It's Time to Change the Way We Consume Music

I love music, and I think the way most of us listen to it is broken.

That's an opinion, and this whole post is opinion, so let me be clear about where it comes from. From 2013 to 2015 I was an Interaction Designer/Developer at Pandora, designing how listening felt and then building it. More recently I've been working on a product in royalty recovery and rights administration: helping artists whose music earns most of its money far from where it was made find and collect what they're owed. So I've seen music from the listening side and from the money side. Neither view is pretty right now.

What streaming got right#

It's worth being fair first. Streaming solved real problems.

• Access. Almost any recorded song is a few seconds away, legally, for the price of a monthly subscription or a few ads.
• Piracy. For many listeners, streaming became easier than stealing. That was a real accomplishment.
• Reach. An artist can release music worldwide on the same day as everyone else, without a distribution deal.

I don't want to go back to a world where finding a record meant knowing someone who owned it. The question is what we lost on the way here, and whether it was worth it.

What we lost#

Ownership. Most listeners no longer own music. They rent access to a catalog, and the catalog can change. Albums disappear when licenses lapse. Playlists you built for years live on someone else's servers, on someone else's terms. Owning a record used to be part of your identity. Now a library is a lease.

Attention. When everything is available, nothing is special. The album as a deliberate, front-to-back experience has given way to tracks served in a feed. Songs are increasingly written for the first few seconds, because that's where a skip happens.

Context. Liner notes, credits, who played bass, where it was recorded. That information still exists, but the main listening interfaces mostly hide it. The people behind the music fade into a name and a thumbnail.

Fair pay. Streaming pays rightsholders from a shared pool, split by share of total streams. In my opinion, that rewards scale over devotion. Your subscription money flows mostly to the most-streamed music overall, not to the artists you actually listen to. Some platforms now also set a minimum: Spotify, for example, announced that starting in 2024, a track needs at least 1,000 streams in the previous twelve months before it earns recorded-music royalties. That rule has its reasons, but for small artists it means much of their long tail earns nothing.

What I learned designing for listeners#

Pandora was built on an unusual idea for its time. Instead of only "people who liked this also liked that," the Music Genome Project had trained analysts describe songs by their musical qualities, and the radio was built from those descriptions. You'd start with a song or an artist, and the station would explore outward from what made that music sound the way it did.

Working on that product taught me a few things I still believe:

• A thumbs up is a conversation, not a rating. The best listening interfaces let you shape what you hear in small, low-effort ways, and they respond visibly.
• Lean-back and lean-in are different products. Most people, most of the time, want to press play and trust the stream. A smaller group wants to dig. Trying to serve both with one screen serves neither well.
• Discovery needs a reason. "Because you liked X" is more trustworthy than a mysterious recommendation, and it teaches the listener something about their own taste.

Pandora is now part of SiriusXM, which acquired it in 2019. The broader industry largely went another way, toward on-demand catalogs, algorithmic feeds and playlist placement. I think some of what made early internet radio special got lost in that shift: the sense that the stream was built around the music, not around keeping you on the app.

The money side#

Working on royalty data has been humbling. The simple version of how artists get paid is: a song is played, money flows to the people who made it. The real version is a chain of societies, publishers, distributors, labels and collection agencies across many countries. Each has its own identifiers, formats, deadlines and rules.

Money gets lost in that chain all the time. The common reasons are boring:

• metadata that doesn't match across systems
• a writer or performer who isn't registered where the music is played
• works that can't be matched to an owner, so the money sits unclaimed
• claim windows that expire before anyone notices

For artists whose music travels far from home, every border adds another place for money to fall through the cracks. That's the problem the product I've been building focuses on, especially for African rightsholders with global catalogs. It's not glamorous work. It's matching, reconciling and chasing. But it's where a lot of music's real unfairness lives.

What I think should change#

These are my opinions. I hold them firmly, but I'm open to being wrong.

1. Pay by listener, not by pool. I'd like to see more experiments with user-centric payouts, where your subscription is split among the artists you actually played. It's harder to compute and won't fix everything, but it lines up the money with the relationship.

2. Bring back ownership, in a modern form. Let me buy an album inside the same app I stream it in, keep a copy that's actually mine, and see exactly how much of my money reached the artist. Bandcamp has shown for years that people will pay artists directly when it's easy and the value is clear.

3. Put the people back on the screen. Credits, writers, players, producers and the story of the record should be one tap away, not buried. That's an interface decision. Nothing technical stops it.

4. Make discovery explainable. Tell me why a song is playing. Let me steer the stream with more than a skip button. Design for curiosity, not just retention.

5. Fix the metadata plumbing. This is the least exciting item and the most important one. Shared, open identifiers for works, recordings and rightsholders, and systems that check them before money moves, would recover more income for artists than almost any feature.

6. Change our own habits. Buy a record from an artist you love. Go to the show. Buy the shirt from the merch table, not the reseller. Listen to an album front to back once a week, with nothing else open.

A listening product I'd want to use#

To make this less abstract, here's a sketch of the product I keep imagining. It's a thought experiment, not a roadmap.

• Home is a shelf, not a feed. The first screen shows the records I've chosen to keep, the way a shelf would. Recommendations live one step away, where I go when I want them.
• Every play has a receipt. At the end of the month I can see which artists my money reached, roughly how much, and how that would change if I bought an album instead of streaming it.
• Radio explains itself. A station tells me why each song is there ("same producer," "similar rhythm," "a band this artist cites") and lets me pull on any of those threads.
• Albums are first-class. Track order is respected, gapless playback is the default, and the credits and notes are part of the album page, not a hidden panel.
• Buying is one tap and permanent. If I buy a record, I can download it in a real format and keep it if I ever leave.

None of this needs new technology. It needs different incentives, and a design team that's allowed to optimize for something other than minutes listened.

What builders can do#

If you build products, you have more influence than you think.

• Design for intent, not just engagement. A listening session that ends because someone found what they wanted is a success, not churn.
• Show the money. Even an estimate of where a listener's money goes would change behavior.
• Treat metadata as product. Clean credits and identifiers aren't back-office work. They decide who gets paid.
• Give artists their data. Artists should be able to see who listens, where, and what that's worth, without a label in between.

The takeaway#

Streaming won on access, and it deserved to. But access was only ever half of what music is. The other half is relationship: between a listener and a record, and between that listener's money and the people who made it.

I've designed the listening side, and now I'm working on the money side. From both, the conclusion is the same: we built a system that's very good at delivering songs and not very good at honoring them. It's time to change that.