Launch Cohort Replay
Reconstruct a launch funnel, separate acquisition from activation problems, and design the next measurable experiment.
Built for founders who have launched and have comparable funnel counts.
What you receive:
- Validated cohort funnel with denominators
- Observed bottleneck and uncertainty notes
- Alternative explanations and data gaps
- One experiment brief and seven-day action plan
Why this workflow: Refuses invalid funnels and incomplete cohorts instead of converting mixed time windows into confident advice.
How to use: provide counts for visitors, signups, activated users and paid users by comparable cohort; event definitions and time windows. Then review the evidence, refine the open questions, and use the finished drafts.
This is a reusable Agent Skill, not a standalone app. The core workflow works from supplied text. It does not include scraping, automatic publishing, account connections, or guaranteed commercial results. Download buyers need their own compatible AI agent; AI subscriptions and usage fees are not included. Python 3.8+ is optional for bundled checks where present. The included example is synthetic.
