Data-Driven Growth & Experimentation
Hypothesis-driven content strategy: packaging tests, format experiments, reading results without fooling yourself.
- Lectures
- 7
- Total runtime
- 6h 18m
- Attendance minimum
- 3h 9m
- Prerequisites
- B.CC — Years One to Five
Lecture Playlist
Ordered as a teaching progression, not by popularity. Attendance is recorded once half of a lecture has been watched.
- Lecture 14: Causality (MIT 14.310x, Esther Duflo) The foundation — what a causal claim actually requires, and why randomization is the answer to selection bias 76 min Attended
- Lecture 15: Analyzing Randomized Experiments (MIT 14.310x, Esther Duflo) Analysis and design — reading an experiment's result, and sizing a test before you run it 79 min Attended
- Marketing Experiments: From the Basics to the Cutting Edge Applied testing — A/B, multivariate, and adaptive "earn and learn" designs in a marketing context 61 min Attended
- BIG Faculty Webinar with Professor Mike Luca & Max Bazerman: How organizations learn from experiments Building the practice — how an organization turns experiments into decisions instead of decoration 57 min Attended
- Mike Luca on the role experiments play in addressing discrimination Experiments on platforms — how design choices produce measurable outcomes you never intended 32 min Attended
- Integrating Explanation and Prediction in Computational Social Science — Duncan Watts Capstone: the humility lecture — why post-hoc explanations of success feel true and predict nothing 54 min Attended
- Practicum · working creator Clickbait is Unreasonably Effective A working creator re-packages his own published videos and shows what the view curves did: Lecture 3's packaging test run live, under Lecture 1's standard for what counts as a causal claim (hold the content fixed, change one thing) 19 min Attended
Every course reserves one practicum seat, listed last and marked P. It is filled by someone who has shipped the work — here, Veritasium — because the university lectures establish why something works and the practicum shows it being done under real constraints.