Product Launch

Turn a launch date into an evidence-gated operating plan. A launch is ready only when the promise, product path, owners, measurement, support, and reversal plan are ready at the risk level of the change.

Workflow

1. Establish the launch type, launch date and timezone, audience, desired outcome, primary meaningful event, baseline, approved proof, channels, scope, dependencies, risk level, one accountable owner, backup, and decision-maker.
2. Classify each input as observed, inferred, proposed, stale, or conflicting. Record capability maturity as shipped, planned, or blocked. Claims also require source | as_of | confidence | allowed audience/surface.
3. Read readiness.md. Scale the review to risk; a routine announcement does not need the controls of a migration, paid launch, regulated claim, or irreversible deployment.
4. If the positioning spine or required proof is unapproved, stop the launch plan after producing only a draft messaging spine and readiness register. Do not create publish-ready channel assets or imply that copy can repair product, URL, instrumentation, security, capacity, or proof gaps. Route positioning approval to brand-positioning.
5. When the gates are clear, read sequence-assets.md. Derive every milestone from the actual launch date; assign owner, backup, dependency, review gate, and status to each deliverable.
6. Immediately before external execution, obtain explicit authorization for the exact publishing, founder posting, relationship reply, outreach, spend, or live-system action. Earlier approval to plan or draft is not execution approval. Without authorization, prepare drafts and a launch-day command sheet only.
7. Read measurement-postmortem.md. Verify the meaningful event and guardrails before launch, monitor launch-day decisions, then convert results into specific keep/change/stop actions.

No-Go Rule

A critical risk blocks launch until an authorized decision-maker closes it or explicitly accepts a reversible residual risk. No-go conditions include an unsupported material claim, broken core journey, missing meaningful-event instrumentation, unavailable owner or backup, unresolved security or compliance blocker, unsafe capacity, or an untested reversal path for a material live change.

Output