Cernio — Revenue & Cost Structure
Version: 1.0 Date: 2026-03-25 Source: Founder Handbook (Ch. 68-84), Strategy Docs (02-api-cost-control, 04-billing-credits) Purpose: Detailed breakdown of how money flows in and out of the business
1. Revenue Model Overview
Cernio uses a Hybrid SaaS + Credit model. This is deliberate:| Model | Problem It Solves |
|---|---|
| Pure subscription | Heavy users generate massive AI costs — unlimited pricing creates large losses |
| Pure credits | Users constantly think about cost — reduces product usage and stalls activation |
| Hybrid | Predictable revenue from subscriptions + controlled AI cost via credits |
2. Subscription Revenue
Plan Structure
| Plan | Monthly Price | Annual Price (20% off) | Target User |
|---|---|---|---|
| Free | $0 | — | Trial — deliver WOW moment |
| Pro | $39-79 | $374-758/yr | Individual exporter |
| Team | $99-199 | $950-1,910/yr | Export teams (2-10 seats) |
| Enterprise | Custom | Custom | Large firms, multi-org |
Pricing decision deferred to post-beta. Pro and Team will be priced based on beta user behavior and willingness-to-pay signals. For projection purposes, we use midpoint estimates: Pro = 149/mo.
Plan Limits (Feature Gates)
| Feature | Free | Pro | Team | Enterprise |
|---|---|---|---|---|
| Discovery searches/month | 3 | 50 | 200 | Unlimited |
| Companies per search | 10 | 25 | 25 | 50 |
| Contact reveals/month | 1 | 100 | 500 | Unlimited |
| Saved leads | 20 | Unlimited | Unlimited | Unlimited |
| Organization members | 1 | 1 | 10 | Unlimited |
| Batch operations | No | No | Yes | Yes |
| API access | No | No | No | Yes |
| Auto-headhunt | No | Yes | Yes | Yes |
| AI company analysis | No | Yes | Yes | Yes |
| Follow-up reminders | No | Yes | Yes | Yes |
| Shared lead workspace | No | No | Yes | Yes |
Subscription Revenue Scenarios
| Scenario | Pro Users | Team Users | Enterprise | Monthly Sub Revenue |
|---|---|---|---|---|
| Beta (Month 3) | 5 | 0 | 0 | $295 |
| Early (Month 6) | 20 | 3 | 0 | $1,627 |
| Growth (Month 12) | 80 | 15 | 1 | $7,455 |
| Scale (Month 36) | 400 | 80 | 5 | $37,260 |
Enterprise priced at $500/mo average for projections.
3. Credit Revenue
How Credits Work
Plan limits define a monthly allowance. Usage within the allowance is “free” (covered by subscription). When a user exceeds their plan limit, credits are consumed.Credit Pricing
| Pack | Credits | Price | Per-Credit Cost | Implied Margin (vs AI cost) |
|---|---|---|---|---|
| Small | 50 | $15 | $0.30 | ~80% (on discovery) |
| Medium | 200 | $50 | $0.25 | ~76% |
| Large | 1000 | $200 | $0.20 | ~70% |
Credit Consumption Table
| Action | Credits | Actual AI Cost | Gross Margin per Action |
|---|---|---|---|
| Buyer discovery search | 1 | ~$0.06 | 70-80% |
| Contact reveal (headhunt) | 1 | ~$0.03 | 85-90% |
| Deep company analysis (enrichment) | 2 | ~$0.10 | 75-83% |
| Market intelligence report | 3 | ~$0.15 | 75-83% |
| Batch operation (per company) | 0.5 | ~$0.03 | 70-80% |
Credit Revenue Scenarios
Assumes ~20% of paid users purchase credit packs, average 1.5 packs/quarter:| Scenario | Paid Users | Credit Buyers | Quarterly Credit Revenue |
|---|---|---|---|
| Beta (Month 3) | 5 | 1 | $50 |
| Early (Month 6) | 23 | 5 | $250 |
| Growth (Month 12) | 96 | 19 | $950 |
| Scale (Month 36) | 485 | 97 | $4,850 |
4. Cost Structure
4.1 Fixed Costs (Monthly)
| Category | Item | Cost | Notes |
|---|---|---|---|
| Infrastructure | Hetzner VPS “Galata” (production) | €9.49 (~$10.30) | 8 vCPU, 16GB RAM, 160GB NVMe |
| Infrastructure | Hetzner VPS “Kadikoy” (worker) | €5.49 (~$5.96) | 4 vCPU, 8GB RAM, 80GB NVMe |
| Infrastructure | Domain + DNS (Cloudflare) | ~$2 | dose-related domains |
| Infrastructure | SSL certificates | $0 | Let’s Encrypt via Coolify |
| SaaS Tools | Stripe | $0 base | 2.9% + $0.30 per transaction |
| SaaS Tools | PostHog (analytics) | $0 | Free tier (1M events/mo) |
| SaaS Tools | Sentry (error tracking) | $0 | Free tier (5K events/mo) |
| People | Founder salary | $0 | Bootstrapped, self-funded |
| Total Fixed | ~$18/mo |
Critical insight: Fixed costs are extremely low (~$18/mo) because of self-hosted infrastructure on Hetzner and bootstrapped operations. This is a major advantage — the business reaches profitability very early.
