The A/B Test Lab — Experimentation at Scale

Triage four concurrent experiments at CoEditor.io, redesign the ones that are not decision-grade, and write the Friday memo to the CEO. Practice the Kohavi-Tang-Xu discipline: separate decision-grade results from not-yet-decision-grade results, refuse to flatten honest uncertainty into political certainty, and frame the experimentation programme to Series-C investors with calibrated peer-benchmarked metrics.