Demand for music streaming apps

Over 700 million people currently use music streaming apps, according to ACM, 2024. That number is striking, but the more interesting question is why. The music was always there. Radio existed. CDs existed. MP3s existed. What streaming platforms built was not just a delivery mechanism, it was a relationship between a listener and a product that learned them, reflected them back, and made them feel understood. That is a design achievement, and it is worth examining closely.

People connect with emotional products, and streaming platforms are learning to engineer that connection deliberately.

The platforms competing for those 700 million users are not competing on catalogue size. They all carry tens of millions of tracks. They are competing on how the product feels to use, how quickly it earns trust, and how effectively it builds habits that survive a bad week or a cheaper competitor. Those are emotional and behavioural problems dressed up as technical ones, and the teams treating them as purely technical are losing ground to those that understand the difference.

What follows is our reading of this market through that lens, not a catalogue of features, but an analysis of the psychological machinery underneath them, and what it means for anyone building or competing in this space.

How Big Is the Music Streaming Market?

The scale of music streaming is genuinely difficult to absorb. Hundreds of millions of active users across dozens of platforms, with paid subscription tiers generating billions in annual revenue. Spotify alone carries more than 600 million monthly active users. Apple Music, Amazon Music, YouTube Music, Tidal, and Deezer collectively serve hundreds of millions more. The market has matured quickly from a curiosity into one of the dominant forms of media consumption globally.

Growth has not slowed, but its character has shifted. Early adoption was driven by convenience, streaming was simply easier than downloading, and the catalogue was vast. That phase is largely over. Penetration in developed markets is high, and the users who were going to switch from ownership to subscription have largely already switched. Growth now comes from retention, from winning users away from competitors, and from expanding into markets where smartphone penetration is still climbing.

The financial model compounds the pressure. A platform running a free tier is subsidising those users through advertising revenue or through the hope that enough of them will eventually convert to paid subscriptions. Every churned user costs more than a non-subscriber might suggest, because the resources spent acquiring them, through marketing, through personalised recommendations, through free trial periods, are not recovered if they leave. Retention is the economic foundation the whole model rests on.

Who Are the Major Players and What Do They Offer?

The platforms at the top of this market have converged on a remarkably similar feature set. All of them offer offline listening, cross-device sync, curated playlists, social sharing, and algorithm-driven recommendations. The technical differences between them are real but narrow. Tidal offers higher-quality audio compression. Apple Music integrates tightly with iOS. Amazon Music bundles with Prime. Spotify has the largest social and third-party developer ecosystem. But none of these differences is large enough, on its own, to explain why a user would stay or leave.

The free tier question

Spotify and YouTube Music run ad-supported free tiers. Apple Music and Tidal do not. This shapes who arrives at each product and what they expect. A free-tier user is not the same person as a paid subscriber, their tolerance for friction is different, their emotional investment is different, and the design needs to account for both states rather than treating them as one audience.

Where they actually differ

The real differentiators are harder to describe in a feature list. They live in onboarding, how quickly the product earns trust, how gently it asks for information, how soon it delivers something that feels personal. They live in the small moments between tracks: the recommendation that lands, the playlist that fits a mood the user had not named yet. These are emotional design decisions, and they have more influence on retention than any codec comparison.

Why Content Libraries No Longer Explain Growth

For roughly the first decade of music streaming, catalogue size was a genuine differentiator. Securing exclusive content, signing licensing deals with major labels, and getting to 30 million tracks before a competitor got to 20 million, these things moved the market. That race is functionally over. Every major platform now carries enough music that virtually no listener will exhaust it. The content library stopped being a growth driver the moment it became effectively infinite from the user's perspective.

A library of 100 million tracks is meaningless if the platform cannot surface the right three for this moment.

What replaced it as the battleground is curation and contextualisation. A listener does not want 100 million tracks. They want the right track for this Tuesday morning commute, or this particular mood they cannot quite name, or this dinner party where the guests have incompatible tastes. The platform that consistently gets that right builds a habit. The one that consistently misses builds an argument for switching.

We see this pattern across consumer apps more broadly. Growing download numbers can mask poor retention, and teams often congratulate themselves on acquisition figures while overlooking how many users are quietly leaving at day three, day five, day seven. The music streaming market has largely learned this lesson at scale, the platforms investing most heavily in discovery and contextual recommendation are doing so because they understand that keeping users is harder and more valuable than finding them.