4.2 Variable Costs (Per Action)
AI Provider Costs (Per 1M Tokens)
| Provider | Model | Input Cost | Output Cost | Best For |
|---|---|---|---|---|
| Gemini | 2.5 Flash | $0.15 | $0.60 | Default (cheapest) |
| Perplexity | Sonar Pro | $1.00 | $1.00 | Web search (headhunt, discovery) |
| Anthropic | Claude Sonnet | $3.00 | $15.00 | Complex analysis |
| OpenAI | GPT-4o | $2.50 | $10.00 | Fallback |
Cost Per User Action
| Action | LLM Cost | Search API | Processing | Total |
|---|---|---|---|---|
| Discovery search | $0.02 | $0.03 | $0.01 | $0.06 |
| Contact reveal | $0.02 | $0.01 | $0.00 | $0.03 |
| Deep analysis | $0.05 | $0.03 | $0.02 | $0.10 |
| Market report | $0.08 | $0.05 | $0.02 | $0.15 |
| Batch (per company) | $0.02 | $0.01 | $0.00 | $0.03 |
4.3 Transaction Costs
| Item | Cost | Notes |
|---|---|---|
| Stripe subscription processing | 2.9% + $0.30 | Per monthly payment |
| Stripe credit pack processing | 2.9% + $0.30 | Per purchase |
| Estimated Stripe cost on $59 Pro plan | $2.01 | 3.4% effective rate |
| Estimated Stripe cost on $50 credit pack | $1.75 | 3.5% effective rate |
5. Unit Economics
5.1 Per-User Economics (Pro Plan — Midpoint $59/mo)
| Metric | Value | Calculation |
|---|---|---|
| Monthly subscription revenue | $59.00 | Pro plan |
| Stripe fee | -$2.01 | 2.9% + $0.30 |
| AI cost (30 searches/mo avg) | -$1.80 | 30 × $0.06 |
| AI cost (50 reveals/mo avg) | -$1.50 | 50 × $0.03 |
| Infrastructure allocation | -$0.36 | $18 fixed / 50 users |
| Net revenue per user | $53.33 | |
| Gross margin | 90.4% |
This is exceptionally high gross margin because AI costs are low and infrastructure is self-hosted. Even at pessimistic usage (50 searches, 100 reveals), margin stays above 80%.
5.2 Per-User Economics (Team Plan — Midpoint $149/mo)
| Metric | Value | Calculation |
|---|---|---|
| Monthly subscription revenue | $149.00 | Team plan |
| Stripe fee | -$4.62 | 2.9% + $0.30 |
| AI cost (100 searches/mo avg) | -$6.00 | 100 × $0.06 |
| AI cost (250 reveals/mo avg) | -$7.50 | 250 × $0.03 |
| Infrastructure allocation | -$0.36 | $18 fixed / 50 users |
| Net revenue per user | $130.52 | |
| Gross margin | 87.6% |
5.3 Credit Pack Economics
| Pack | Revenue | Stripe Fee | AI Cost (if all discovery) | Net | Margin |
|---|---|---|---|---|---|
| Small (50 credits) | $15.00 | $0.74 | $3.00 | $11.26 | 75.1% |
| Medium (200 credits) | $50.00 | $1.75 | $12.00 | $36.25 | 72.5% |
| Large (1000 credits) | $200.00 | $6.10 | $60.00 | $133.90 | 67.0% |
Credit margins are lower than subscription margins but still healthy (67-75%). Credits serve as overflow monetization, not primary revenue.
5.4 Blended Unit Economics Summary
| Metric | Conservative | Expected | Optimistic |
|---|---|---|---|
| ARPU (Monthly) | $45 | $65 | $85 |
| Gross Margin | 78% | 85% | 90% |
| Estimated CAC | $30-50 | $15-30 | $5-15 |
| Estimated LTV (24-36 mo retention) | $1,080 | $1,560-2,340 | $2,040-3,060 |
| LTV:CAC Ratio | 22-36x | 52-156x | 136-612x |
LTV:CAC is extremely favorable because of near-zero CAC (personal network, content marketing, trade fairs) and high gross margins. Even the conservative estimate far exceeds the 3x benchmark for healthy SaaS.
