Streaming services received an average of 106,000 new tracks every day in 2025.
That did not happen by accident.
The music industry removed nearly every physical constraint on releasing music. It built businesses around unlimited access. Then it acted surprised when unlimited access produced an unlimited supply.
I am old enough to remember when releasing music required a physical object to move through the world.
Somebody had to manufacture it. Somebody had to decide how many copies to press. Others had to warehouse it, ship it, pitch it and convince a retailer to provide shelf space.
There was only so much room on the truck, in the warehouse and on the shelf.
That system had serious problems. It excluded artists who deserved opportunities. It favored relationships, budgets, timing and politics.
I am not romanticizing a model in which a small group could keep worthy music from reaching listeners.
However, the physical model contained something the current system largely lacks: friction.
Music Distribution Once Forced Decisions
Friction is not always bad.
If a label planned to manufacture 1,000 CDs, somebody had to believe there was a reason to press them.
If a distributor intended to ship boxes across the country, somebody had to believe those boxes were worth shipping.
Likewise, if a retailer gave one artist shelf space, it probably passed over another artist.
Those decisions were not always fair. Nor did artistic merit alone drive them.
Still, they were decisions.
The friction also filtered out some fraud, infringement, rights problems and people who could not stand behind their music.
Money, manufacturing and logistics forced everyone to answer basic questions before a project entered the marketplace.
Is the music ready?
Is there an audience?
Does anyone have a plan?
Is someone prepared to work the release?
Digital distribution removed most of those constraints. At first, that felt miraculous because it was.
Artists who could not get a meeting could suddenly reach listeners around the world.
Independent labels that were too small or specialized for traditional retail could compete. Meanwhile, genres that made no sense to national chain-store buyers could find audiences directly.
I built Syntax Creative on that opportunity, so I am not confused about the benefits of digital distribution.
But access also brought volume.
Distribution Changed From Curation to Plumbing
Once the truck disappeared, nobody had to ask whether another release would fit.
Once the warehouse became a database, almost everything could remain available.
As access became the product, parts of distribution shifted from curation to plumbing.
The pipe is not free.
Delivery, metadata management, royalty accounting, reporting and takedowns all cost real money.
So do fraud detection, rights administration and platform compliance.
Distributors should receive payment for providing those services.
The more important question is what the business model rewards.
Does a distributor make money when the music succeeds? Or does it make money merely because another artist paid to place another release into the pipe?
Imagine paying a booking agent the same flat amount for every date placed on a calendar.
Would that agent pursue an arena show instead of a child’s birthday party?
The agent would still have an incentive. However, the incentive would favor accumulating dates rather than securing the right dates.
The same principle applies to managers, labels and distributors. Compensation shapes behavior.
Many distribution services advertise that artists retain 100% of their royalties.
However, the distributor may collect an upfront charge and recurring subscription fee before the music generates a stream.
Under that model, the distributor’s primary customer is not the listener. It is the artist.
That is not automatically wrong.
Some artists need nothing more than an inexpensive upload service. They do not need a label, marketing team, publicity campaign or radio promotion.
They may not need playlist strategy or an artist-development program either. They simply want their music available.
There is nothing inherently wrong with selling access.
However, access and career development are not the same product.
The Industry Rewards Uploads More Than Outcomes
A distributor paid for every upload has less financial reason to ask whether the music is ready.
It also has less reason to ask whether an audience exists or whether anyone intends to work the release.
This is not a question of artistic permission. People should be free to make whatever they want.
Record whatever you want. Release whatever you want. Experiment, fail, improve and try again.
But the industry cannot build an economy around limitless uploads and then feign shock when the supply becomes unmanageable.
According to Luminate’s 2025 Year-End Music Report, digital service providers received an average of 106,000 new tracks each day in 2025.
The marketplace contained more than 253 million audio ISRCs.
Of those tracks, 88% received 1,000 streams or fewer during the year. Meanwhile, 120.5 million received no more than 10 streams.
That is not merely a music-discovery problem.
It is a structural problem.
The Shelf Is Infinite, but Attention Is Not
The issue is not simply that listeners heard millions of songs and rejected them.
Many of those songs never reached a meaningful number of listeners.
