Good morning. I'm Bryan Goldberg, head of investor relations at Spotify. It's my privilege to welcome everyone in the room and on the live stream to our 2026 Investor Day. Over the next several hours, the Spotify management team is going to be providing a detailed look into our business, where we've been, where we are today, and the growth opportunities we see ahead. Before we begin, we are going to be making some forward-looking statements that are subject to risks and uncertainties. Please review the safe harbor statement on the slide behind me. If reading's not your thing, then simply listen to the Snoop Dogg.
Hey, hey. What's cracka lackin'? It's your boy, Big Snoop Dogg, and you already know the vibe. Welcome to Spotify Investor Day. Before we slide off into all of that good stuff, I got to lace you up real quick with a little something we call the safe harbor. Yeah, stay with me. Look, today you're going to hear some forward-looking statements. That means we talking about where the company might be headed, projections, expectations, estimates, and all that future talk. Let me keep it real with you. The future ain't set in stone. Things can change. Actually, results could come out looking real different, materially dif ferent from what's being discussed today.
If you really want to understand all the risk and what could shift that picture, go ahead and tap into the company's filings with the SEC. That's where it's all laid out. On top of that, you're going to be hearing some non-IFRS measures. I know that sound fancy, all that means is we sometimes look at the numbers in a way that helps tell the fuller story. If you want to see how that all lines up with the standard numbers, check the appendix in the presentation. It's all broken down for you right there. One more thing before I let you all ride. Nothing you hear today means there's any obligation to come back and update these forward-looking statements later on if things change.
Today is today. That's how we rock it. All right? Just like that, you're safely harbored. Let's get into it. Over to the team. Over and out from the D-O-double G.
Please welcome to the stage Co-CEOs of Spotify, Alex Norström and Gustav Söderström.
Good morning, everyone. I'm Alex.
I'm Gustav.
Whether you've been following our journey since the beginning or taking a fresh look today, thank you for joining us. Today is our opportunity to share with you why we are so excited about value creation for Spotify. Spotify is in the business of delivering creativity and culture to the world, helping artists, creators, and authors connect with audiences and grow their careers. Gustav and I, we've been with Spotify since the early days. I got the haircut to show for it. I'm not sure if you're aware, but Spotify turns 20 this year, and this marks our third Investor Day. We've seen the company thro ugh so many chapters, and we can say with confidence that the opportunity ahead has never been greater. Let us tell you about where we are today.
We're in 184 markets with 761 million active users, nearly 300 million are subscribers. This massive group of paying passionate fans is not only double the size of any other music service, but also much more engaged than any other music service.
That's why almost two-thirds of all premium music streams happen on Spotify. Far more than our subscriber base would suggest.
The majority of these subscribers come back over 25 days a month, with over 100 million of them spending more than 28 days a month with us. That's 100 million people with almost 100% DAU over MAU .
That also means that 3.5% of the world subscribes to Spotify, giving us more than 96% of the world left to win over.
There have been more than 10 billion playlists created, and our global user base now generates 3.4 trillion taste signals every day across all of our verticals and surfaces, and that's up 43% since just the top of this year.
Spotify is also one of the world's largest music communities, with nearly 500 million people subscribing to someone else's playlist and over 45 million people enjoying collaborative playlists each month, and almost 50 million people listening together in real-time using Jam.
Spotify is also more than music today. Over 500 million people, half a billion people, stream the video podcast on Spotify, up nearly 50% year-over-year.
In just a couple of years, we've already captured roughly 20% of the audiobooks market in the U.S. As of today, you all are going to hear much more from our members of our team about where we've been and where we're going. As most of you know, Spotify was born during the piracy era, when the music ind ustry was in free fall. From day 1, we set out to solve problems that others thought were unsolvable. This mindset has defined everything since. Before we look ahead, Gustav, why don't you take us back a bit?
Let's do it. Let's go back to 2018, our first Investor Day. The big question that many of you were asking yourselves then was, "Okay, Spotify was right about access versus downloads, but in a world of competitors with walled gardens and massive distribution advantages, how can you possibly win?" Our answer was counter-positioning, built on three ideas. First was Premium. We focused on maximizing reach while lowering the barrier to entry and building engagement. Second was ubiquity. Rather than sort of building a walled garden of our own, we chose to be everywhere, across devices and ecosystems. Third was personalization. We invested early in machine learning as the core driver of retention. That's pretty obvious, maybe, in the age of AI, but it was not obvious back in 2013 when we started.
It wasn't.
Of course, investors also wanted to understand how do you win in a category where everyone licenses the same catalog? Our solution was that we would use the freemium tier to build global scale, then we would win by out-innovating on the product itself. We didn't just rely on one or two differentiated features. We built a platform that continuously turns user behavior into new product experiences and ways to create value. Spotify became the R&D department of the music industry, where new offerings are tested at scale, adopted across the ecosystems. When we met again for our next Investor Day in 2022, four years later, we had delivered on that strategy and quite a bit more, actually.