Track retention at three days, seven days, and thirty days separately. A drop between each interval tells you something different about where the product is losing people and why.

The Emotional Loops Built Into Streaming Platforms

The best streaming platforms are delivering emotional experiences that happen to contain music. The distinction matters because it changes what you measure, what you build, and what you protect when budget pressure arrives.

An emotional loop in a streaming app typically works like this. The user arrives in a particular state, tired, energised, distracted, focused. The platform reads behavioural signals: what they skipped, what they replayed, how long they stayed on a playlist before switching. From those signals it constructs a response, a recommendation, a Discover Weekly, a mood-tagged playlist, that meets the user where they are. When that response lands accurately, the user feels seen. That feeling is the loop. It is what brings them back.

People connect with products that feel emotional rather than merely functional. The indicators of that connection are measurable: session time, frequency of return visits, what users say about the product on social media, and whether they recommend it to people they know. These behaviours do not come from a product that works. They come from a product that resonates. A music app that works is one that plays audio without crashing. A music app that resonates is one a user would feel a small loss without.

Session time and return visit frequency are worth separating. Long sessions indicate depth of engagement in a given moment. Frequent returns indicate that the product has built a habit. Both matter, and a strong product shows both.

How Onboarding Hooks Users Before They Hear a Note

On average, 77 per cent of apps lose their daily active users within the first three days of download, according to our own work across consumer products. Even strong products see a 40 to 50 per cent retention drop after day three. The gap between those two figures is where good onboarding lives, it is the difference between a product that earns a habit and one that gets deleted quietly on a Sunday afternoon.

Music streaming apps have been thoughtful about this, though not always consistently. The first screen a user sees carries an enormous amount of weight. It has to communicate what the product is, earn enough trust for the user to continue, and ask as little of them as possible, all before they have heard a single second of audio. Ask too much too soon and you create the anxiety that pushes people away. Ask too little and the product feels aimless.

Headspace, which operates in an adjacent space to music apps but faces the same onboarding challenge, earns trust within the first three seconds by removing the two things that most commonly break trust on wellbeing apps, a sign-up gate and a diagnostic questionnaire. There is no "how are you feeling today?" prompt that presumes an intimacy the user has not yet established with the product. Music platforms can take the same lesson: the goal is to deliver value before asking for commitment, not to interrogate the user before letting them in.

Getting the app store listing right also plays a role before onboarding even begins. If a listing accurately describes what the app does, the people who download it already understand what they are getting. They arrive with correct expectations, which means the product does not have to spend its first moments correcting a mismatch. That alignment between the listing and the actual experience is one of the simplest ways to reduce early abandonment.

Personalisation as an Emotional Retention Tool

Personalisation in music streaming is often discussed as a technical achievement, recommendation algorithms, collaborative filtering, audio analysis. Those things matter, but the reason personalisation retains users is emotional, not technical. When a playlist arrives that matches a mood the user had not articulated, or a weekly recommendation contains an artist they had forgotten they loved, the product has demonstrated that it knows them. That experience of being known is what keeps people.

What personalisation signals to the user

Personalisation signals investment. A product that personalises well is telling the user that their history, their taste, and their behaviour have been noticed and used to serve them better. That is a form of attention, and attention creates attachment. The alternative, a generic experience that treats every user identically, communicates the opposite. According to Deloitte, 69 per cent of customers say they are more likely to purchase from a brand that personalises experiences, which suggests the commercial stakes are not abstract (Deloitte, Consumer Trends Smartphone Usage).

The limits of personalisation

Personalisation can tip into surveillance if it surfaces things the user did not realise the platform had noticed. A recommendation that feels eerily accurate is a small pleasure. A recommendation that reveals how closely the platform has been watching can feel invasive. The emotional response to those two experiences is entirely different, and the boundary between them shifts depending on the user's existing trust in the platform. Getting this right is a design problem, not an algorithm problem.

When API Design Shapes the User Experience

Most users have no idea what an API is, and yet the quality of a streaming platform's API affects their experience directly. Streaming apps rely heavily on third-party integrations, smart speakers, car systems, fitness apps, social platforms, games. The quality of those integrations depends on the quality of the API that exposes the platform's data and controls to developers outside the core team.

Spotify's developer ecosystem is a useful reference point here. By building a well-documented, accessible API, Spotify allowed third-party developers to build tools that extended the platform's reach into contexts the core team would not have prioritised. That expanded the emotional surface area of the product, users encountered Spotify in their running apps, their games, their smart home systems. Each encounter reinforced the habit and deepened the relationship.

