Prior: Beta(1,1) — uninformative. Updates on observed data.
Posterior Distribution
Bayesian A/B — Retention
D1 / D7 / D30 retention test
Retention Posterior
Interpretation
ARPU Comparison
Mann-Whitney U + Bootstrap CI
ARPU Distribution Estimate
Revenue Impact
Growth Tools
ROAS · LTV · Ad Revenue · Cohort Revenue
ROAS Curve Projection
Model cumulative ROAS over time across spend scenarios
Upload Cohort Data
CSV or XLSX · Each row = one acquisition date cohort
📂
Drop file here or click to browse
.csv · .tsv · .xlsx · .xls
Configuration
Column detection is automatic — override if needed
Expected format:
Columns: Day · CPI · Installs · Spend · Revenue Day 0 · Revenue Day 1 … Revenue Day N · D0 ROAS · D1 ROAS …
Each row is one acquisition date cohort. Revenue columns are per-install or total (configure below).
📉 Retention adjustment (optional — refines curve)
When enabled, the projected ROAS curve is multiplied by a retention survival curve fitted to your D1/D7/D30 inputs — so projections beyond your observed data window decay realistically instead of extrapolating linearly.
Cohorts
—
Total Installs
—
Avg CPI
—
Avg D0 ROAS
—
Avg D7 ROAS
—
Proj. Breakeven
—
ROAS Curve by Day (Avg across cohorts)
Actual observed + power-law projection
Revenue per Cohort by Day
Each line = one acquisition cohort
Budget Scenario Comparison
Projected total revenue at each monthly spend level
Projected ROAS Milestones
Fitted curve extrapolated from observed data
Day
Proj. ARPU/Install ($)
ROAS %
Status
Load data to see projections
Cohort Summary Table
One row per acquisition cohort — revenue by day, ROAS, CPI
No data loaded
Analysis & Recommendations
Load cohort data to see automated analysis.
LTV Calculator
Multiple models
LTV Curve
Ad Revenue Estimator
DAU × Sessions × eCPM model
Revenue Breakdown
Cohort Revenue Model
Monthly install cohorts + retention curve
Monthly Cumulative Revenue
Sample Size Calculator
How long do I need to run this A/B test? Power analysis for conversion, retention & revenue tests.
Conversion Rate Test
Two-proportion Z-test power analysis
Power Curve
Sample size vs detectable effect
Test Duration Scenarios
Retention Rate Test
D1 / D7 / D30 cohort sizing
Cohort Size vs Detectable Effect
Timeline
ARPU / Revenue Test
T-test for continuous revenue metric
Standard Deviation Impact
How SD affects required sample size
Revenue Guardrails
Multi-Variant Test Planner
Bonferroni-corrected sample sizes
Variant Comparison Table
⚠️ Multi-Variant Warnings
North Star Metric Dashboard
Define leading indicators, weight them, and track your composite North Star score. Paste CSV or enter manually.
Define the metrics that predict your North Star. Weights must sum to 100.
North Star Score
—
Composite score
0Target
North Star Trend
Leading Indicators Over Time
PM Playbook
Frameworks, guides, and mental models for every stage of product work
Portfolio Insight
Arturo Ruiz Valdés
CPO
Chief Product Officer · PerBlue · Vancouver, WA
Product leader with 10+ years building mobile games at scale. Contributed to multi-million dollar franchises including Summoners War: Sky Arena, Disney Heroes: Battle Mode, and Legendary: Game of Heroes.
At PerBlue I lead product strategy across multiple live titles. I've soft-launched 10+ games, from prototype and design through soft launch, live ops, and long-term monetization optimization.
I built PM Dispatch because I kept rebuilding the same spreadsheets and stat calculators in every role. This is the tool I wish existed when I started — all the prioritization frameworks, significance tests, and growth models in one place, free and without friction.
10+
Years in Product
10+
Games Launched
3
Major Franchises
20+
Toolkit Tools
Mobile GamingMonetizationLive OpsUA EconomicsSoft LaunchGacha DesignBayesian A/BProduct StrategyP&LGame Design
Arturo Ruiz Valdés
Chief Product Officer · Game Industry
Building products that reach millions. PM Dispatch is the free toolkit I wish I had when I started.