The new question was, okay, Spotify is now a great product, but will it ever truly be a great business with solid margins? Even more, can you ever actually go beyond music? Our answer to this question was the Spotify machine. The idea was pretty simple. It was to take all the capabilities that we had built for music and then extend them into new verticals, creating new business models, combining music, podcasts, and audiobooks into a single unified experience.
All right. Let's take a look at what we have delivered since 2022. Let's start with users. We've added nearly 340 million people to the platform. The pace of this progress puts us well on the path to 1 billion users. We grew our subscriber base by over 110 million to reach 293 million subs. This makes Spotify one of the largest subscription businesses in the world. In revenue, we had a currency-neutral CAGR of 18%, reaching EUR 17 billion in 2025. That almost reaches the long-term target we set. We achieved a 32% gross margin in 2025, up from 25% in 2022, beating our 30% goal. We shifted operating margin over 18 percentage points from around negative 6% to a positive of almost 13% in 2025. Importantly, this has flowed through to cash.
Free cash flow went from close to zero in 2022 to nearly EUR 3 billion in 2025. We've got a scaled, profitable business with a large and growing audience and multiple engines of growth. The Spotify you see today is very different from the one you saw four years ago. How did we achieve this while continuing to grow users, subscribers, and revenue at high rate? Well, first, we renewed our contracts with the rights holders in the music industry. We did this twice to make our music business model much more sustainable. Second, we grew our marketplace business at a much higher velocity, which contributed to improving our gross margin. Third, we reimagined our ads and podcasting businesses. While these were pretty hard pivots, they're putting us in a much better position long-term.
We licensed one of the most compelling catalogs of audiobooks and added it to Spotify Premium in over 20 markets. Next, we launched global improvements to Spotify Free, making it even more competitive as a growth driver. Finally, we redesigned and reduced our org size to match our roadmaps, which resulted in operating leverage. There are countless other wins we're not covering here. The progress is evident.
I love those audiobooks. All right, let's jump into what you're asking now in 2026. We think the first question you have on your minds is, okay, you're a real business now. Is this it? Is it mostly just incremental growth from here on? Our answer, absolutely not. You will hear much more throughout the day, but here are two key takeaways that I want to take with you. Spotify is not a single-product company, and our next chapter is not just about scaling the verticals that we already have. Sure, it's about continuing to improve our current monetization, but also about opening up new ways to monetize. Second, Spotify's opportunity to grow goes well beyond just pricing. There is actually no such thing as an average user, which you will hear more about today.
Engagement and willingness to pay follow a power law, and we see clear opportunities to capture more value from our most engaged users. The second question that I'm guessing a lot of you are asking is, what does AI mean for Spotify? Is it a tailwind or is it actually a headwind? Guess what? It's a tailwind. Our view on AI is both pretty straightforward, but also somewhat contrarian to what some in the Valley may tell you. We do not believe that the advantage for us comes from owning our own general reasoning frontier model, what is called an LLM. Instead, we believe that general domain reasoning, like maybe coding, math, will stay widely available, especially given the massive investment and intense competition that you are seeing between the hyperscalers right now.
Instead, our bet is that buying this capability on the open market will continue to be the most cost-efficient strategy. The advantage from us is that it comes from applying general intelligence to something that is proprietary, dynamic, and deeply personal. It essentially comes down to the unique data and context that only Spotify has about users. What is often called taste. Taste is continuously refreshed, it's grounded in real behavior, and very hard to replicate. Rather than training general large language models, we are training what we call our Large Taste Model. We have also referred to this as our large personalization model in the past. You know it's the same thing. This strategy lets us retain significant advantages and still benefit instantly from any frontier advances, and the very aggressive price-performance-cost curve that we're seeing there.
We will cover this in much more detail today.
At Spotify, we don't capture people's attention with empty calories. We do it by delivering an experience that people truly value, and they stay because the product earns its place in their day, every day. After nearly two decades, we understand that growth is driven by engagement and retention. Today, most platforms compete for time on app, whether it's minutes per day or hours per session. Spotify performs well on that dimension. From 2021 to 2025, streaming hours per subscriber increased globally by 10%. We've never believed that all time is created equal. This is where Spotify breaks from much of the consumer internet. We are not trying to spark a binge. We are trying to become a trusted companion across more moments in people's lives.
We measure value not only by how long people listen, but how often they choose to return. Days in a month, not just minutes in a session. That distinction matters. Spotify fits into the morning commute, the study session, the workout, the dinner prep, the evening wind down, and the story before bed. Importantly, a growing number of those moments are spent with Spotify around the world, and people listen across every device. Their iPhone, Garmin watch, Tesla, Roku, and PS5, just to name a few. In fact, the number of times that you come back to us in a month and the number of devices you use might be among the most important metrics that we monitor. This is how we think about engagement on Spotify.
