Data
Model unit economics
CAC, payback and LTV by cohort, computed rather than asserted.
Fill it in
Acquisition, plan, revenue and churn.
The blended number hides the answer.
In months, which is the number that decides spend.
Your prompt
Compute the unit economics. Compute CAC from ALL acquisition cost, including salaries and tools, not just ad spend. Media-only CAC is the number that makes a channel look profitable right up until the team is counted. Split by channel. Blended CAC is an average of a channel that works and one that does not, and every decision made on it is wrong in both directions. Compute LTV from actual retention curves, not from one over churn. That formula assumes a constant hazard, and real churn is front-loaded, so it overstates LTV by a lot in exactly the businesses that quote it. Lead with PAYBACK in months. It is the number that decides whether spend can increase, and it does not need a survival assumption to be true.
Use Model unit economicsOpens with everything above already filled in.
Why this works
LTV computed as one over churn assumes a constant hazard, and real churn is front-loaded, so it overstates the number in exactly the businesses that quote it. This uses actual retention curves, splits CAC by channel, and leads with payback in months.