Music no longer competes only with other music.
It competes with every podcast, short video, television series, game and notification.
It also competes with every news cycle, work obligation and family responsibility that demands a person’s time.
The digital shelf may be infinite. Human attention is not.
That is the scarcity the industry cannot eliminate.
We solved much of the distribution problem. In the process, we created an enormous discovery problem.
Instead of slowing down, the industry continued making it easier to add more music.
Home-recording technology improved. Remote collaboration became routine. Production tools became less expensive.
Meanwhile, marketing advice became endless. Artists heard constant pressure to release more often, satisfy algorithms and remain visible.
Then generative AI opened the valve even further.
AI did not create the underlying problem. Instead, it accelerated a system already designed to reward abundant supply.
The deeper problem is simple.
Music supply can grow without a practical ceiling, while listener demand cannot.
The number of available tracks can double. The number of hours in a listener’s day cannot.
We Replaced Gatekeepers With Tollbooths
Physical distribution had gatekeepers. Digital distribution has tollbooths.
A gatekeeper says no.
A tollbooth says yes, provided someone pays.
Artists understandably resented the old gatekeepers. In many cases, they had good reason.
However, removing every gatekeeper did not remove every commercial interest.
It changed who gets paid and when.
Consider what happens when a release has no audience, marketing, strategy or story.
Perhaps it has no reason to exist beyond the fact that someone finished making it.
The artist receives access. The distributor receives payment. The platform receives another file. The database receives another ISRC.
But did the artist get closer to a sustainable career?
Did listeners receive better service?
Did the music economy become healthier?
Questions about standards quickly become uncomfortable because people hear the word “elitism.”
Nobody should want to return to a system in which a small group decides which art may exist.
But rejecting the old walls does not require pretending that discernment has no value.
Uploading a Song Is Not Launching a Song
Availability is not demand.
Distribution is not artist development.
Being present on every major service does not mean an audience has meaningfully encountered the release.
The old physical system forced people to answer many of these questions early.
Manufacturing, freight, warehousing, sales and returns placed money at risk before the music reached consumers.
The digital system did not eliminate the cost. It moved the cost elsewhere.
The cost appears when worthwhile music gets buried beneath an impossible volume of competing releases.
It appears in fraud-prevention systems, metadata cleanup, duplicate claims and takedowns.
It also appears in ownership disputes and royalty platforms that process microscopic earnings across enormous catalogs.
Digital services create monetization thresholds because parts of the long tail barely function as a commercial market.
Meanwhile, legitimate artists wonder why reaching listeners feels harder than ever.
Worldwide distribution has never been easier. Being heard has rarely felt harder.
Not Every Song Needs Global Distribution
The answer is not to rebuild every old barrier.
Instead, the industry should stop pretending that removing barriers eliminated the need for judgment.
Artists need honest counsel.
Not every song requires the same release strategy. Not every demo needs worldwide distribution.
Likewise, not every musical idea needs to become an ISRC immediately.
Sometimes the best decision is to wait.
Improve the song. Test it in front of people. Build an audience.
Strengthen the visuals. Clarify the story. Develop a reason for listeners to pay attention before adding another file to the system.
Distributors should also explain clearly what they sell.
Is the product access, administration, accounting, marketing or strategy?
Does it include audience development, or is it merely a dashboard?
There is a legitimate market for each service. However, companies should not confuse one with another.
A distribution account is not a career.
The Music Flood Was a Business Decision
The music industry spent years celebrating the democratization of distribution.
Much of that celebration was justified.
Great music reached the public that would never have survived the old system.
Independent artists gained leverage. Small labels built worldwide businesses.
Meanwhile, specialized music found listeners without needing permission from a national buyer.
But democratization without discernment can become a landfill.
We can debate royalty thresholds, fraud policies, payment models and artificial intelligence.
We can also debate whether tracks with negligible activity should generate revenue.
Those are all legitimate discussions.
Before debating the remedies, however, the industry should acknowledge how it arrived here.
The flood was not accidental.
We removed almost every constraint except the one constraint nobody can remove: human attention.
That one is not going anywhere.
An earlier version of this commentary appeared on Hypebot.