Now, the other key unlock to our effectiveness is our verticals. Music came first and remains the core of Spotify. Layering podcasts on top adds more days of engagement. Most recently, we've introduced hundreds of millions of people to audiobooks, and the result has created our most active and well-retained group of all. Those who use all three verticals, music, podcast, and audiobooks, are engaging with Spotify almost every day of the month.
It's pretty crazy.
It's pretty remarkable. Now we have improved all three dimensions, use cases, devices and verticals. We've done so consistently over the last five years. This is exactly the type of engagement and retention that drives the growth of Spotify. We're not stopping there. We recently introduced fitness, which we believe has the potential to become a meaningful vertical in its own right. We have our eye on several more. Even our pricing is a retention story. We've raised prices multiple times now with minimal churn. Because users, they see the value and stick around. What you'll see today builds on that trajectory, continuing the Spotify growth story. What does this all mean for our goals through 2030?
What does it mean?
What does it mean? Well, here's what we expect to deliver. A mid-teens revenue CAGR, a gross margin of 35%-40%, an operating margin above 20%, and strong growth in free cash flow. We also remain committed to our North Stars, 1 billion subscribers, $100 billion in revenue, and over 40% in gross margin. Let me give you more insight into our approach to capital allocation. We will keep a strong balance sheet. This provides us with the flexibility to execute and reinvest in our strategy. We will also continue to explore inorganic investments, like we always have, that will strengthen our existing businesses and accelerate our strategy. We will keep countering dilution from stock-based compensation through share repurchases. We will also continue to see strong cash flow generation.
Even with these efforts, we plan to start returning excess capital to our shareholders. Christian will elaborate more on these plans. We think this is an exciting outlook.
Very exciting.
Now I want to take you into how we plan to achieve these goals. We're betting on four big ideas. The first is that the world operates as a power law. For Spotify, that opens up significant monetization. Let me unpack what that means. In the industry, we often talk about averages, like Gustav said. There is no such thing as an average user. It's just a way to represent very different underlying usage and willingness to pay. This is our demand curve. So far, Spotify has focused on capturing two sections of that curve. With our scale-free tier, we capture and monetize the very long tail of people using Spotify, supported by ads. And this is what gives us a massive universal TAM. Next, our pay tiers. You meet the demand with a range of options.
For example, the EUR 6.99 Student plan and the EUR 12.99 individual Premium plan. That means today we have almost 300 million people on what is one of the world's largest subscription platforms. We have expanded this Premium segment both up and down by making more plans more affordable with Family, Student, Duo, and others. Also by raising price, as we've added more features and offerings into Premium, again, with minimal churn. There are lots of people in Premium who are prepared to spend even more based on their incredible usage. Now, for some time, we have been talking about add-ons that would let us capture the full head of that curve. Now, we all thought that music was going to be first out the gate.
I certainly did.
It turns out that it was our newest vertical, Audiobooks, that got there first. Let me tell you about Audiobooks+. As Audiobooks and Premium rolled out across our first 22 markets, we saw listening grow significantly. What stood out was a group of highly engaged users consistently hitting their monthly usage limits. That was a clear signal of unmet demand. We introduced Audiobooks+, allowing those users to extend their listening by purchasing an add-on for additional hours. The response has been strong. In less than a year, more than 1 million users are already paying for Audiobooks+ on top of their Spotify subscription. Even more importantly, these are very valuable users. These users, these subscribers, have lifetime values that are multiples of Premium-only users. They spend far more, and they stay much longer.
We see this power law of usage across all of our verticals, creating significant potential. In many industries, this type of demand curve typically gets divided between different players.
Exactly.
We are one of the few companies in the world that has the business models and the skills needed to cover the entire demand curve. From the outside, Spotify has always looked like a simple 2-step funnel, Free and Premium, large in size with a massive TAM, but ultimately a capped ARPU. What we're outlining today is actually a platform for a diverse set of higher ARPU products, each with a smaller individual TAM, but much higher ARPUs. Thanks to our scale, this represents many millions of users and a tremendous upside. This is just the beginning of our next monetization chapter. Today, Charlie will show that music fans will be able to interact with their music in entirely new ways. Something that we know that you've all been waiting for.
Maybe.
Yeah. As Roman and Maya will explain, we will also allow users to create truly personal podcasts for the first time, in addition to unlocking creator membership add-ons. Finally, Owen will talk to us about what is next for Audiobooks, which I won't spoil, but it turns out that even Audiobooks+ didn't satisfy the full demand curve of our user base. Book lovers are asking for Audiobooks++ and +++ and so on.
You totally spoiled it now, by the way.
I did. Across all of these use cases, the opportunity is consistent. We will give our most engaged users the ability to pay for more so that they can use more, control more, and access more.
This brings us to the second big idea that we're doubling down on. Spotify is moving from single player and passive to multiplayer and interactive. When our users adopt new behaviors, even before the product fully supports them, we pay attention. That signal is actually one of our most valuable advantages. We see what people do, what they choose, what they share, and even what they say in natural language these days to us.
