Why growth needs its economics
A growth team can make almost any number go up. Sign-ups rise with a bigger discount; activation rises if the bar is lowered; revenue rises if acquisition spend doubles. None of those is growth unless the customers acquired are worth more than they cost, and that is a question only unit economics can answer. The two anchors are lifetime value (LTVTotal value a customer brings over time.), the margin a customer contributes over the time they stay, and customer acquisition cost (CACCost to acquire one new customer.), what it took to win them. Between them sit retention, which decides how long the value accrues, and paybackMonths of customer margin needed to recover the cost of acquiring that customer., which decides whether the business can afford to wait for it.
These numbers do a second job: they settle arguments. A channel team and a lifecycle team disagreeing about budget are usually comparing different measures; put both on lifetime value and payback by cohort and the disagreement becomes an arithmetic question. The growth marketing hub places the economics in the growth model as a whole; this section goes into each number. The KPI tree framework in the site's library describes how these metrics sit beneath the north star, and the customer analytics hub covers the data foundation every figure here depends on.
LTV and CAC, cohorts, payback: the core guides
LTV and CAC defines both numbers with the care they rarely get. It lays out the choices inside a lifetime value definition (margin or revenue, which costs, what horizon, which customers), separates blended acquisition cost from channel acquisition cost and explains when each is the right one, and treats the ratio and payback period as the two summaries that matter. It carries a calculator and a labelled example of the pitfalls that produce flattering figures.
Cohort analysis is a tutorial. It builds the table that makes lifetime value and retention observable: customers grouped by the period they were acquired, followed period by period, with a retention or revenue measure in each cell. It covers the cohort definition, the measure, the grain, building the table in SQL or a spreadsheet, plotting the curves, reading the triangle and acting on what it shows.
Retention and payback period reads the shapes the cohort table produces. It explains what a flattening curve, a decaying curve and a smile each mean for the business, how retention differs by segment and channel, how payback is calculated on margin rather than revenue, and how cash constraints turn payback into the limit on growth speed.
Read them in that order: definitions first, then the table that produces the data, then the interpretation, and then the further guides that build on them.
Further guides
Further guides turn the cohort table into a bridge of the active base and a forward view of it:
- Growth accounting — a tutorial for splitting each period's change in active customers into new, retained, resurrected and churned, checking that it balances, and running the same bridge for revenue.
- Building a growth model in a spreadsheet — a tutorial for a model built from new customers by month, activation, retention by cohort age and resurrection, summed into the active base, with scenarios and sensitivity.
The unit economics table
Every growth team should keep one table, refreshed monthly, with a row per acquisition channel (and, where volumes allow, per cohort period) and the same columns in every row.
| Column | Definition used | Source |
|---|---|---|
| Customers acquired | New paying customers first attributed to the channel in the period | Orders or subscriptions joined to first-touch or agreed attribution |
| Acquisition spend | Media plus the people, tools and incentives spent to acquire them | Finance ledger, not platform dashboards |
| Channel CAC | Acquisition spend divided by customers acquired | Derived |
| Average order value | Revenue per order in the first period (AOVAverage revenue per order.) | Order data |
| Contribution margin per order | Order value less cost of goods, fulfilment, payment and variable service costs | Finance, agreed definition |
| Retention curve reference | Which cohort curve this channel's customers follow | Cohort table |
| LTV on contribution | Margin per period multiplied by expected periods retained, discounted where the horizon is long | Derived from the curve and the margin |
| Payback period | Periods until cumulative contribution equals channel CAC | Derived |
| Ratio | LTV on contribution divided by channel CAC | Derived |
The table has no figures here because the definitions are the point: two teams with different definitions of margin or of "acquired" will produce different numbers from the same data and argue about the wrong thing. For paid channels specifically, the ROAS calculator gives the first-order check on whether revenue covers spend; return on investment (ROIProfit relative to cost.) on contribution over the retention horizon is what the table adds.
Two channels, one table, a different answer
An online pet-supplies retailer compares two acquisition channels in its unit economics table for March. Paid search acquired 1,240 customers for £43,400, a channel CAC of £35; paid social acquired 2,060 customers for £51,500, a CAC of £25. On acquisition cost alone, social looks better. The order data shows paid search customers placed first orders averaging £64 with £21 of contribution; social customers averaged £38 with £11 of contribution. The cohort table shows search customers following the retailer's "replenishment" curve, with 48% still ordering after six months, while social customers follow the "one-off treat" curve at 19%. Twelve-month LTV on contribution works out at £118 for search and £31 for social; payback is just under two months for search and, for social, arrives only in month eleven, when cumulative contribution finally passes the £25 CAC. The ratio is 3.4 for search and 1.2 for social. The budget decision reverses: search gets the increase, and social is kept only for the product lines whose customers reorder.
Decisions the numbers drive
Unit economics are not reporting; they are the inputs to a few recurring decisions.
How much to spend on acquisition, and where. Channels with payback inside the business's cash horizon and a healthy ratio get more; channels that never pay back get cut or confined to the segments where they do. The budget follows the table, not the platform dashboards.
How fast to grow. Payback period sets the limit: a business that recovers acquisition cost in a few months can reinvest quickly; one that waits a year needs capital to grow at the same rate. The retention and payback guide covers this constraint, and the spreadsheet growth model tutorial shows the ceiling the current inputs imply, where the customers lost each month catch up with those gained.
Where to put product and lifecycle effort. If lifetime value is limited by early churn, activation work beats acquisition work. If it is limited by order frequencyAverage number of orders each customer placed in the period., replenishment and lifecycle messaging beat both.
What a customer is worth to a partner or an acquirer. The same table, with its definitions written down, is what due diligence asks for.
The KPIA metric tied directly to an objective. tree makes these decisions legible: the north star on top, the loop stage rates beneath, and lifetime value, acquisition cost and payback as the economic layer that says whether moving the north star was worth the money.
Common economics problems
| Symptom | Usual cause | Guide |
|---|---|---|
| LTV looks healthy but the business loses money | Lifetime value computed on revenue, not contribution margin | LTV and CAC |
| Blended CAC is fine while every channel looks bad | Organic and referral customers hiding paid channel costs in the blend | LTV and CAC |
| Retention "improves" every month without any change | Immature cohorts mixed with mature ones in a single rate | Cohort analysis |
| Payback period keeps lengthening as spend rises | Marginal customers from saturated channels retained worse than early ones | Retention and payback period |
| Finance and marketing report different CAC figures | Different cost inclusions and different attribution windows | LTV and CAC |
| Curves look fine in aggregate but one segment is collapsing | No segment or channel split on the cohort table | Retention and payback period |
| Active customers stay flat while new customers rise | Rising churn netted against rising acquisition | Growth accounting |
| Targets are set without knowing what the inputs can deliver | No forward model built from the business's own cohorts | Building a growth model in a spreadsheet |
Where the data to build any of these does not yet exist, the customer analytics hub is the place to start, because every number in this section is only as good as the events and identities beneath it.
