Day-one ROAS lies to you
On a free-or-cheap-subscription model, almost none of a fan’s value shows up on the day they subscribe. It arrives later — a PPV unlock next week, a tip, a renewal next month. Measure a campaign by its day-one return and you’ll kill the ones that were quietly on their way to profit.
Cohorts fix this by refusing to judge a fan too early.
How a cohort matures
- Bucket by acquisitionEvery fan joins the cohort of the week they were acquired, so the whole group ages together.
- Track spend over timeFollow realized spending at D0, D7, D30 and beyond — the curve of value actually landing, not a single snapshot.
- Compare at equal ageThe honest comparison is this week’s D7 against last month’s D7. Same age, apples to apples — that’s where the trend lives.
A fresh cohort should look low — its spending hasn’t happened yet. Low-and-climbing is healthy; the mistake is treating a young cohort’s number as its final one.
Scale on the trajectory
Once you know how your cohorts mature — how a given D7 tends to become a D30 — you can commit budget on the trajectory instead of waiting a month for the verdict. Combined with Predicted Value, which feeds that expected spend to Meta early, you’re optimizing and scaling toward lifetime value from the start.
Questions
What is a spending cohort?
A group of fans bucketed by when you acquired them — usually the week. You then track how much that group spends over time: on day 0, day 7, day 30 and beyond. Because everyone in the cohort started together, you can compare like with like and watch value accumulate rather than judging a fan by their first day.
Why not just use day-one ROAS?
Because most OnlyFans revenue arrives after day one. A cohort that looks unprofitable on the day you acquired it can be strongly positive by day 30 once PPV, tips and renewals land. Judging on day-one numbers makes you kill campaigns that were actually working — cohort ROAS shows the real trajectory.
Why does a young cohort’s ROAS look low?
Because it hasn’t matured yet — the spending simply hasn’t happened. A recent cohort naturally reads low and climbs as it ages. The signal is in comparing cohorts at the same age (this week’s D7 vs last month’s D7), not in a fresh cohort’s raw number.
How does this change how I scale?
It lets you scale on where a cohort is heading, not where it starts. If your D7 numbers reliably mature into a healthy D30 ROAS, you can commit budget earlier and with more confidence, instead of waiting a month to find out.
Scale on lifetime value, not day one
Watch each cohort’s ROAS mature and commit budget with the full picture.