That's right.
Years ago, we saw something interesting in how our playlists were used. One person would start a playlist, then their friends would text back and forth to add songs, and then the original creator could share it with the group. At the time, Spotify wasn't designed for collaboration, but the signal was very clear. People didn't just want to listen, they wanted to listen together. We added on-platform messaging and created the shareable playlist. Now it's obviously everywhere, with more than half a billion people subscribing to someone else's playlist. Then we went all in on these network effects, introducing Jam, a way for users to listen with their friends and maybe family in real time. With nearly 50 million people jamming together, we are seeing even more engaged users.
Now, the collaborative playlist is also another popular Spotify invention, which turned playlisting into something that groups build, share, and actually revisit. With almost another 50 million people streaming from a collaborative playlist, we've made music something that you just do together. All of this results in a pretty simple fact. You simply need to be on Spotify, end of story. Our experience is not about amassing followers or meeting strangers. It's about your real life connections and hanging out with your people. The next big idea, surprise surprise, centers on AI. The world is moving towards generation, where our users are in control. Our goal is to give them exactly that. Spotify launched in the era of curation, where the world's music was organized by our passionate users into tens of billions of playlists.
Building one of the richest collections of human taste signals that was ever assembled, actually. Next came a recommendation, powered by algorithms and machine learning. We turned those billions of curation signals into features like Discover Weekly, maybe Release Radar, and personalized surfaces that made discovery effortless for hundreds of millions of people. Now, we're entering the era of generation, where the experience isn't just selected for you from a catalog, it's actually shaped by each of our users in real-time around their taste, context, and intent. For example, with generative AI, for the first time in history, it's now possible to create and consume truly personal media, individual media. A podcast that was made for you, an audience of you, or maybe at most, you and a friend.
When you ask Spotify to, I don't know, give me the news, we will deliver a daily brief built around your interests, your inbox, your calendar, even the presentation that you're going to have later today. This media obviously needs a trusted private space, not a public feed. We've built exactly that. We're already seeing people take advantage of this with tools like Save to Spotify, which we launched a couple of weeks ago, where you can use your agent of choice, like maybe Claude Code, OpenClaude, Codex, to create individualized content. Like a narrated travel plan for your trip to Barcelona, or maybe dive deeper into a subject like token economy, based on what you've shared with your agent, like notes, files, articles, and more.
Today, there is no media player for both public and private content, or put differently, there is no media player for the generative age. We believe Spotify will become that. Another way to think about this generative era is that computers finally understand English. This puts infinite creativity and control back in the user's hands. So for example, Prompt a Playlist, where you can ask it to create a Playlist using your entire listening history, going up to 20 years back almost, or to create one around who won Best New Artist. A fan favorite, "Make me a playlist with main character energy and send it to me every Monday morning to get me pumped up and confident for the week." This is a real user example, not made up.
It's yours. It's yours.
If you want to shape Spotify more broadly, Taste Profile lets you tell Spotify who you are and even who you want to be. No one else offers this level of control. Spotify truly becomes your media service.
That brings us to our fourth big idea, time well spent. The most precious commodity that we each possess is our time. Yet, across much of the internet, too many platforms treat time as something to be captured and not really respected. Feeds keep you endlessly scrolling, often leaving you feeling regretful and empty. We believe there's another path. We enable experiences that generate love and bring joy and inspiration and insights, so that the time that you spend with us doesn't deplete you, it energizes you. We actively measure not just time spent, but how much of the time users consider valuable versus regrettable. Spotify consistently ranks among the most valuable time that people spend online. In our surveys, users report feeling good almost 90% of the time that they spend on Spotify.
That stands in sharp contrast to broader industry patterns, where users, especially younger ones, report actually regretting up to 40% of the time that they spend on other platforms.
Even more, in some cases.
Even more. In some cases, they regret up to close to 70% of their time.
It's pretty crazy to me.
This is what makes Spotify different. We are not maximizing engagement at any cost. We're building for something more satisfying, and more importantly, durable. This is not just good for users, it's good for business too. It turns out that all time is not created equal, and people are willing to spend and pay for time that they value. With that, let me invite another Gustav. I'm living in a world of Gustavs. This one is Gustav Gyllenhammar, to show you how this all translates into subscriber growth around the world.
Thank you both. Hi everyone. Historically, we have framed our opportunity to the smartphone. Music is not tied to one device. It is universal. Hardware is expanding beyond the phone, and software is redefining what listening can be across every surface. Now, when we think about TAM, our ambition is total population, because everyone has a relationship with music. What I want to show you is how our freemium system plays out, free fueling paid, country by country, culture by culture, all around the world. We are very proud that over 3% of the world's population pays for Spotify on a monthly basis. The global average hides an even more important story. When you look market by market, you can see that we're already capturing a significantly higher portion of the potential audience.
