You are Elena Bassi, chief marketing officer of Lumina Media Group, a music-and-podcast streaming company in Milan. A rival's €4.99 price cut has pushed monthly churn from 2.6% to 4.1%, and the board has given you exactly €6.0 million and ninety days to turn Lumina's own listening data into a service subscribers will stay for. Before anything else, you must decide which data to switch on.
The situation
Lumina has 3.8 million subscribers in Italy and southern Europe, paying €10.99 a month. It earned €214 million in 2025 at a 9% EBITDA margin, about €19.3 million, with 540 employees. Three weeks ago a global platform cut its price to €4.99. Since then monthly churn has climbed from 2.6% to 4.1%: about 57,000 extra cancellations a month, roughly €7.5 million of revenue a year if it holds. Finance has ruled out matching the price.
What Lumina has is behaviour: every skip, every replay, every playlist re-sorted by mood, every search that found nothing worth playing, logged for 3.8 million listeners. Most of it sits unused; until your promotion the analytics team mainly built dashboards for the unit that sells advertising. This morning the board gave you €6.0 million, capital and operating spend together, for the whole year, and there will be no more. You must fund exactly one service, with no hedging across ideas, bring a plan in ninety days, and show churn bending within two quarters of launch, or the money goes back to ad sales.
Your data team has catalogued eight signal streams, each with its coverage of subscribers, a quality score from 1 to 10 and a cost to activate, drawn from the same €6.0 million. Four service ideas are on the table: a creator and curator marketplace, a weekly discovery guarantee, a mood-and-context personalisation layer, and an audience-insight product for advertisers. Each needs particular signals, and a service whose signals were never switched on launches without the data it needs.