01
✓ Clean ROAS you can trust
Whale-adjusted, statistically sound, aligned across every revenue source — MMPs, ad networks and stores.
Data Analytics
for UA.
This is not a reporting tool. It is a decision tool —
built on whale detection and a Bayesian statistical model.
Cohortful takes your campaign data, applies a single statistical methodology to every cohort, and returns a confidence range with a clear signal on every campaign: scale or stop.
1468
cohorts analyzed
136
analyst days saved
$6.2M
UA spend processed
PROBLEM
Standard tools give you a number.
Not the truth.
Your campaigns look profitable. ROAS is positive, you keep spending. But the money doesn't come back. Because the number AppsFlyer, Adjust and most ad networks show you includes whales who could never appear in your next cohort — and the standard ROAS calculation has no way to account for it.
This isn't a data problem.
It's a methodology problem.
Cohortful fixes it.
HOW IT WORKS
One methodology. Every cohort. Every time.
Cohortful uses a Bayesian statistical model. Instead of a single average number, it computes a probability distribution — accounting for cohort size, detecting whale users, and using your historical cohorts as a prior.
The result: not just a value. A signal you can trust.
Two analysts looking at the same data will make different calls — different revenue alignment choices, different treatments of small cohorts. Cohortful removes that variability. Same methodology applied to every cohort, automatically, without human interpretation in the middle.
METHOD 01
Sophisticated revenue modelling
We carefully align every revenue source — MMPs, ad networks and
stores — into one clean, accurate picture.
+
METHOD 02
Revenue variability analysis
We measure exactly how much whales are distorting each cohort and
how spread out the real revenue is inside it.
=
- Probability distribution, not a single average
- Whale detection and exclusion — built in
- Zero human variability across cohorts
REAL EXAMPLE
Same cohort. Same day. Opposite decision.
Cohort: 300 users · CPI $3 · D7 spend $900 · Top 10% of users generate 70% of revenue — including 3 whales at $40 each.
| Standard tools | Cohortful | |
|---|---|---|
| D7 Revenue | $252 | $252 |
| D7 ROAS | 0.28x (actual) | 0.09x–0.21x, 90% CV* (expected) |
| Signal | SCALE | STOP |
* CV — Confidence Variability range. Shows where the true ROAS is likely to fall based on Cohortful's Bayesian model, accounting for whale probability and cohort size.
Cumulative revenue share
How revenue is distributed across users in a cohort. The further the
curve bends from the diagonal, the more revenue is concentrated in a
few users.
10% users → 70% revenue
Cumulative revenue
Equal distribution
Same revenue. Same day. Standard tools take the actual number at face value. Cohortful asks what happens next — and the range says this cohort won't repeat. You scale on $252. The next cohort delivers $130.
WHAT YOU GET
From data to decision. In minutes.
02
✓ Clear signal on every cohort
Scale or Stop — with the reasoning behind it.
From D0.
03
✓ One methodology across your entire UA portfolio
No analyst variation.
No "it depends who ran the numbers."
Every cohort treated the same way, every time.
Join early access
Fill in the form and we'll be in touch within a few business days.