Installs into .claude/skills of the current project.
Are you the author of Unit Economics?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/adamriofc-unit-economics)
---
name: unit-economics
description: "Model customer-level economics — LTV, CAC, contribution margin per unit, and the LTV:CAC heuristic — for Indonesian digital and D2C businesses."
argument-hint: <arppu> <churn> <cac>
risk_level: LOW
rule_type: professional-standard
quality_tier: expert-reviewed
allowed-tools: bash
capability:
purpose: [financial_analysis, unit_economics_modelling]
not_for: [certified_audit_opinion, public_offering_prospectus]
requires: [<arppu> <churn> <cac>]
produces: [bep_units, bep_revenue, contribution_margin, margin_of_safety]
consumes: [context.scale, marketing.cac]
deterministic: true
cross_domain_relevance:
strategy: high
marketing: medium
tax: medium
---
# Unit Economics
Evaluates whether each customer (or each unit sold) is profitable after acquisition and delivery costs.
## Core Metrics
* **LTV** = ARPPU (average revenue per customer per period) × Retention Horizon (1 ÷ monthly churn rate).
* **CAC** = Total acquisition cost (ads, sales, onboarding discounts) ÷ number of new customers.
* **Contribution margin per unit** = Price − variable cost (see break-even-analysis).
* **LTV : CAC** — an industry heuristic, not a hard rule: ≥ 3 is considered healthy; < 1 = loss per customer.
## Rules
* CAC payback period: how many months of the customer's monthly cash inflow it takes to recover CAC — target < 12 months for bootstrapped businesses.
* Separate organic vs paid acquisition: only paid costs enter acquisition CAC; organic is tracked separately.
* Monthly churn is computed from cohorts, not a total average.
## Scope & Safety
* **Use for**: evaluating ad channels, pricing & discount decisions, product prioritization.
* **Do not use for**: external financial reporting — these are management metrics, not PSAK.
* LTV:CAC ≥ 3 is an industry heuristic; validate with your own channel's actual data.
* LTV projections are sensitive to churn assumptions — test scenarios at ±2 churn points.
## Worked Example
Input (local SaaS): ARPPU 150,000/month, churn 5%/month, CAC 800,000, contribution margin 80% of revenue.
Output: horizon = 1 ÷ 0.05 = 20 months; LTV = 150,000 × 20 = **3,000,000**; LTV:CAC = 3,000,000 ÷ 800,000 = **3.75** (healthy, above the heuristic of 3); payback = 800,000 ÷ (150,000 × 0.8) = **6.7 months**.