In Sweden, our most established market, paid penetration is approaching 50% of the population. That's more than 10 times the global average. We don't think this is just Swedish pride, because other established markets are not far behind. In established markets like the U.K., Netherlands, Germany, and Australia, penetration typically ranges from 25%-50%, while faster-growing markets, they are earlier on this journey. Brazil is at 12%, the Philippines is at 4%, just above the global average. Our growth does not come from a single lever. It comes from a repeatable global playbook that we adapt locally to maximize growth. Simply put, first we attract listeners, then we build engagement. From here, we convert to subscriptions, then we deepen retention further, and over time, we grow revenue per user. Every market, every stage, it works.
This is exactly where our conviction for 1 billion subscribers come from. It is not a top-down time exercise, but built from what we've already seen playing out in the markets where this model has been running the longest. AI is accelerating every stage of the playbook with faster localization and a hyper-personalized premium funnel every single time a user opens the app. Our conviction is also grounded in something structural that we are very proud of. Spotify's business is both global and local in a way that is extremely hard to replicate. Global scale gives Spotify brand power, data advantages, network effects, while local expertise makes Spotify feel native to each market with the right artists, cultural moments, tiers, pricing, and partnerships. It all starts with winning listeners.
Before someone becomes a power-user, subscribes, buys an add-on, they choose Spotify. They first come to Spotify because they're looking for music. They choose to stay on Spotify because it quickly feels like the most personal, social, and culturally connected place to listen. They also choose Spotify because it is free. It can be easy to forget how magical this experience is. Remember when you first downloaded the app, you had instant access to all the world's music. Free is where it all begins. 71% of subscribers spend time on free before converting to Premium. Thanks to our investments in the free experience, we are creating an even bigger base of free users globally, migrating them to our paid product over time.
Our updated free experience that rolled out globally last year, it has improved engagement and retention across the board, resulting in double-digit engagement lifts with outsized positive impact for Gen Z. Since our last Investor Day, we have increased our number of free users by roughly 90%. This free growth is accelerating, and we believe we are on the right path to reach 1 billion users before 2030. Music is universal, but taste is local. What works in Stockholm doesn't automatically work in Nashville, Rio, Mumbai. Only 47% of music streamed on Spotify is in English. Local music tops the charts and dominates listening in most countries around the world. In 90% of the cases, the top artist is performing in a local language.
To truly win, our product has to speak the language of the market, both literally and culturally. That's why Spotify DJ, one of our highest engagement features, has just expanded from English and Spanish into French, German, Italian, and Brazilian Portuguese. We will continue to expand this feature to more languages in the future. This expansion will happen much faster. What once took years of market-by-market investment now takes a fraction of the time. We are combining nearly 20 years of taste data with the speed of AI. This is where our global platform also becomes a local advantage. We combine superior discovery, personalization, and brand relevance with deep local understanding, strong partnerships with artists and other creators. Take our recent BTS album campaign. It was uniquely Spotify.
You saw millions of fans engaging in-app while thousands of fans enjoyed our concerts and fan events we had in New York, Seoul, São Paulo, Mexico City, Tokyo, Jakarta, and Manila. It was truly one global moment made local and interactive everywhere. When we get this right, product love, brand love, they reinforce each other. Word of mouth and artist advocacy, they are two of our most important acquisition levers. They're leading to astounding growth in every market. Free is where the relationship begins. At our festival, premium is the headliner. It is the best place to be a fan. It's a clear destination for the most engaged users who want more from music, podcasts, audiobooks, video, and fitness. The freemium model is all about driving this conversion.
10 years ago, only 32% of our U.S. and Canadian users paid for premium. Today, 60% are on premium. This is off of a base that is four times the size it was a decade ago. In Brazil, we've doubled the conversion rate since 2016 from 22% to 44% today, while the user base has grown 14 times, resulting in a 27x growth in subscribers. In our earlier stage markets across Asia, Middle East, and Africa, we're seeing massive free user growth. These two forces, significant and strong conversion in established markets alongside massive free user growth in developing markets, means our global conversion rate has actually stayed around 40% for many years. This is exactly how the model is supposed to work. This doesn't just happen.
Over the last 20 years, we have fine-tuned one of the world's highest-performing freemium engines. We know how to best design our free and premium tiering system, when to bring premium marketing and offers to the right user at the right time, and now AI is turbocharging this. Over time, users can move through the portfolio as their needs and willingness to pay grow, student to individual to duo to family, and eventually into add-ons and higher value experiences. A user who enters on individual and moves on to family and then buys an Audiobooks+ subscription is our highest LTV cohort. To help you understand the full picture, I want to visit three markets. They show the arc of our model from established, to scaling, to enormous long-term opportunity. We will start with the U.S.
