Radio is finding an unexpected new role with younger music listeners: helping them escape the predictability of their own algorithms.
A new MIDiA Research study finds that 42% of live-radio listeners ages 20 to 34 use the medium for music discovery, compared with 32% of live-radio listeners overall. Human DJs, changing playlists and the shared nature of live programming are becoming more valuable at the same time personalized streaming feeds are getting increasingly good at predicting what listeners already like.
The 42% figure applies specifically to people ages 20 to 34 who already listen to live radio. It does not mean 42% of everyone in that age group uses radio for discovery. That distinction matters when measuring how large the shift really is.
The findings come from MIDiA’s new report, From Reach to Resonance: Radio’s New Role in Music Discovery, built from consumer surveys and qualitative interviews. MIDiA published the younger-listener findings September 10th.
This is not a story about Spotify suddenly losing a generation to FM radio. Streaming remains the larger monthly music format in MIDiA’s data. What is changing is more specific and potentially more interesting: some younger listeners appear to be rediscovering the value of letting another person choose what comes next.
Music Discovery Jumps Among Younger Radio Listeners
The age difference is the clearest number in MIDiA’s research.
Among people who already listen to live radio, 32% overall say music discovery is one reason they tune in. That rises to 42% among listeners ages 20 to 34, putting discovery much closer to the center of what younger radio audiences want from the format.
Radio’s biggest advantage overall is even simpler. Half of live-radio listeners cite the fact that it is free, MIDiA’s findings show.
Those numbers do not establish a full-blown “radio renaissance.” Younger listeners are not necessarily returning to the broadcasting habits of their parents or grandparents, and the modern definition of live radio reaches beyond a car dashboard.
Today’s landscape includes terrestrial stations alongside online and community-driven operations such as London’s NTS, Rinse FM and Balamii. A listener can therefore embrace the basic idea of radio without owning a conventional radio at all.
What connects those experiences is not the antenna. It is the loss of complete control.
A live station decides what plays next. The listener cannot endlessly skip, reorder the queue or ask the system to generate another playlist based on yesterday’s behavior. For an audience accustomed to nearly unlimited control over music, that restriction can become part of the appeal.
Infinite Choice Is Starting to Feel Predictable
Streaming transformed music by removing scarcity. Almost any song can be summoned immediately, while recommendation systems continuously learn what a listener plays, skips, saves and repeats.
That convenience remains enormously powerful. MIDiA finds 56% of consumers across the U.S., U.K., Canada and Australia use music streaming monthly, compared with 53% who listen to radio.
Streaming has therefore narrowly overtaken radio in MIDiA’s broader consumption data. It has not been displaced by it.
Discovery is where the tension becomes more interesting.
MIDiA analyst Laura Fisher argues that years of unlimited choice have pushed personalized feeds toward increasingly narrow loops. In her analysis of the trend, Fisher describes streaming as making discovery something an algorithm increasingly performs on a listener’s behalf.
The upside is obvious: fewer songs that feel completely irrelevant. The downside is familiar to anyone who has opened a personalized playlist and found another variation on music they already know.
Recommendation systems are designed to predict what an individual listener is likely to enjoy. A human DJ can make a different calculation entirely.
A presenter might play something because it is strange, historically important, locally relevant, unexpectedly connected to the previous record or simply because they believe an audience should hear it.
That difference turns unpredictability into a feature.
Human DJs Can Take Risks Algorithms Avoid
A good radio sequence does more than place songs beside one another.
A presenter can tell listeners who produced a track, where an artist comes from, why an older record suddenly matters again or what connects two songs that an algorithm might never pair. Context becomes part of the discovery rather than information a listener has to seek out afterward.
MIDiA identifies human curation, contextualization and dynamism as qualities that can distinguish radio from automated recommendation. Radio World highlights the 42% discovery figure while pointing to the same contrast between programmed recommendation systems and human tastemakers.
Stations such as NTS demonstrate what that can look like musically. Rather than building every show around a broad genre label or a listener’s previous behavior, individual hosts can move between obscure catalog records, emerging artists, regional scenes and new releases because the person programming the show hears a connection.
Live programming can also change while it is happening. A listener sends a message, someone calls the station, a host reacts to a record or follows one selection into an unexpected part of the catalog.
None of that guarantees better music discovery. Human programming can be repetitive too, particularly on heavily formatted commercial stations built around narrow rotations.
