Two decades after Daniel Ek and Martin Lorentzon registered Spotify Technology S.A. in Sweden, the company they founded is the world’s largest paid music service, with more than six hundred million monthly active users and roughly two hundred and forty million paid subscribers. It is also one of the most controversial businesses in the global creative economy: the most-blamed application by working musicians, the most-defended by the record labels who used to fight it, and the most-studied by economists trying to make sense of how streaming changed what a song is worth. Every conversation about the contemporary music industry — about per-stream royalties, about catalogue value, about why songs are getting shorter — eventually arrives at the same Swedish application.

The Bargain That Saved the Major Labels

In 2006, when Ek and Lorentzon began pitching the idea to the major record labels, the global recorded-music industry was in the worst shape of its modern history. Revenues had fallen from a peak of around twenty-five billion dollars in 1999 to less than half that figure by 2006, with file-sharing applications such as Kazaa, LimeWire and BitTorrent doing meaningful business. Apple’s iTunes Store had stabilised the decline in legal downloads but had not reversed it.

The deal Ek negotiated — equity stakes in Spotify for Sony, Universal and Warner in exchange for permission to license their catalogues — has been re-litigated in financial filings and historical accounts ever since. The labels eventually liquidated those stakes for billions of dollars after Spotify’s 2018 direct listing. The application’s bargain was that streaming would convert pirates into paying customers without cannibalising the still-meaningful download economy. The bet largely worked. Global recorded-music revenues bottomed out in 2014 and have grown every year since, surpassing the 1999 peak in nominal terms in 2023.

The Per-Stream Economics

The most consistent criticism of Spotify is that the per-stream payment to rights holders is small. Public estimates put the average per-stream royalty at approximately three to five U.S. cents per thousand streams, depending on the country, the subscriber tier and the contract. Of that figure, the major label typically receives the largest share, with smaller percentages flowing to the publisher, the performer and any session musicians who hold publishing or master rights. The artist’s actual take-home depends on how their deal is structured.

Spotify’s defence is that the per-stream figure is the wrong unit of analysis. Each user pays a fixed monthly fee — currently ranging from a few dollars in lower-income markets to eleven or twelve dollars in the United States and Western Europe. A share of that subscription, after taxes, payment-processing fees and corporate overheads, is paid to rights holders in proportion to streams. From Spotify’s perspective the company is distributing roughly seventy per cent of revenue to rights holders — among the most generous payout ratios in the application’s history.

The frustrations of independent musicians are largely about the inequalities inside the rights-holder bucket rather than about the size of the bucket. Major labels retain a far larger per-stream cut than independent labels or self-published artists. Master-rights renegotiations in the past five years — particularly Taylor Swift’s celebrated re-recording project — have reshaped how some catalogues are priced.

The Discovery Algorithm

The most influential of Spotify’s product surfaces is the Discover Weekly playlist, introduced in July 2015. The algorithm that builds Discover Weekly — and its sibling features Release Radar, Daily Mix and the now-ubiquitous personalised home screen — relies on a combination of collaborative filtering and natural-language processing of contextual signals such as user playlists and editorial mentions. Most critically, it uses a deep learning model that compares audio fingerprints of songs to match listeners with sonically similar tracks.

This system has made Spotify the dominant tastemaker of contemporary music. A song that enters a high-profile editorial playlist or that the algorithm decides to surface widely in Discover Weekly can climb from obscurity to chart-relevance within weeks. Many genres — chillwave, lo-fi hip-hop, hyperpop, sertanejo universitário — have measurably been shaped by what Spotify’s algorithms reward. Critics describe this as “algotorialism” and argue that it pushes artists toward formats and lengths optimised for the system rather than for artistic expression.

The Podcast Detour

Between 2019 and 2022, Spotify spent more than a billion dollars acquiring podcast companies, including Gimlet Media, Anchor, The Ringer, Megaphone and Parcast. The strategy, articulated by Daniel Ek at the company’s 2019 investor day, was to capture audio attention beyond music and to escape the per-stream label economics that constrained music margins. The cornerstone of the strategy was a series of exclusivity deals with marquee podcasters, of which the Joe Rogan deal — variously reported as worth between a hundred and two hundred million dollars — was the most prominent.

The strategy worked partly. Spotify did become a meaningful podcast platform, second only to Apple Podcasts in many markets and dominant in some. The economic returns on the acquisitions, however, were disappointing. The company wrote down several podcast assets between 2022 and 2024 as part of broader cost-cutting, made significant layoffs in the podcast division and modified its exclusivity strategy to allow flagship shows to publish elsewhere. The 2024 Joe Rogan renewal was a more open arrangement, with the show distributing on YouTube and other platforms in addition to Spotify.

