Unit Economics

Origin & evolution

The idea is rooted in cost accounting and the older notion of contribution margin per unit, which industrial firms used long before software existed. The startup framing crystallised around two metrics🔢MetricStrategyA unified metric that measures progress, health, or behaviour across the productView reference →: customer acquisition cost (CAC) and customer lifetime value (LTV). The lifetime-value side traces to direct-marketing and database-marketing work in the 1980s and 1990s, where firms began valuing a customer by projected future profit across the whole relationship, not by a single sale.

The pairing became canonical for software through David Skok's SaaS Metrics writing, which set two now-standard rules of thumb: an LTV-to-CAC ratio of roughly 3:1 or better, and a CAC payback period under twelve months. Bill Gurley and other investors reinforced the discipline by warning against vanity growth funded by buying customers who never repay their acquisition cost.

The thinking sharpened again after 2021. As cheap capital tightened, investors returned to unit economics as a screen, and the field grew stricter about what counts: fully loaded CAC including sales salaries, LTV computed on gross margin rather than revenue, and cohort👥CohortGrowthA group of users sharing a common characteristicView reference →-based retention rather than a single blended churn number. The metric did not change; the tolerance for fudging it did.

How it works in practice

A subscription🔂SubscriptionSales & RevenueA recurring subscriptionView reference → app charges 20 GBP per month at a 80 percent gross margin, so each customer contributes 16 GBP of margin monthly. Average customer life is 25 months, giving an LTV of 400 GBP. Paid acquisition costs 180 GBP per customer all-in. The ratio is 2.2:1, below the 3:1 bar, and payback takes about eleven months.

That single calculation reframes the roadmap🛣️RoadmapProduct SpecificationA strategic plan of features and milestonesView reference →. The team can lift the ratio three ways: raise price, cut churn so customers live longer, or lower acquisition cost. They model a retention feature⭐FeatureProduct SpecificationA product capability or featureView reference → that extends average life from 25 to 32 months. LTV rises to 512 GBP, the ratio crosses 2.8:1, and payback is unchanged. Now the retention work has a financial case, not just a product hunch, and the team can rank it against pure acquisition spend on equal footing.

Unit economics vs. its neighbours

• Revenue stream💵Revenue StreamBusiness ModelA source of revenueView reference → names where money comes from. Unit economics judges whether that money, per customer, beats the cost of earning it. A healthy-looking stream can hide poor economics underneath.
• Cost structure🧮Cost StructureBusiness ModelA cost category or structureView reference → is the whole-business view of money going out, including fixed overhead. Unit economics isolates the variable, per-customer slice, which is what tells you whether scaling helps or hurts.
• Metric is any single measured number, such as MRR or churn rate. Unit economics is a composed judgement built from several metrics (CAC, LTV, margin, payback) into a verdict on per-customer viability.

In the graph

In the Unified Product Graph, unit_economics💹Unit EconomicsBusiness ModelPer-unit economic metrics (CAC, LTV, etc.) is a leaf in the Business, GTM and Growth region, attached to the model by business_model_measured_by_unit_economicsBusiness Modelmeasured byUnit Economicshierarchy. The placement is deliberate: it hangs as a sibling of revenue_stream💵Revenue StreamBusiness ModelA source of revenueView reference → and cost_structure🧮Cost StructureBusiness ModelA cost category or structureView reference → under the same business model💰Business ModelBusiness ModelThe business model canvas or definitionView reference →, so the graph holds both the inflow and the cost of inflow in one structure. That lets a query test viability directly, reading the stream that earns and the economics that judge it together, which is the exact pairing that keeps a fast-growing model honest about whether the growth is worth funding.