Spotify CEO Says Generative AI Enables Real-Time Deep Research on 700 Million Users
Spotify CEO Gustav Söderström stated that generative AI has allowed computers to finally understand English. Previously, only about 1 million developers could communicate with computers, but now everyone can interact using natural language.
In the past, user research teams would interview only 10 users and infer insights for 761 million users. Now, Spotify allows each user to directly converse with algorithms in English through the Prompted Playlist feature, experiencing real-time adaptation to personal needs.
Market mechanisms show that the demand for personalized control drives Spotify's funding from traditional algorithm recommendations to generative AI interactions. Event-driven changes have shifted subscriptions and retention from passive pushes to user-led prompts, benefiting Spotify users and the platform that adopt natural language control, while traditional music service providers relying on static recommendations face pressure.
Source: Public Information
ABAB AI Insight
Gustav Söderström has previously driven the application of AI in recommendations and discovery as Spotify's product head, having deeply integrated machine learning into playlist generation. This emphasis on generative AI upgrades user research from sample inference to full real-time dialogue, continuing the company's transformation from data-driven to conversation-driven.
On the capital front, Spotify is concentrating resources on Prompted Playlist and natural language agents, aiming to reduce churn and enhance conversion by giving users direct control over algorithms. The motivation is to transform 700 million users from passive consumers to active co-creators in the AI era, capturing higher ARPU and ecosystem stickiness.
Similar to Netflix's early evolution of personalized recommendations and the current penetration of generative AI in consumer applications, Spotify is in a transitional phase from algorithmic black boxes to user-controlled generative factories.
Essentially a technological replacement, natural language prompts replace traditional small-sample research and static recommendations, concentrating capital from developers/algorithm experts to direct interaction with hundreds of millions of user intents, driving the consumer internet from platform-led to user-driven experience reconstruction.
ABAB News · Cognitive Law
While small-sample research seems scientific, full natural language dialogue represents a structural leap in deep user research. Selling passive algorithms to retain users contrasts with selling user prompts to gain control; the ultimate sale is the personalized pricing power driven by conversations with everyone. The company lacks data but needs a channel for users to directly command algorithms; the winners will reshape consumer relationships using generative AI.