The Audiobook Bet

Spotify’s most recent vertical expansion is audiobooks, launched in late 2022 in the United Kingdom and expanded globally through 2023 and 2024. Premium subscribers receive a monthly allotment of audiobook listening hours, with additional credits available for purchase. The product directly competes with Amazon’s Audible and Apple Books.

Early figures shared in 2024 earnings calls indicated that audiobook adoption had been faster than the company internally projected and that, importantly, audiobook listeners exhibited higher retention metrics on Spotify Premium overall. The audiobook category has not yet meaningfully shifted Spotify’s overall financial profile but is widely expected to be a long-term contributor.

The Profitability Inflection

For most of its public history, Spotify has been described as a high-growth, low-margin business. Operating profitability has been elusive because of the high cost of content — the royalty share to rights holders is structurally large — and the cost of acquiring and retaining subscribers. The company posted its first full year of operating profit in 2024, helped by price increases in major markets, podcast cost reductions and improved advertising performance on Spotify’s free tier.

Investors have responded warmly. The share price, after a difficult 2022, recovered substantially through 2023 and 2024 as the company demonstrated cost discipline. The 2024 announcement of Spotify Wrapped’s expanded advertising integration and a new artist services product (Marquee, expanded in 2023) signalled an emphasis on monetisation depth rather than user growth.

Wrapped: The Annual Cultural Event

Spotify Wrapped, the annual personalised year-in-review feature launched in 2016, has become one of the most reliable cultural events in social media. Every December, hundreds of millions of users share screenshots of their listening data, propelling Spotify into trending topics on every major platform for several days. The feature’s marketing value to Spotify is significant; the labels increasingly orient their year-end campaigns around it; and competing services have launched their own versions with varying success.

Wrapped is also a window into the company’s data capabilities. The granularity of the personalised data — favourite genres, most-listened minutes, “audio aura” personality types — illustrates how much Spotify knows about each listener. That knowledge is the same knowledge that fuels Discover Weekly and the algorithmic ranking decisions that shape contemporary music.

Spotify and the Geography of Music

One of the most underappreciated aspects of Spotify’s influence is its role in flattening geographic distribution barriers for music. Twenty years ago, a Latin-American or African artist usually had to be signed to a multinational label and physically distributed into Western retail chains to reach an audience in Europe or North America. Today, the same artist’s music sits next to Taylor Swift’s in the global catalogue, with the algorithm — not a regional buyer — deciding what gets surfaced.

The result has been a measurable rebalancing of the global music economy. Latin-American artists now consistently appear in the Spotify Global Top 50. Afrobeats acts from Nigeria and Ghana have moved from being niche imports to charting routinely in the United States and the United Kingdom. K-pop’s expansion outside Asia, while predating Spotify, has been amplified by its algorithmic discovery surfaces. The 2024 IFPI Global Music Report explicitly credited streaming services for the increased share of non-English-language tracks in international charts.

The AI-Generated Music Question

The most disruptive policy question on the platform in 2024 and 2025 has been the rapid growth of AI-generated music. Platforms such as Suno and Udio can produce convincing songs in seconds, and a meaningful share of newly uploaded tracks to streaming services in 2024 were partly or wholly AI-generated. Spotify’s response has been a layered policy: track-level disclosure when AI is materially used, removal of obvious infringement (including voice-clone imitations of recognised artists), and per-account limits on uploads that pattern-match to automated farms.

Some commentators have noted that the existence of AI-generated music undermines the per-stream economics of human artists, because any share of streams captured by AI tracks is a share taken from the human catalogue. Others have argued that AI-generated music is simply the latest in a long history of automated music — drum machines, samplers, MIDI orchestration — and will be absorbed by the industry. The truthful answer is that we do not yet know what the equilibrium will be. We do know that the question is now real.

The Influence of a Single Application

Spotify is not the largest music service in every country. YouTube Music has a larger ad-supported audience in many emerging markets, Apple Music dominates the iPhone-heavy enterprise segment, and Tencent Music’s three apps — QQ Music, Kugou and Kuwo — together serve more users in China than Spotify globally. Yet for almost every conversation about contemporary music — about how artists are paid, about what the audience sounds like, about how the next viral track will spread — Spotify is the default reference. The application has done what very few music services have ever done: it has shaped what music is made, not just how it is delivered.

Whether that legacy is positive or negative depends on whom you ask. For listeners, the application has been transformative: more music, more easily, for less money than at any previous point in history. For working musicians, the verdict is more contested. For the global recorded-music industry as a whole, the application is the saviour that almost killed them and then rebuilt their revenues to record levels. Twenty years on, Spotify is no longer the upstart. It is the incumbent — and the next twenty years will measure how well it can keep adapting to a music economy that it, more than any other actor, has authored.