A GTM experiment brief that produces useful learning
Define a channel or message test clearly before activity begins.
Understand the decision
State the uncertainty to reduce. Connect a defined audience, action, expected signal, and the larger decision that signal informs. Ordinary campaign activity is not an experiment when no plausible result would change the plan.
Put the practice to work
Limit simultaneous changes. Record audience source, message, offer, channel, timing, owner, sample rule, and stopping condition. Choose measures near the uncertainty: reply quality may test problem resonance better than impressions, while activation may reveal more than sign-ups.
Review and improve
Decide in advance what outcomes support continuing, revising, or stopping. Preserve context and unexpected effects, including a clear negative result. That record prevents an unsupported idea from returning later under a different name.