Survival analysis for conversion rates
summary
Shopify reframes conversion as time‑dependent and shows that a snapshot 5.8% (800 of 14,000) after one week underestimates eventual spenders; cohorts diverge so one game hits 17% vs 25% at six days. Analysis uses Python lifelines (KaplanMeierFitter) to get cohort survival curves, warns against rounding durations, and shows Kaplan‑Meier misleads with multiple outcomes. For plan choices the talk builds a competing‑risks solution using cumulative incidence (statsmodels) to correctly attribute conversions across plans.
AI-written summary · about
chapters (6)
- 0:00Survival analysis basics
- 3:30Time-dependent conversion rates
- 10:06Cohort comparisons with survival
- 12:20Survival analysis mistakes
- 15:35Multiple conversion events
- 21:36Key takeaways
related talks
52mSurvival Analysis of Customer Retention using Postgres, SQL and R
29mThe Power of Analytics in Game Development for Meta Horizon Worlds
26mProgressive player populations
21mData Analytics for Better Product Decision Making
27mLife, Death, and Shopping
48mHow to Use Analytics to Optimize Your Product
34mOptimizing E-commerce: Data Science for Better CX & Sales
18mLifetime value forecast – How to predict your game's revenue