Event Funnel & Cohort Retention Simulator

Event Funnel

Install → Signup 55%

Signup → First action 60%

First action → Purchase 22%

Daily churn (retention decay) 0.12

Installed

–

Funnel conversion

–

Day-7 retention

–

Day-30 retention

–

How it works

Simulated users are spawned at the top of the funnel and pass through three real gates — install→signup, signup→first action, first action→purchase — rolling against each stage's conversion rate independently, exactly like a production event-tracking pipeline counting real taps. Every user who reaches Purchase joins the current signup cohort; every ~7 simulated seconds that cohort closes and a fresh one starts. Each closed cohort's ongoing activity is then modeled with a stretched-exponential retention curve, the standard shape analysts fit to real day-N retention data.

retention(t) = exp( −(t/τ)^p ),  τ = 1/churn,  p = 0.65
funnel conversion = purchases / installs  (measured, not modeled)

Install → Signup / Signup → First action / First action → Purchase — the three stage conversion rates; lower them to see drop-off pile up earlier in the funnel.
Daily churn — shrinks τ, steepening every cohort's retention curve so users fall off faster after converting.
Auto-spawn — toggles the steady stream of simulated installs so you can freeze the funnel and read exact counts.

Real-world relevance: this loop — a measured event funnel feeding a modeled cohort-retention curve — is exactly how mobile analytics tools (Amplitude, Mixpanel, Firebase) report onboarding health and long-term stickiness side by side.

Passed gate Dropped off Retention curve