The Art (& Science) of finding the most important AI Use Cases - Part 2
Case Study - Spotify’s Discover Weekly Feature
In the last post, we discussed how teams should go about identifying AI uses. Today, we discuss one of OG AI features that changed the trajectory of its company.
Spotify’s Discover Weekly
Spotify could’ve taken the easy route with standard recommendation algorithms—“people who liked this also liked that.” Instead, they built something more intentional: Discover Weekly—a personalized playlist that updates every Monday morning, curated specifically for each user based on their unique listening habits. It wasn’t just a smarter recommendation engine—it became a weekly ritual, blending AI with user delight in a way that felt almost magical. Back in 2015!
This was one of the first true examples of personalization that too, at the scale of the internet. Discover Weekly didn’t just suggest more of what users already liked—it predicted what they might love, but hadn’t discovered yet. In the world of recommendation systems, this rare magic is known as serendipity.
But what made this use case truly stand out wasn’t just the tech—it was the deep insight into a user journey: Monday mornings, when people are craving something fresh, energizing, and uplifting. This is applied AI at its finest—solving a real, emotionally resonant need in a way that builds habit, loyalty, and even a bit of joy.When Discover Weekly launched in July 2015, it went crazy viral. It quickly became one of the most celebrated examples of personalized AI in consumer tech—praised not just for its innovation, but for the way it resonated with users.
Here are some powerful numbers that tell the story:
1 billion tracks were streamed within just 10 weeks of launch
5 billion tracks streamed in the first year
Over 2.3 billion hours of music have been streamed through Discover Weekly since its launch
A 30% increase in user engagement was observed
Nearly half of Spotify’s user base tried the feature shortly after its launch—and many turned it into a Monday morning ritual
As one Spotify exec put it:
“It’s not just one of our most successful personalized features—it’s one of our most successful product features, period.”
Five years of discovery and engagement through Discover Weekly
What truly sets Discover Weekly apart is the deep insight into a specific user moment: Monday mornings—a time when people are naturally looking for something fresh, energizing, and motivating. This wasn’t the result of copying what others were doing. It was the outcome of first-principles thinking.Spotify didn’t just build a smarter recommendation engine—they created a weekly experience that felt personal, timely, and even a little magical. Nothing like this had existed before, and that’s exactly what made it so powerful.
💡 Takeaway:
The most powerful AI use cases don’t start with models—they start with a clear understanding of your user’s journey. It meets users where they are, when they need it most. That’s the power of first-principles thinking.
When you combine first-principles thinking with a deep understanding of your users, you don’t just deliver features—you create habits, emotions, and loyalty. That’s the true art (and science) of applied AI.