The U.S. is the world's biggest music market, and we have been capturing more of it every single year. Our free tier is increasing engagement and retention amongst U.S. users, and this engagement is converting. MIDiA data shows that we have grown our U.S. premium market share by 8-10 percentage points over the last six years with gains every single year without exception. Where we have established a large subscriber base, like in the U.S., is also where we have the most permission to introduce additional paid products. Our U.S. Premium users, they are deepening the relationship with Spotify. They're adding podcasts and audiobooks on top of music, meaning more listening days per month, stronger retention year-over-year, and this deep subscriber engagement leads to higher ARPU as we can serve our power users with products like Audiobooks+.
With paid penetration still under 20% in the U.S., there's a lot of runway left here. Across our established markets show the same similar story. Now let's move to Brazil. Brazil is a passionate music market where culture moves fast. Spotify has built exceptional brand strength in Brazil. What matters most is what we're building beneath this growth. It's infrastructure that compounds. Local artist partnerships, culturally resonant editorial, campaigns that embed Spotify into the moments people care about. Like our Q1 summer hit celebration, over 300,000 fans voted in-app, crowning Anitta as the queen of the Brazilian summer, and local payment options. Pix reaches almost every adult in Brazil, we were one of the first merchants to integrate Pix Automático, the recurring version in 2025, at a 90% lower cost than taking credit cards.
Since 2022, we have doubled net additions in Brazil. 2025 was our strongest year ever for subs intake, and this momentum is accelerating. In 2026, revenue is going to grow with more than 30% year-over-year. Let's close out with India. 150 million smartphones are shipped in India each year. According to World Data Lab, India is contributing the largest share of new consumer growth in the world this year, surpassing China for the very first time. India's potential is enormous. When we launched in India in 2019, we were the 10th player in one of the most competitive streaming markets in the world. Today, we are the leader in audio streaming. India now sits alongside the U.S. as one of our largest markets by MAU, and it has some of the highest brand love of all of our markets.
The subscription economy opportunity is enormous. Even with unique dynamics such as low data costs, high ad tolerance, today, less than 10% of our Indian users are on Premium. This isn't a challenge. It is our runway. User growth, especially when coupled with brand strength, market leadership, has been a consistent predictor of subscription growth across all our markets. We are adapting with local entertainment partnerships, deeper payments integration like UPI autopay, which accounts for over 90% of our intake at an 85% cheaper effective rate than credit cards. We've also introduced a ne w pricing tier, Premium Platinum, priced at just over 2x the cost of our standard premium offering. The early signals are very strong. In just a few months after launch, with minimal marketing, more than 7% of our subscribers are already on this higher value tier.
Even here, where we're still building the base, the demand curve is already showing us there is no such thing as an average user. It points to significant upside as we scale further. Overall, conversion is accelerating in India. In 2025, we added three times as many net subscribers as in 2022, and our subscriber count is seven times higher compared to when we presented at the last Investor Day. With 1.4 billion people, rising consumer spending, we can imagine a future with more than 150 million subscribers in India. If you zoom out, this playbook is working everywhere. This growth creates the foundation for the next part of the story, monetization. Over the last 20 years, we've proven people will pay for music globally.
For much of our history, we have kept pricing low, which helped us first introduce streaming at scale. Three years ago, we hadn't moved the baseline price in 15 years. Now, pricing strategy is showing up in the P&L. We're continuing to scale while raising price, because we keep expanding value, churn remains low, while the revenue impact is highly accretive. Pricing will continue to be a tool in our toolbox. We think long and hard about when to increase. You can expect us to continue on this journey. As we've said before, we always want the user to win. They should always feel like they're getting more value. This is how we ensure that their time and money is well spent. It's the perfect recipe for lifetime value.
With highly retained and engaged subscribers, this revenue opportunity goes beyond price. This is the power law in action. Markets with the highest level of engagement and where we can see the strongest subscription performance, the highest level of podcast and audiobook adoption, and the highest willingness to purchase additional item. Today, we have nearly 300 million Premium subscribers. This is the starting point, not the ceiling. Every market in this company's history has followed the same arc. We know this playbook works, and now AI is helping to accelerate every stage. The path to 1 billion subscribers is not just a long-term aspiration. We believe this is an extrapolation of what we've already proven, continent by continent, market by market, user by user.
To talk about why people stay, why they spend, and why they love Spotify more than any other audio product on Earth, I'm going to pass things on to Nicole.
Thank you, Gustav. Every December, Wrapped takes over. Social feeds fill up at levels usually reserved for championship games or the biggest award shows. A Spotify product experience quickly becomes something people use to express who they are and connect with others. That's because Spotify is where product and culture meet. Last year alone, Wrapped generated more than 620 million shares. Why is Spotify able to deliver a marketing campaign that becomes a cultural phenomenon year after year? It starts with a simple idea, no regrets. Every day, we make deliberate choices so that time with Spotify feels worth it. We take ordinary moments and make them more engaging, more personal, and more meaningful through the experiences we create. This belief and the choices we make as we build trust with the user shapes how we design.
