A confidence interval on every forecast · demo below runs sample data

Run the move
before you make it.

Type a decision. NuMoon plays it out on your real data — margin, churn, LTV, attribution — and tells you exactly where you’d land in 30, 60, or 90 days. Not benchmarks. Your actual cohorts.

Your scenario
Or pick a preset
scenario · Raise tier-2 prices 10%
Ship
now
actual · 60d trailing · demo data forecast · with confidence band
MRR Δ · 30d
+$14,200
Churn Δ
+2.1%
Margin Δ
+$9,100
Confidence
87%
01 · Grounded

Your numbers.
Not benchmarks.

Every forecast runs on your actual cohorts, your actual margins, your actual attribution windows. No “industry average” hand-waves.

02 · Honest

Misses published.
In your dashboard.

Every forecast ships with a confidence band, and predicted vs. actual is tracked on every shipped move — so you audit the accuracy, not us.

03 · Reversible

Guardrails on
every shipped move.

A/B at 30% traffic. 14-day rollback window. NuMoon refuses moves outside the brand-specific risk envelope automatically.

The demo ledger · sample data · what your receipts will look like

Predicted vs. actual.
Every move gets a receipt.

demo median error
±5.6%
demo moves below
8
inside ±10% band
7 / 8
Move
Module
Window
Predicted
Actual
Δ
Pause Meta camp #4 · reallocDemo brand · DTC apparel
marketing
30d
+$9,820
+$10,140
+3.3%
Renegotiate 3PL tier-2Demo brand · home goods
ops
90d
+$15,000
+$14,200
−5.3%
Raise tier-2 pricing +10%Demo brand · subscription
product
60d
+$14,200
+$13,860
−2.4%
Apollo Vol-3 · 312 leadsDemo brand · B2C software
sales
30d
+$8,900
+$9,420
+5.8%
Sweep $48k → T-BillsDemo brand · DTC apparel
cash
90d
+$1,420
+$1,448
+2.0%
Klaviyo abandon-2 re-armDemo brand · DTC apparel
marketing
14d
+$1,400
+$1,540
+10.0%
Reorder SKU #44 · air bridgeDemo brand · DTC food
ops
30d
+$8,900
+$8,170
−8.2%
Hire 1 PT CS rep · draftedDemo brand · B2C software
ops
60d
−$1,200
−$1,420
+18.3%
↑ demo dataset · 8 sample moves. Your own ledger starts building from your first shipped move, audit-logged in your dashboard.See your own ledger
Sharper every week · trained on your approvals

The brain learns
from every miss.

Every reject signals “wrong move.” Every approve-then-revert signals “right shape, wrong magnitude.” Every approve-and-hold signals “trust this pattern.” The brain re-trains on your trust signals and gets sharper at predicting moves you would approve.

Signal 01reject

“Wrong move.”

A rejected move teaches the brain a constraint — the criteria you applied get weighted into the priors, so the same shape of move stops surfacing.

tightens the risk envelopeper brand
Signal 02approve, then revert

“Right shape, wrong size.”

A reverted move tells the brain the direction was right but the magnitude wasn’t — it sharpens the dollar estimate, not the idea.

recalibrates magnitudekeeps direction
Signal 03approve & hold

“Trust this pattern.”

A move you approve and keep is the strongest signal — the brain promotes the pattern and looks for the next instance of it across every module.

promotes the playbookcross-module
Free · no login · ~4 minutes

Try the simulator.

Connect two sources. Type one scenario. See where it lands before you commit a dollar.