Unit economics is the bridge between engineering decisions and the company's income statement. Understanding this connection is what separates a FinOps-aware engineer from one who only thinks about infrastructure.
Cloud Costs as COGS
For software companies, cloud infrastructure costs are classified financially as Cost of Goods Sold (COGS) — the direct cost required to deliver your product to a customer. This is distinct from operating expenses like salaries or marketing.
Gross Margin is calculated as:
$$\text{Gross Margin} = \frac{\text{Revenue per Unit} - \text{Cost per Unit}}{\text{Revenue per Unit}} \times 100\%$$
The Margin Squeeze in Practice
Imagine you sell a SaaS subscription for $10/month per user:
| Scenario | Revenue/User | Cloud Cost/User | Gross Margin | Interpretation |
|---|---|---|---|---|
| Healthy | $10.00 | $2.00 | 80% | Strong margin; room for R&D, sales, profit |
| Tight | $10.00 | $5.00 | 50% | Acceptable but leaves little room for error |
| Danger Zone | $10.00 | $9.00 | 10% | Barely covering infrastructure; no room for growth |
| Underwater | $10.00 | $12.00 | -20% | Losing money on every user; unsustainable |
The "Underwater" scenario is more common than you might think in early-stage startups that prioritize growth over efficiency. Unit economics makes this visible before it becomes a crisis.
Benchmarks by Industry
Healthy SaaS gross margins (after COGS including cloud) typically range from 70–85%. If your cloud unit costs are pushing gross margins below 60%, it is a signal that architectural optimization is a business priority, not just a technical nicety.
The Engineering Implication
Every architectural decision has a margin implication:
- Choosing a managed service (e.g., RDS) vs. self-managed database affects cost per transaction.
- Over-provisioning "just in case" directly reduces gross margin.
- A poorly written N+1 database query that causes 10x more DB calls than necessary translates directly to a higher cost per API call, which translates to a lower gross margin.
Unit economics makes this chain of causation visible and quantifiable.