It keeps us focused on experiences people value and deliberately choose to come back to, making us more efficient in how we invest our time and effort. AI is accelerating that dramatically. It is helping us move from signal to insight faster than ever before, interpreting user intent, identifying what matters earlier, and giving our teams a clearer view of where Spotify can create even more value. That matters because great product design starts with human judgment. AI does not replace that judgment, it strengthens it. It helps us focus our teams on the opportunities most likely to deepen the relationship between Spotify and our users. That means we can deliver a better Spotify to more than 760 million users while staying disciplined about what deserves to be built and scaled.
When users feel their time with Spotify is time well spent, that encourages exploration. When people feel Spotify is genuinely enriching their lives through music, they become much more open to discovering podcasts, fitness, audiobooks, and other experiences across more moments of their day. In April, we partnered with Morning Consult and Burson on a major brand affinity study across six major markets. What stood out was not just the ranking, but how people described their experience. Across major platforms, Spotify ranked number one for time well spent. Asked the inverse, which service they never regret using? Again, Spotify ranked number 1. Two questions, same answer. The signal's even stronger with Gen Z, one of our fastest-growing audiences. Gen Z was our strongest generational cohort on time well spent, with almost 90% reporting overall satisfaction with Spotify.
For Gen Z, Spotify is not just something they use. It consistently feels worth it. That matters, especially at a time when parents, myself included, are increasingly mindful of the time their kids and teens spend with screens. Not only is this the right product philosophy, it is also good business. It builds trust, it strengthens habits, and it makes people more likely to recommend the experience to others. The single biggest source of new Spotify users is still a recommendation from an existing user. That only happens when the product consistently delivers. Brand love is not just a sentiment metric. It is acquisition leverage. It improves the efficiency of paid marketing, reinforces conversion from Spotify Free to Spotify Premium, and strengthens retention once people subscribe.
In fact, just last week, we launched a special on-platform experience to celebrate Spotify's 20th birthday. In the first six days, almost 100 million people engaged in the experience, helping drive our single biggest day of subscriber intake ever. When you choose what to listen to, shape a playlist, or go deeper into a song's DNA, you are investing attention. On Spotify, you leave with something, a mood shifted, an idea sparked, or a connection deepened. We hear from users again and again how Spotify is their companion throughout the day, even enhancing the experience of chores, cooking, and focus time. Music, books, and podcasts are not disposable formats. They a sk for attention and intention, and that effort creates meaning. Because in a world of infinite content, the winners will not be the platforms that take the most time.
They will be the ones that make time feel most valuable. Now, Nat will show you how that philosophy comes to life in the product. Thank you.
Thank you. At Spotify's scale, product development starts with noticing user behavior, whether it's a playlist people share, a prompt they repeat, a feature used in ways we just didn't expect. The opportunity is not to chase every signal, but to identify the ones that can improve retention, conversion, and long-term value. That is the model we're focused on. Observe demand, build the right experience, measure whether it deepens habit, then scale what strengthens the business. SongDNA and About the Song are good examples. Fans don't just want the track, they want the meaning, the collaborators, the samples, the influences, and the creative process behind it all. We turned that desire into a native Spotify experience.
Since launching in March, more than 70 million subscribers have used SongDNA, generating more than 265 million interactions, with Gen Z leading the way as early power users. About the Song shows the same pattern, with three out of four users rating the feature positively. The important point is not just that users like these features. It is tha t deeper context creates higher intent sessions, turning listening into discovery, fandom, and repeat engagement. That gives users more reasons to stay inside Spotify, and it gives creators and rights holders better ways to connect with the audience already there. That is what separates meaningful discovery from empty engagement, and it also shows how time well-spent changes the way we build.
When you are clear on the kind of time you are trying to create, you stop investing in features that simply fill the space, and focus instead on experiences people actively choose, return to, and share with others. That focus becomes even more powerful when the path from vision to execution gets shorter. AI is dramatically reducing the time and cost of turning ideas into real product experiences. Months become weeks become days, and work that once took entire planning cycles can now happen far faster. As Nicholas will discuss later, at Spotify, any em ployee can now create a new Spotify experience and instantly make it available in our internal app store for others to try. This kind of AI-powered prototyping expands who can contribute ideas.
It helps teams align around something tangible, and it accelerates learning about what deserves deeper investment. This speed and focus benefits the product and the users in three ways: more control, deeper connection, and richer personalization. First, control. People want Spotify to respo nd to them, not the other way around. That is why we build features that let users tailor the experience around the moment that they're in. Video toggle is a simple example. Users can decide when video adds to the experience. That kind of control builds confidence. You can see it in the way people spend time curating their exact Olivia Dean or Justin Bieber playlists, and then later unwind with a Bobby Lee or Amy Poehler podcast.
When people can customize the experience, they trust it, return to it, and make it part of their daily lives. Second, connection. As generative AI increases the volume of content, connection to the people behind the art becomes even more important. Spotify goes beyond helping people find something to listen to or to watch. We help them understand who made it, why it matters, and where to go next. Third, personalization. Taste Profile gives people more direct control over how Spotify understands them across music, podcasts, and audiobooks. Increasingly, users can engage with Spotify in natural language and have the experience adjust to them in real time. This is where AI becomes a true tailwind. Spotify is not just predicting what might come next.
