Keep a GTM Experiment Log
A log for hypotheses, tests, metrics, and decisions.
Set the scope
Keep a GTM Experiment Log starts with a written boundary. Name the event, the source, the owner, the evidence, and the point where the work must stop.
The main risk is that a test can be run without a hypothesis or a decision rule. Record the source and time before changing a status.
Capture the evidence
A usable record includes hypothesis, audience, metric, result, owner, and decision. Keep the source beside the record so another reviewer can reconstruct what happened.
The GTM lead confirms the decision or exception. The support role can record the test, check the metric, and route decisions, but should not fill a gap with an assumption.
Separate preparation from judgment
Decisions about strategy and investment decisions remain with the authorized owner. Use three labels: proceed, pause for missing evidence, and escalate.
Apply the labels to real examples. The record should show the current state and the next owner.
Run the review
Review missing hypotheses, incomplete results, and rework. Sort by age, owner, and exception reason.
For example, a channel test ends with no decision. Record what evidence was available and who owned the next step.
Pilot one workflow
Start with one experiment. Test the checklist on a small batch and review exceptions with the owner.
Expand only after the workflow produces a reviewable record and a predictable review rhythm.
Related GTM guides
Frequently asked questions
What should stay with the accountable owner for GTM experiments?
Decisions about strategy and investment decisions and any exception outside the written rule stay with the authorized owner.
What should the support role record?
The approved action, source evidence, current status, exception, and next owner.