How Data Scientists Use Bayesian A-B Testing for Smarter Decisions

What this episode covers

Episode 115 of The Data Science Podcast with Fexingo dives into Bayesian A/B testing and why it's replacing traditional frequentist approaches in industry. Lucas explains how a simple probability framework gives data scientists a more intuitive answer: 'there's a 95% chance variant B outperforms A by at least 2%,' instead of a confusing p-value. The episode walks through a concrete example from an e-commerce checkout flow, showing how prior distributions, likelihood updates, and posterior sampling work in practice. Luna pushes back on the subjectivity of priors, and Lucas shares how Netflix and Spotify use Bayesian methods for multi-armed bandit problems. They also discuss the 'peeking problem'—why frequentist tests break when you check results early, and how Bayesian methods handle it gracefully. A quick, grounded introduction for anyone who has run an A/B test and wondered if there's a better way. #BayesianABTesting #DataScience #ABTesting #BayesianStatistics #PriorDistribution #PosteriorDistribution #MultiArmedBandit #Ecommerce #Netflix #Spotify #MachineLearning #Statistics #DataDriven #PeekingProblem #Probability #FrequentistVsBayesian #FexingoBusiness #Technology Keep every episode free: buymeacoffee.com/fexingo

PodParley-generated summary based on available episode metadata and transcript content.