The design implication is that API quality is not purely a technical consideration. A poorly designed API produces integrations that feel clunky, slow, or disconnected, and those integrations carry the platform's brand into experiences the platform cannot control. A listener whose Spotify disconnects from their car speakers three times on a commute is having a Spotify experience, and it is a bad one. The API is the product, in those moments.

When evaluating third-party integrations, test them in the context the user will actually experience them, in a car, on a phone with variable signal, through a smart speaker in a kitchen. The controlled environment test and the real-world experience can be very different things.

How Platforms Measure Emotional Engagement

Standard engagement metrics tell an incomplete story. Session length, daily active users, monthly active users, these numbers reveal that someone was in the product, but say nothing about what they were feeling while they were there. A user who spent 40 minutes replaying one song because it was the only thing managing their anxiety is technically more engaged than a user who spent 20 minutes discovering three new artists they loved. The numbers suggest the opposite of what actually happened.

The more useful signal is behavioural rather than aggregate. Dwell time on particular features, skip rates on recommendations, replay frequency, the speed at which a user moves between tracks, these tell you something about emotional state. A user skipping rapidly through recommendations is communicating frustration. A user replaying the same track five times in a row is communicating a very different emotional need. Both are available in the data if you choose to read them that way.

Social signals matter too. What users say about a product on social media, whether they recommend it to friends, and whether they share playlists publicly are all indicators of emotional connection rather than mere functional satisfaction. A user who tells a friend they have to try a particular platform's Discover Weekly is giving that platform something no advertising campaign can reliably produce. That behaviour does not come from a product that works correctly. It comes from a product that made someone feel something worth sharing.

Silence is not approval. Users frequently abandon apps without submitting a complaint, cancelling formally, or leaving a review. Their departure is quiet, and a team that mistakes the absence of negative feedback for product health will miss the problem until the retention numbers start to move. Proactive measurement, tracking what users do rather than waiting to hear what they say, is the only reliable way to catch these patterns early.

What This Means for Competing in the Streaming Market

Anyone building in this market now is building into a space with established habits, strong brand loyalty, and well-funded incumbents. The technical barriers to entry are lower than they were, cloud infrastructure, licensing frameworks, and developer tools have all matured. The emotional barriers are higher. A new entrant has to disrupt an existing relationship between a user and a product that has spent years learning them.

Where the gaps actually are

The gaps are in in emotional fit for underserved audiences, the classical listener whose discovery needs are poorly served by algorithms trained on pop consumption patterns, the fitness user whose running pace should be setting the tempo rather than the other way around, the listener in a non-English-speaking market whose local music is treated as secondary. These are real unmet needs, and they are emotional before they are technical.

Building for the long term

Consumer app revenue compounds over time through volume, retention, and habit, not through pricing that reflects the cost of building the thing. A streaming platform that tries to recoup its content licensing costs through subscription pricing will price itself out of the markets where it most needs to build a user base. The financial model only works at scale, and scale only comes from retention, and retention only comes from emotional connection built over time. That is not a cycle that rewards impatience.

• Identify the audience whose emotional needs the current platforms serve poorly
• Design onboarding that meets users in their actual state, not an idealised one
• Build personalisation that feels like attention rather than surveillance
• Measure emotional engagement through behaviour, not just session counts
• Treat the API and third-party ecosystem as extensions of the core experience

Conclusion

The demand for music streaming apps is driven by the experience of being understood by a product, the daily loop of opening an app, finding something that fits exactly where you are, and feeling that the product is, in some small but real way, on your side. That experience is designed. It is the result of decisions about onboarding, personalisation, measurement, and emotional architecture that most users will never consciously notice but will feel every time they open the app.

The platforms winning users are the ones that have stopped treating music delivery as the product and started treating emotional resonance as the product. The technical layer exists to serve that, not to define it. A recommendation algorithm that surfaces the right track at the right moment is remarkable engineering in service of a fundamentally human need, the need to feel heard, even by software.

For anyone building in this space, the lesson is the same as it is across consumer apps more broadly. Download numbers are not retention. Retention is not engagement. Engagement is not emotional connection. Each step in that chain requires a different kind of design thinking, and the platforms that understand all four are the ones building something that lasts.

If you are building a streaming product, a consumer app, or any digital experience where retention is the thing that actually matters, let's talk about your product.

Frequently Asked Questions