Design a better water bottle and test it
Overview
This question evaluates experimental design, product analytics, hypothesis formulation, metric selection, sample size/power/MDE estimation, segmentation, and risk and rollback planning within the Analytics & Experimentation domain for Data Scientist roles.
Propose 10 mutually exclusive design improvements for a commuter-focused reusable water bottle (e.g., insulation, grip, cap mechanism, filter, materials, volume markings, ergonomics, leak-proofing, cleaning ease, accessories). For each idea, specify: target segment, hypothesis, one primary success metric, and an experiment (A/B or multivariate) with unit of randomization, power/MDE back-of-envelope, expected test duration assuming today=2025-09-01, risks (novelty, seasonality), and stopping/rollback criteria.