (if demanded) │
│ • Snapshots (if demanded) │
│ • Subset sync (if demanded) │
│ • First case study │
│ │
│ KPIs: │
│ • 5+ paying customers │
│ • $1K+ MRR │
│ • <10% monthly churn │
│ • Positive NPS │
│ │
│ ┌─────────────────────────────────────────────────────────────────┐ │
│ │ GATE 3 DECISION (End of Month 9): │ │
│ │ │ │
│ │ IF <5 customers OR >20% churn: │ │
│ │ → Product-market fit problem │ │
│ │ → Deep customer interviews │ │
│ │ → Consider pivot │ │
│ │ │ │
│ │ IF 5+ customers AND <10% churn: │ │
│ │ → Proceed to growth phase │ │
│ │ → Scale marketing │ │
│ │ → Build requested features │ │
│ └─────────────────────────────────────────────────────────────────┘ │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ PHASE 4: GROWTH (Month 10-24) │
│ ═════════════════════════════ │
│ │
│ Month 10-12: │
│ • Mock API feature (if demanded) │
│ • Team features │
│ • Scheduled syncs │
│ • Content marketing │
│ │
│ Year 2: │
│ • Python Phony (Revenue Focus) ★ │
│ • Conference presence (Laracon, PyCon) │
│ • First enterprise customer (Python-based) │
│ • SSO/SAML │
│ │
│ KPIs (Year 2 end): │
│ • $100-150K ARR │
│ • 50-80 paying customers │
│ • <5% monthly churn │
│ • Clear path to $500K+ ARR │
│ │
│ ───────────────────────────────────────────────────────────────────── │
│ │
│ PHASE 5: SCALE & EXIT (Year 3-5) │
│ ════════════════════════════════ │
│ │
│ Year 3: Enterprise features (SOC2, audit) │
│ Year 4: $600K-800K ARR, exit preparation │
│ Year 5: $800K-1M ARR, exit execution │
│ │
│ Exit target: $3-8M (5-8x ARR multiple) │
│ │
└─────────────────────────────────────────────────────────────────────────┘Part 5: Summary - The Winning Strategy
5.1 Key Strategic Decisions
┌─────────────────────────────────────────────────────────────────────────┐
│ STRATEGIC DECISIONS SUMMARY │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ DECISION 1: NEVER METER USAGE │
│ ───────────────────────────── │
│ • Unlimited generation, users, and AI agents at EVERY tier, │
│ including Free │
│ • Competitors cannot match this (LLM cost floor, seat-based │
│ business models) │
│ • Marketing: "Unlimited generation · unlimited users · │
│ unlimited AI agents" │
│ │
│ DECISION 2: VALUE METRIC = THE CONNECTED DATA SOURCE │
│ ──────────────────────────────────────────────────── │
│ • One number that scales with value delivered │
│ • Tables, rows, sync volume, seats, agents: never counted │
│ │
│ DECISION 3: LOCAL-FIRST DATA PLANE │
│ ────────────────────────────────── │
│ • Compute runs in the customer's infrastructure; production │
│ data never reaches Phony servers │
│ • Cloud = control plane only → ~92-93% gross margin │
│ (Enterprise ~85%) │
│ • Trust story vs Tonic/Gretel routing data through their cloud │
│ │
│ DECISION 4: FIVE TIERS KEYED ON DATA SOURCES │
│ ───────────────────────────────────────────── │
│ • Free $0: 1 preview source (manual sync, basic PII, 7-day │
│ history) │
│ • Starter $49: 1 source, daily scheduled sync, full PII │
│ • Team $199: 3 sources, CI/CD triggers, PII inventory │
│ • Business $599: 10 sources (+$39 ea), compliance pack, SSO │
│ • Enterprise from $15K/yr: unlimited, self-hosted/air-gapped │
│ │
│ DECISION 5: PACKAGES ARE GIT-BASED AND FREE │
│ ──────────────────────────────────────────── │
│ • No hosted registry as a paid feature │
│ • Distribution rides on git — free forever, builds adoption │
│ │
│ DECISION 6: VALIDATE BEFORE BUILD │
│ ───────────────────────────────── │
│ • Waitlist from Day 1 (with OSS launch) │
│ • Founding member pre-sales before MVP │
│ • Gate decisions at Month 3, 6, 9 │
│ • Build only what customers ask for │
│ │
│ DECISION 7: MVP = MySQL ONLY │
│ ──────────────────────────── │
│ • PostgreSQL, Mock API, etc. come AFTER first revenue │
│ • Fastest path to validation │
│ • Add features based on demand, not assumptions │
│ │
└─────────────────────────────────────────────────────────────────────────┘5.2 Why This Strategy Wins
┌─────────────────────────────────────────────────────────────────────────┐
│ WHY WE WIN │
├─────────────────────────────────────────────────────────────────────────┤
│ │
│ vs TONIC FABRICATE: │
│ ├── They: $10 credits/month (≈100-1000 records) │
│ ├── We: Unlimited generation (truly unlimited) │
│ ├── They: LLM-based (slow, non-deterministic) │
│ └── We: Statistical (fast, deterministic, CI/CD friendly) │
│ │
│ vs TONIC STRUCTURAL: │
│ ├── They: $199/mo + $19/table ($1,149/mo at 50 tables) │
│ ├── We: Flat per connected data source — tables never counted │
│ ├── They: Route your production data through their cloud │
│ └── We: Data plane stays inside your network │
│ │
│ vs FAKER: │
│ ├── They: Static word lists, no learning │
│ ├── We: Statistical learning, custom models │
│ └── We: Both free, but we're more powerful │
│ │
│ vs AI PLATFORMS (Gretel, MOSTLY AI): │
│ ├── They: Credit-based, expensive at scale │
│ ├── We: Unlimited, predictable cost │
│ ├── They: AI/ML complexity │
│ └── We: Simple, transparent, fast │
│ │
│ OUR UNIQUE COMBINATION: │
│ ┌───────────────────────────────────────────────────────────────────┐ │
│ │ ✓ Truly unlimited generation, users, and AI agents │ │
│ │ ✓ Local-first data plane (prod data never leaves your │ │
│ │ network — competitors route it through their cloud) │ │
│ │ ✓ Flat per-data-source pricing (no usage/seat/agent meters, │ │
│ │ no credit anxiety) │ │
│ │ ✓ Statistical learning (smarter than Faker) │ │
│ │ ✓ DB sync + hosted mock endpoints (unique combination) │ │
│ │ ✓ Self-serve (no sales call needed) │ │
│ │ ✓ Open source core, git-based packages (trust, community) │ │
│ │ ✓ Fast (100K+ records/sec vs 10/sec LLM) │ │
│ │ ✓ Deterministic (same seed = same output) │ │
│ │ ✓ ~92-93% gross margin funds the generosity │ │
│ └───────────────────────────────────────────────────────────────────┘ │
│ │
└─────────────────────────────────────────────────────────────────────────┘