Define the batch before the decision
A tiny batch has explicit boundaries: which inboxes send, how many prospects receive mail, and over what window. Deciding scope first keeps the approval honest, because an unbounded launch cannot be meaningfully approved.
- Pick a subset of ready inboxes rather than the whole workspace.
- Cap prospect count well below the plan's daily capacity.
- Agree on the stop condition before sending starts.
Review the evidence bundle
The go/no-go review reads the same evidence the launch gate enforces: workspace readiness status, per-inbox SMTP and IMAP readiness, active alerts, and fresh seed-placement results against the configured threshold. Stale or missing evidence is a no-go, not a judgment call.
- Fresh passing seed placement for the sending domains.
- No unresolved critical alerts on the workspace.
- Blocked export rows resolved or explicitly excluded from the batch.
Record the outcome either way
A no-go is a useful result. Record what failed, what changed, and when the review reruns. A go decision should note who approved it and which evidence was reviewed, so later batches can compare against the first one instead of restarting the argument.