A working plan for selecting, scheduling, running, and learning from experiments. Use this to manage experiment ideas from initial review through results and follow-up.
Experiment backlogAdd each proposed experiment with enough context for another team member to understand the change and the problem it addresses.
PriorityScore each idea consistently, using the same scale across the backlog.
Schedule and ownersSet a realistic run window, name the accountable owner, and capture dependencies before committing dates.
Analysis and planSpecify the audience, pages, success measure, baseline, and expected effect before launch.
TimelineTrack the planned and actual run period for every experiment.
Results and follow-upEnter the observed change, whether the experiment was implemented, and what the evidence means for the original hypothesis.