Members & Revenue · Retention Sentinel

Churn is found at T-72h, not at the renewal failure

The sentinel scores every paying member daily, picks a save-play from a library that contains no discounts, and runs it across WhatsApp, in-app and voice. Discounting is structurally disallowed — discounted cohorts churn 1.7x faster at the next renewal.

@keep

Paying members watched

214,000

Save rate

61.4%

+12.4pts · since discounts were removed

Revenue retained

₹1.40 Cr

annualised, this cycle

Renewals in next 72h

3,109

412 flagged at risk

At-risk cohorts

Each has one play, and the play is measured

Cohort	Members	Save rate	Play	Lift
Pro · month 2	18,400	61%	Research digest onboarding	+9.2pts
Pro · renewal T-72h	3,109	74%	Extension over discount	+12.4pts
Elite · usage down 50%	940	82%	Analyst round-table invite	+7.8pts
Starter · never traded	26,800	38%	Academy lesson + paper-trade nudge	+14.1pts
Lapsed 30–90 days	41,200	21%	Winback with new-feature proof	+5.6pts

Retention trend

Month-2 retention, 7 weeks

MonTueWedThuFriSatSun

71.2% → 79.4% since the sentinel took the cycle.

Play library

No discounts in it, deliberately

· One-month extension at the same price
· Audio digest so research stops feeling like homework
· Analyst round-table invite for Elite members
· Portfolio risk check that shows immediate value
· Academy lesson matched to their last losing trade
· Human callback for any payment-trust issue

Hard rules

· No offer to a member with an open support ticket
· No more than one save attempt per member per cycle
· Any economics change above ₹2L goes to Aniket
· Cancel is always one tap, never buried

See the open offer approval →

Why discounting was banned

The finding that changed the play library

20% discount cohort

1.7x

faster churn at renewal #2, and ARPU permanently reset lower.

Extension cohort

+9pts

save rate, with realised ARPU flat and no price expectation set.

What it means

A discount buys one renewal and sells the next two. Time and usefulness cost less and hold better.