Measuring launches

Define success before launch day, in writing, with a time window. Vague goals ("buzz," "adoption") cannot fail, so they never inform the next bet. The first 30 days after GA are still part of the launch.

A useful success card

Example for "scheduled reports":

Compare against a baseline or a holdout if you can (A/B testing). For many T2 launches a simple before/after with cohort cuts is enough: teams that saw the announcement email versus organic discovery.

Walking through a funnel leak

Launch week: 12,000 unique visitors to the blog, 800 clicks to the feature, 90 teams activate. Vanity says "great traffic." The pre-written goal was 400 activated teams in 30 days. At day 14 you are at 140.

Diagnosis: onboarding drop at permission scopes. You ship a permission wizard in week 3; day-30 activations hit 380. Without the pre-defined metric, the team would have celebrated the blog and missed the funnel leak.

Also watch leading indicators on day 0–3: signup spikes, error budgets, support volume by tag, sales questions. Lagging revenue metrics often take a full quarter. Do not wait that long to decide if the launch mechanics worked.

Iterate, double down, or sunset

Schedule a post-launch review around T+14 days: what moved, what broke, what feedback themes dominate, what ships next versus what is won't-do. Feed that into iteration and feedback.

Some launches reveal the feature should not grow. Usage stuck under a few percent of actives, support cost high, strategic fit gone — that is a sunset candidate, not a moral failure. Sunset with the same seriousness as launch: notice period (often 60–90 days for paid B2B), migration path, stop selling immediately, firm end date. Ambiguous "we are phasing it out someday" guarantees nobody migrates.

Common mistakes

• "We'll see how it goes." If success cannot fail, you will not learn.
• Celebrating blog traffic. Measure the funnel to first value.
• Soft-launch forever, or a sunset with no date. Calendar the decision.

Check yourself

12,000 blog visitors, 90 activations, goal 400 in 30 days. Did the launch work? Not yet. Traffic is vanity. Diagnose the funnel (here, permission scopes) against the pre-written goal.

Activation is healthy but almost nobody heard about the feature. Double down, iterate, or kill? Double down on distribution. The product path works; awareness does not.

Key takeaways

• Write adoption, activation quality, outcome and guardrails before launch day.
• Measure the funnel, not just the blog. Watch day 0–3 leading indicators.
• T+14 review: double down, iterate, or sunset with a date and a migration path.

Further learning

ProdPad's GTM webinar walks release → beta → launch metrics with concrete cross-functional ownership.

Back to the map: launching.