But the underlying mechanism is different. An algorithm attempts to infer individual preference at scale; a DJ can curate for a room full of people and occasionally decide that surprise matters more than perfect personalization.
For younger listeners who have spent most of their music lives inside recommendation systems, that can feel less old-fashioned than unexpectedly new.
Radio Now Means More Than AM and FM
The stations benefiting from this interest do not necessarily resemble traditional Top 40 radio.
MIDiA points specifically to NTS as an example of radio’s renewed discovery role becoming visible at the cultural edges. The London-based online station is known for specialist programming, deep catalog selections and shows where a DJ’s individual taste is part of the reason listeners tune in.
Rinse FM and Balamii occupy related territory, connecting live radio with local scenes, club music and communities that do not always fit neatly into mainstream playlist categories.
That helps explain why younger audiences can simultaneously be streaming natives and radio listeners. The formats are not mutually exclusive.
Someone can use Spotify or Apple Music when they know exactly what they want, TikTok or YouTube when a song enters their social feed, and live radio when they want someone else to take over.
Radio’s role can therefore be complementary rather than competitive. Streaming excels at access and personalization; a specialist station can provide a path into music the listener did not know enough about to request in the first place.
That distinction becomes particularly useful for emerging artists and niche genres. Recommendation systems can surface unfamiliar recordings, but a DJ can attach a name, scene, story or musical relationship to what the audience is hearing.
Streaming Still Leads Music Discovery
Any claim that younger listeners have rejected algorithms for radio would go well beyond the available evidence.
Other 2026 research continues to put streaming firmly at the center of music discovery. YouGov’s U.S. data, for example, finds streaming recommendations cited by 59% of Americans whose music listening has increased during the past year, compared with 43% citing radio and music channels.
YouGov’s findings reinforce the scale streaming still has as a discovery engine even while younger audiences show renewed interest in other ways of finding music.
The more interesting development is not radio defeating streaming.
It is the possibility that streaming has become so dominant that listeners increasingly want a second discovery mode built around qualities personalization intentionally removes.
One system says: here is more music based on everything we know about you.
The other says: you do not know what is coming next.
For discovery, both can be useful.
Shared Listening Gives Radio Another Advantage
Personalization is inherently individual. Radio does not have to be.
When thousands of people hear the same unexpected song at roughly the same moment, discovery becomes a shared event rather than a private recommendation generated for one account.
An August study commissioned by iHeartMedia found that 91% of surveyed Gen Z radio listeners said radio gives their social circles something to discuss.
The research comes from one of the largest commercial radio companies in the United States, giving iHeartMedia an obvious business interest in demonstrating the format’s continued relevance. Its findings are therefore best read as additional industry evidence rather than independent confirmation of MIDiA’s conclusions.
The underlying idea nevertheless points toward something streaming services have spent years trying to build through collaborative playlists, social sharing and listening activity.
Radio begins with the shared feed.
For artists, that can change the context around discovery. Hearing a song through an algorithm may produce another stream. Hearing a DJ introduce the artist, explain why the track was selected and place it beside other music can give the recording an identity beyond its position in a personalized queue.
That does not automatically create a fan. It does give the listener more information about why a song might be worth remembering.
MIDiA’s report title, From Reach to Resonance, reflects that distinction. Getting a track in front of someone and making that listener care about it are not necessarily the same achievement.
Radio’s Opportunity Is to Stop Copying Streaming
The irony of the moment is that radio spent years adapting to the platform that was supposed to replace it.
Stations built apps, added on-demand listening and competed with streaming’s convenience. Streaming services, meanwhile, borrowed radio terminology, stations, programmed playlists and increasingly passive listening experiences.
MIDiA’s argument is that radio’s next opportunity may depend on moving in the opposite direction.
Instead of becoming a slightly less personalized streaming service, stations can emphasize the qualities algorithms have difficulty reproducing: recognizable hosts, local knowledge, specialist taste, live interaction, context and the possibility of hearing something that does not fit neatly inside a listener profile.
Fisher describes the opening as a chance for radio to strengthen what makes it distinctive rather than imitate streaming. The full MIDiA report draws on surveys conducted from 2022 through 2025 across the U.S., U.K., Canada and Australia.
Streaming still reaches more people monthly in that research, and algorithms remain one of the most important engines for finding music. Nothing in the data suggests younger listeners are about to abandon them.
What the numbers do suggest is that perfect personalization may not satisfy every discovery impulse. Sometimes finding a new favorite song starts with giving up the ability to choose the next one.