Super Data Science: ML & AI Podcast with Jon Krohn 386: Cohort Analysis
Jul 24, 2020
Explore the fascinating world of cohort analysis, where business performance is examined beyond just totals. Discover how tracking customer journeys with timestamped interactions offers deeper insights. Learn to contrast traditional funnel metrics with this innovative approach using real-life examples, like e-commerce retention patterns. Jon also shares how cohort analysis will benefit hiring and product improvements at his company. Don't miss a free webinar invitation to master these techniques with Tableau!
AI Snips
Chapters
Transcript
Episode notes
Cohorts Track First Interaction, Not Conversion Time
- Cohort analysis groups customers by their first interaction date rather than when conversions occurred.
- This reveals how a specific signup cohort progresses through the funnel over time.
Discovery Via The Lean Startup
- Jon Krohn discovered cohort analysis discussed in Eric Ries's The Lean Startup.
- That prompted him to share the technique and explore its value for product experiments.
Compare Cohorts Over The Same Time Window
- When measuring cohort conversions, evaluate cohorts after a reasonable time window rather than by conversion date.
- Compare cohorts (e.g., April vs May) by following their progression over the same post-signup period.