/ledger

Unit Economics Reference

Purpose: Attribute cloud cost per unit — per customer, per tenant, per transaction, per feature, per user, per request. Unit economics transforms "we spend $X/month" into "we spend $Y per paying customer" — the basis for pricing, deal profitability, and product decisions.

Scope Boundary

• ledger unit-economics: Per-unit cost attribution (this document).
• ledger tagging (elsewhere): Tag taxonomy that enables unit economics.
• ledger estimate (elsewhere): Pre-change cost diff.
• ledger finops-framework (elsewhere): Parent capability domain.
• Magi (elsewhere): Business-level forecasting.
• Pulse (elsewhere): KPI definitions that unit economics ties to.

Why Unit Economics

Total cloud spend tells leadership little. Unit cost answers decisions:

Core Equations

Gross Margin = (Revenue − COGS) / Revenue
Contribution Margin = (Revenue − Variable Costs) / Revenue
LTV = ARPU × Gross Margin × (1 / Churn)
LTV / CAC = Lifetime Value / Customer Acquisition Cost
Payback Period = CAC / (ARPU × Gross Margin per month)

Cost per unit

Cost per unit = Σ (attributable cloud cost) / (unit count in period)

Attribution is the hard part.

COGS Decomposition for SaaS

Attribution Techniques

Direct attribution

Costs that belong to one unit:

• Dedicated instance per tenant.
• Per-tenant database cluster.
• Per-user AI API call cost.

Tag at source; read from Cost-and-Usage-Report.

Proportional attribution

Costs shared across units; split by usage:

• Shared Kubernetes cluster → split by pod-hours per tenant.
• Shared database → split by row count or query count per tenant.
• Shared S3 bucket → split by bytes-read per tenant.

Requires usage metrics per tenant (Pulse).

Unattributable (shared fixed)

Platform costs that don't split per unit:

• CI/CD infrastructure.
• Internal tools.
• Corporate IT.

Treat as overhead / fixed cost; don't force into unit economics.

Kubernetes-specific

Tools: OpenCost (CNCF), Kubecost, StormForge. Attribution via:

• Pod-level resource requests × price.
• Workload labels → team / tenant.
• Namespace → team boundary.

Unit Types

Per-customer

Granularity: individual paying customer. Use for: pricing decisions, deal-profitability analysis.

Per-tenant (B2B SaaS)

Granularity: org account. Use for: enterprise contract renegotiation, tenant migration decisions.

Per-transaction

Granularity: payment / order / message / API call. Use for: variable-cost pricing, margin per transaction.

Per-feature

Granularity: feature use. Use for: feature kill/keep decisions, AI feature profitability.

Per-user

Granularity: active user (paid or free). Use for: free-tier sustainability, freemium conversion economics.

Fixed vs Variable Cost Classification

Contribution margin = (Revenue − Variable costs) / Revenue. Contribution margin > 0 is the minimum viability bar for a new tier or feature.

Benchmark Ranges (SaaS, 2024)

Unit-economics FinOps usually focuses on improving gross margin (reducing COGS) without harming the other metrics.

Workflow

DEFINE      →  unit type (customer / tenant / transaction / feature / user)
            →  period (monthly / quarterly)

ATTRIBUTE   →  direct costs (tags required)
            →  proportional costs (usage-based split)
            →  unattributable fixed costs (overhead bucket)

COMPUTE     →  total attributed cost per unit
            →  revenue per unit (from finance / Pulse)
            →  gross margin per unit
            →  contribution margin per unit

SEGMENT     →  by plan tier (Free / Pro / Enterprise)
            →  by cohort (by signup quarter)
            →  by geography

IDENTIFY    →  unprofitable segments (contribution < 0)
            →  high-cost outliers (whale users)
            →  feature-level unprofitability

RECOMMEND   →  pricing adjustments
            →  optimization targets (where to reduce COGS)
            →  kill-candidates (features / tiers below contribution margin)

HANDOFF     →  Magi: business-level forecast
            →  Pulse: usage metrics required
            →  Launch: pricing changes
            →  Tagging: missing tags for better attribution

Output Template

## Unit Economics: [Product / Segment]

### Definition
- **Unit type**: [customer / tenant / transaction / feature / user]
- **Period**: [month / quarter]
- **Revenue source**: [finance / billing system]

### COGS Decomposition
| Category | $ per unit | % of cost | Trend |
|----------|-----------|-----------|-------|
| Compute | [...] | [...] | [...] |
| Storage | [...] | [...] | [...] |
| Data transfer | [...] | [...] | [...] |
| Databases | [...] | [...] | [...] |
| Third-party | [...] | [...] | [...] |
| Observability | [...] | [...] | [...] |
| CDN | [...] | [...] | [...] |
| Support | [...] | [...] | [...] |
| AI/ML | [...] | [...] | [...] |
| **Total** | [$N] | 100% | — |

### Margin
- **Revenue per unit**: [$N]
- **Gross margin**: [%]
- **Contribution margin**: [%]
- **Benchmark comparison**: [healthy / warning / bad]

### Segment Breakdown
| Segment | Unit cost | Revenue | Gross margin |
|---------|-----------|---------|--------------|
| Free | [...] | [0] | negative (expected) |
| Pro | [...] | [...] | [...] |
| Enterprise | [...] | [...] | [...] |

### Outliers
- **Top 5% costliest customers**: [what they cost, what they pay]
- **Unprofitable segment**: [segment, reason]
- **Feature-level unprofitability**: [feature, contribution]

### Recommendations
- Pricing: [adjustments]
- COGS reduction targets: [categories]
- Kill / unbundle: [features / tiers]

### Attribution Gaps
- Untagged cost: [%]
- Needed tags: [list → handoff to `tagging`]

### Handoffs
- Magi: business forecast update
- Pulse: usage metrics required
- Launch: pricing change rollout
- Tagging: coverage gaps
- Scribe: investor-deck-ready summary

Anti-Patterns

Deliverable Contract

When unit-economics completes, emit:

• Unit type definition + period.
• COGS decomposition (9 categories).
• Margin metrics (gross + contribution) vs benchmarks.
• Segment breakdown (plan tier, cohort, geo).
• Outlier identification (whales, unprofitable segments, features).
• Recommendations (pricing, reduction targets, kill-candidates).
• Attribution gaps report.
• Handoffs: Magi, Pulse, Launch, Tagging, Scribe.

References

• FinOps Foundation — Unit Economics capability (2024 framework)
• Cloud FinOps (Storment & Fuller, 2nd ed, O'Reilly)
• SaaS CFO / Benchmarkit — SaaS benchmark reports
• KeyBanc Capital — SaaS Survey (annual)
• OpenCost — CNCF project for Kubernetes cost allocation
• Kubecost — commercial Kubernetes unit economics
• AWS Well-Architected Cost Pillar
• Azure Well-Architected Cost Optimization
• Google Cloud FinOps best practices
• David Skok — SaaS unit economics primer
• SaaStr — content library on unit-economics for SaaS leaders