6. Revenue per Search Analysis
This is the core economic equation:| Monthly Searches | AI Cost | Revenue (at $0.25 avg credit) | Gross Profit |
|---|---|---|---|
| 1,000 | $60 | $250 | $190 |
| 10,000 | $600 | $2,500 | $1,900 |
| 100,000 | $6,000 | $25,000 | $19,000 |
| 1,000,000 | $60,000 | $250,000 | $190,000 |
Note: Most searches will be within plan limits (not consuming credits). The subscription fee covers the expected usage. Credits only kick in for power users exceeding limits.
7. Break-Even Analysis
Minimum Viable Revenue
| Expense | Monthly |
|---|---|
| Infrastructure (Hetzner + domain) | $18 |
| AI API costs (at 10 users, 300 searches) | $18 |
| Stripe fees (at 5 Pro subscribers) | $10 |
| Total minimum costs | ~$46/mo |
Break-Even Point
| Scenario | Required Users | Monthly Revenue | Notes |
|---|---|---|---|
| Cover infra only | 1 Pro user | $59 | Infrastructure paid for by single subscriber |
| Cover infra + AI | 2 Pro users | $118 | All variable costs covered |
| Founder salary ($3K/mo) | ~55 Pro users | $3,245 | Modest founder compensation |
| Founder salary ($5K/mo) | ~90 Pro users | $5,310 | Comfortable founder compensation |
| Sustainable business ($10K MRR) | ~170 Pro users | $10,030 | Reinvestment capacity |
Key insight: With $18/mo fixed costs, the business is technically profitable from the first paying customer. The real break-even is founder salary — achievable with 55-90 Pro subscribers.
8. Revenue Expansion Opportunities
Near-Term (Ring 1-2)
| Opportunity | Mechanism | Revenue Impact |
|---|---|---|
| Pro → Team upgrades | Multi-user export teams | 2.5x ARPU increase |
| Credit pack purchases | Power users exceeding limits | 10-20% revenue uplift |
| Annual billing | 20% discount for upfront payment | Cash flow improvement |
Medium-Term (Ring 3)
| Opportunity | Mechanism | Revenue Impact |
|---|---|---|
| Market intelligence reports | Premium add-on (3 credits each) | New revenue stream |
| HubSpot/CRM integration | Premium connector | Feature differentiation |
| Trade fair tool | Mobile companion | Activation channel |
Long-Term (Ring 4+)
| Opportunity | Mechanism | Revenue Impact |
|---|---|---|
| Enterprise API access | Programmatic discovery | High-value contracts |
| AI Export Copilot | Premium AI assistant | Premium tier justification |
| Data licensing | Anonymized export intelligence | New revenue category |
| White-label | Platform for export agencies | Enterprise deals |
9. Cost Optimization Levers
| Lever | Mechanism | Potential Savings |
|---|---|---|
| Provider routing | Use Gemini Flash (cheapest) for classification, Perplexity only for web search | 30-50% AI cost reduction |
| Search cache | 30-day TTL on export_ai_search_history — identical queries don’t hit AI | 20-40% fewer API calls |
| Batch processing | Classify 25 companies per API call instead of 1-by-1 | 10-20x fewer API calls |
| Model selection | Use smaller models for simpler tasks (classification vs analysis) | 20-40% token cost reduction |
| Self-hosted models | Run open-source models on GPU VPS for high-volume tasks (future) | 60-80% cost reduction |
10. Pricing Psychology
Exporters think in terms of lead generation capacity, not AI compute:| Framing | Bad | Good |
|---|---|---|
| Discovery | ”Uses 1 AI credit" | "Find 25 potential buyers” |
| Contact reveal | ”Costs 1 credit" | "Get the purchasing manager’s email” |
| Pro plan | ”$59/month for AI" | "$59/month = 50 buyer searches + 100 contacts” |
| ROI | ”AI tool subscription" | "One distributor deal pays for 10 years of Pro plan” |
A single successful distributor relationship can generate 59/mo, Cernio costs $708/year — the ROI is self-evident.
Summary
| Metric | Value |
|---|---|
| Revenue model | Hybrid SaaS + Credits |
| Primary revenue | Subscription (Pro 149 midpoints) |
| Secondary revenue | Credit packs ($15-200) |
| AI cost per search | ~$0.06 |
| Gross margin (subscription) | 85-90% |
| Gross margin (credits) | 67-75% |
| Fixed costs | ~$18/mo (self-hosted Hetzner) |
| Break-even | 1-2 Pro subscribers (infra) / 55-90 (founder salary) |
| LTV:CAC ratio | 22-612x (depending on assumptions) |
Next: 03-financial-projections.md — 3/6/12 month and 3-year projections