RevenueOS
Specimen · a claim ledger, as the client receives it
This is the shape of the report a client would receive. Every program shows its verdict, its interval and the counts behind them. Each program is measured against leads deliberately held back from it, and the portfolio line pools the two programs that were designed identically.
Specimen: a seeded demo tenant. Every number below is fixed and illustrative. RevenueOS does not publish customer results, and this is not one.
In a live workspace running a proof program, this page would be recomputed from that workspace’s own enrolled sends each time it is opened, and the same question would always get the same answer. Here the numbers are frozen so the example stays readable. Experiment enrollment and the proof engine are built but switched off in production today, so no customer program is being graded yet.
The scale: every estimate and interval on this page is a difference in points per lead between a program and its own holdout, so 0 means no difference. The front page’s angle examples use a ratio instead, where 1.0× means no difference.
How to read it: the verdict is whether the interval clears zero. The intervals are anytime-valid, so you can read them whenever you ask, not only at a scheduled stop. Where a read cannot be distinguished from zero, the estimate is withheld instead of printed with a caveat.
+1.6 pts per lead
anytime-valid 90% CI +0.4 to +2.8 pts · valid under continuous monitoring
n proof 859 · holdout 1150 · floor 380
Enrollment-weighted portfolio estimate — a claim about these programs together, never any single program.
meetings · confirmatory · not decision-bearing
anytime-valid 90% CI +0.1 to +1.3 pts · valid under continuous monitoring
+2.3 pts per lead
anytime-valid 90% CI +0.6 to +4.1 pts · valid under continuous monitoring
n proof 412 · holdout 549
meetings · confirmatory · not decision-bearing
anytime-valid 90% CI +0.2 to +1.7 pts · valid under continuous monitoring
anytime-valid 90% CI -1.4 to +5.2 pts · valid under continuous monitoring
n proof 63 · holdout 158
meetings · confirmatory · not decision-bearing
anytime-valid 90% CI -2.1 to +3.8 pts · valid under continuous monitoring
anytime-valid 90% CI -0.9 to +2.8 pts · valid under continuous monitoring
n proof 447 · holdout 601
meetings · confirmatory · not decision-bearing
anytime-valid 90% CI -0.6 to +1.4 pts · valid under continuous monitoring