Growth · Insight Engine
An insight that doesn't change a decision isn't one
Each finding below carries its evidence, a confidence level and the specific thing that changed because of it. Four of the five have already rewritten a play the team was running.
Findings this month
18
Acted on
14
Plays removed entirely
2
discounting, CTV spend
Mean confidence
91%
Language, not intent, is the activation blocker
Owner: Growth
Evidence
- ·Telugu/Malayalam cohorts abandon broker linking 2.4x more often
- ·Same cohorts convert above average once a native-language voice walkthrough runs
Decision taken
Make native-language voice the default for 6 states, not an experiment
Confidence
94%
Discounting Pro renewals destroys the next renewal
Owner: Growth
Evidence
- ·Discounted cohorts churn 1.7x faster at renewal #2
- ·Extension cohort holds ARPU flat with a 12.4pt better save rate
Decision taken
Remove discount from the save-play library entirely
Confidence
91%
Members who ask Stoxo 3+ questions in week 1 never churn in month 1
Owner: Product
Evidence
- ·Churn 2.1% vs 19.4% for zero-query members
- ·Effect holds across plans and cities
Decision taken
Onboarding's only job becomes: first 3 Stoxo questions
Confidence
88%
Compliance rejections cluster in 4 phrase patterns
Owner: Compliance
Evidence
- ·82% of blocks are implied returns, past performance, urgency, or missing disclaimer
- ·Creative Forge now pre-filters those at generation time
Decision taken
Bake the four patterns into every generation prompt and creator brief
Confidence
96%
Academy certification is the cheapest paid-plan funnel we have
Owner: Academy
Evidence
- ·18% of certificate earners open a paid plan page within 24h
- ·Cost per paid start ₹142 vs ₹913 on Meta prospecting
Decision taken
Fund Academy as an acquisition channel, not an education perk
Confidence
86%