You don't need to become a finance expert — a few concepts applied at the level of a single unit are usually enough to reason soundly about whether a feature makes business sense.
1Aggregate Growth Can Hide Per-Unit Losses
A business or feature can look successful by total volume while actually losing money or value on every individual unit — growth in that state makes the underlying problem worse, not better, since each new unit adds to the loss rather than fixing it.
2CAC and LTV Define Whether Growth Is Healthy
Cost to acquire a customer (CAC) compared against the value that customer generates over time (LTV) is the core check for whether growing the customer base is actually building value or actively losing money per new customer — a simple, transferable framework for reasoning about many feature and pricing decisions.
3Step-by-Step Breakdown
Unit economics means looking at the cost and value of a single unit — one customer, one transaction, one API call — rather than only aggregate totals. A business can have a large total revenue and still lose money on every individual unit sold, which aggregate numbers alone can hide.
Two core numbers: CAC (cost to acquire one customer) and LTV (total value one customer generates over their relationship with the product). A healthy business generally needs LTV meaningfully higher than CAC — otherwise, growth is actively losing money per new customer gained.
Why can a business have high total revenue and still be in trouble from a unit economics standpoint?
- →It's impossible — high total revenue always means a healthy business
- →Total revenue can hide a situation where each individual customer or unit sold actually costs more to acquire and serve than it generates in value, which only shows up when you look at the per-unit numbers
- →Unit economics is only relevant for very small companies
- →Aggregate revenue and unit economics measure identical things
This connects directly to earlier lessons: a feature's real cost (including AI/API costs at scale, covered earlier) needs to be weighed against the value it's expected to generate per user — the same cost-at-scale thinking from AI features generalizes to any feature with a real per-use cost.
How does the cost-at-scale thinking from the AI features module connect to unit economics more broadly?
- →They're unrelated — cost-at-scale only applies to AI features specifically
- →Both are about evaluating a feature's real per-unit cost against its value at realistic usage volume, rather than assuming a feature that's cheap in a demo stays cheap at scale
- →Unit economics only applies to entire businesses, never individual features
- →AI features are exempt from normal unit economics considerations
Level Up 🚀
Advanced cheat sheets, SEO tricks, and interview prep for this topic.
Browser Support
Fully supported.
Fully supported.
Fully supported.
Fully supported.
Accessibility (A11y)
1Accessibility Investment Has Real, if Harder to Measure, Unit Economics Too
Improving accessibility can expand the addressable market (more potential customers) and reduce support costs (fewer confused-user tickets) — both real unit economics effects, even though they're often harder to attribute directly than a straightforward CAC/LTV calculation.
// Harder to measure directly, but real:
// wider addressable market + lower support cost per userSEO Implications
- 1
Target 'unit economics for engineers' and 'CAC LTV explained for product engineers' rather than generic finance/startup content
Readers here want the concepts applied specifically to feature and product engineering decisions, not a general startup finance course.
Best Practices
Sanity-Check New Features Against Rough CAC/LTV Impact
For a feature that materially affects acquisition or retention, do a rough estimate of its likely effect on CAC or LTV — even an imprecise estimate is often enough to catch a feature that looks good in isolation but would hurt the business's core unit economics.
Frequent Bugs
Evaluating a feature's success purely by an aggregate metric (total signups, total revenue) without checking whether it's actually improving or worsening the underlying per-unit economics.
Pair aggregate growth metrics with a rough per-unit check (cost and value per customer or per use) to catch a feature that's growing volume while quietly making unit economics worse.
Real-World Examples
Growth That Was Actively Losing Money
A team celebrated a marketing-driven signup spike, but a rough CAC/LTV check revealed the campaign's CAC was nearly double the average customer's LTV — every new signup from that channel was a net loss, a fact the aggregate signup-count metric alone had completely obscured.
// Aggregate: signups +40% (looked great)
// Per-unit: CAC $180, LTV $95 -- losing money on every new customer from this channel