Unit economics

Paid ads aren't magical. They work when cost to acquire a customer is materially less than the customer's lifetime value. Before you spend a dollar, know the math.

The core metrics

• CAC: cost to acquire one paying customer = spend / customers
• LTV: lifetime value of a customer (gross profit over their lifecycle)
• Payback period: how long until CAC is recovered in gross profit
• LTV:CAC: should be 3:1 or better to be sustainably profitable

The allowable CAC

Your LTV, divided by your target LTV:CAC ratio, is your allowable CAC. If LTV is $600 and you target 3:1, allowable CAC is $200. Any campaign whose CAC runs above that is losing money.

Payback period matters too

LTV:CAC of 4:1 over 5 years looks great on paper but kills cash flow. You need 4 years of capital to fund it. Payback period under 12 months is healthy; under 18 is acceptable; over 24 is a business model problem.

The blended vs marginal CAC

Early dollars of ad spend find the easiest customers (cheap CAC). Later dollars find harder ones (expensive CAC). At scale, blended CAC rises. What matters: the next dollar's CAC, not the average.

Unit economics by segment

One tenant or product segment can have 10:1 LTV:CAC while another has 1.5:1. Blended looks fine; the mix hides a dying segment. Segment by acquisition channel, geography, product, cohort. Kill the losers.

What to do with this

• Compute LTV:CAC per segment, not blended, one 1.2:1 segment can eat the margin of 3 profitable ones while the average looks fine
• Track payback period as rigorously as LTV:CAC, a 4:1 ratio with 30-month payback is a cash trap, not a growth engine
• Use observed LTV, never modeled, until you have 18+ months of data, projected LTV is where bad unit economics hide
• Watch marginal CAC as you scale, blended looks fine while the next dollar loses money, that's how operators overspend into decline
• Kill unprofitable segments even when they feel emotionally hard, the capital reallocation to healthy segments is where real growth comes from