Dina Bavli·Nov 6Digital Experiments: The Statistics Were Never the Hard PartEleven parts later, the durable lesson is not a formula. It is how to make a defensible decision before, during and after an experiment.
Dina Bavli·Nov 3Digital Experiments: When the Treatment Is a Language ModelWhat non-determinism actually changes in an experiment — and what randomization, measurement and causal design still demandA response icon1A response icon1
Dina Bavli·Oct 30Digital Experiments: The Experiment Worked. It Cost $2.5 Billion.From statistical evidence to product decisions: economic significance, Simpson’s paradox, ITT, p-hacking, and the limits of optimization
Dina Bavli·Oct 27Digital Experiments: Can We Trust This Result?The strongest-looking result in the dashboard can still come from an experiment that quietly stopped being the one you designed
Dina Bavli·Oct 23Digital Experiments: How Much Evidence Is Enough?Power, sample size and duration only make sense after you decide what evidence would actually change the decision
Dina Bavli·Oct 20Digital Experiments: You’re Not Testing What You ThinkTreatment design is the contract between what your hypothesis says, what production delivers, and what your analysis eventually claims
Dina Bavli·Oct 16Digital Experiments: Your Control Group Isn’t a Control GroupAssignment mechanisms, interference, and SUTVA — when randomizing by user quietly corrupts what you measured
Dina Bavli·Oct 13Digital Experiments: Not Every Experiment Is an A/B TestWhat to do when the intervention already shipped, randomization is impossible, or the system itself refuses to stay inside treatment and…
Dina Bavli·Oct 9Digital Experiments: Your Metric Is a Business DecisionWhat happens when the metric improves and the product outcome gets worse?
Dina Bavli·Oct 6Digital Experiments: The Outcome You Never Get to SeePotential outcomes, counterfactuals, and why causal inference begins with a value that is missing by construction.