It is becoming more interactive, listening, interpreting, and responding while building a deeper understanding of who you are and how your taste is changing. That shifts the relationship. The more you invest, the more Spotify feels personal, useful, and relevant. As the platform grows, every new single helps us understand users better and increases the value they get. The richer that system becomes, the smarter and more effective Spotify gets across every vertical you will hear about today. To take you deeper into where it all began and where the opportunity is still expanding, here is Charlie Hellman, Head of Music.
Thanks, Nat. Even at the scale we've achieved, hundreds of millions of subscribers, Spotify continues to strengthen its role as the primary driver of growth for music. In 2025 alone, we paid more than $11 billion to the music industry, increasing more than 10% year-over-year. That's more than double the growth rate of the music industry's other revenue sources combined. Our all-time payouts now exceed $70 billion. No retailer in history has put this much money back into music, that, in turn, has created the capacity for more successful artists than ever. The number of artists generating more than $100,000 a year from Spotify alone has more than tripled since 2017.
That earning potential is powered by our subscriber base expansion, our personalized recommendations that connect any type of artists to the right fans, and our marketplace promotional tools that allow artists to bet on themselves and propel their own success. Our promotional tools for artists have become the most relied-upon marketing engine across the industry for both new releases and older catalog, fro m the biggest labels in the world to the do-it-yourself artists. That value we're providing to the industry is reflected in the increased demand we see for these tools. The gross profit contribution of marketplace tools has grown 4X in the last four years. This is what we do. We identify hard problems in music for artists, for fans, and build solutions that move the industry forward.
We did it with piracy, creating a legal experience better than the illegal one. We did it with streaming, turning the access model into real income. We did it with discovery, helping more artists find and connect with audiences, even as the volume of music has grown exponentially. It's only continued. Technology's been expanding music creation for decades, from bedroom production tools to do-it-yourself global distribution. More artists have been able to reach audiences and create than ever before. Spotify's always benefited from that growth, and our job is to make sure that that growth translates into real opportunity for artists. Today, we're entering a new moment. Generative AI is accelerating creation at an unprecedented pace. In many ways, this feels like a continuation of the same trend.
More music, more choice, more ways for fans to discover what they love. At the same time, things are starting to feel like the Wild West. Alongside new original work, there's a surge of covers, remixes, reinterpretations built on existing music. Without a rights system in place, artists can lose control of their work, and value can be created without it flowing back to the people who made it. That matters most for established artists, the ones who spent years building iconic bodies of work. This is exactly the kind of problem Spotify was built to solve. Done right, fans' desire to reinterpret music is an opportunity for revenue, for driving attention back to the original work. Catalog growth has always been good for Spotify.
We want to make sure that this next wave is good for artists, too. That means building a legal, responsible experience that's better than today's rogue alternatives, where fans can create, and artists can choose to participate and benefit. Today, we're announcing landmark licensing agreements with Universal Music Group and Universal Music Publishing Group. For the first time, fans will be able to legally create covers and remixes from participating artists' and songwriters' catalogs with both the original artist and the songwriter sharing in the value created. This will launch as a paid add-on, and our Premium users will be able to try it out first. All users on Spotify will be able to enjoy these songs. Fans create, they share, they bring new audiences back to the music they love.
Artists and songwriters participate in the value that their work inspires as a brand-new source of income on top of what they already earn on Spotify. For Spotify, it unlocks fan spending amongst our super users that doesn't exist today. The goal is simple: ensure that when fans create, artists benefit. This era of generation doesn't need to threaten the future of music. Because we built the system legal, trusted, and aligned, we can make sure that the value flows back to the people who created it. We're building with artists and songwriters. That means consent, credit, and compensation by design. That's how Spotify operates, and it's how we'll lead the industry into what comes next. Now, over to Joe.
There we go. All right. Thank you, Charlie. Spotify has helped shape culture for years. It's probably not lost on anyone here that the most streamed artist in the world for the last six years sings in Spanish. We're proud to have played a role in helping Bad Bunny's music reach every corner of the world. Here's what most people don't realize. That doesn't happen automatically in some algorithm. We have people who live and breathe music culture all around the world. They know what's bubbling up before it hits a chart. When they spot something real, we can help it travel in ways no other platform can. That combination is hard to copy. Most platforms operate globally from a distance. We don't.
We have the local instinct and the global reach, and they're working together in ways that no one else can replicate. In a world that is increasingly reliant on AI, the human expertise behind it becomes more scarce and more valuable. As Gustav said earlier, AI is only as good as the information it has access to, and we have two decades of it, two decades of editorial taste and cultural instincts from the world's best music editors. You combine that with a deep understanding of how hundreds of millions of people actually listen, and you have a foundation no one else can touch.
I mean, anyone can tell you that Bad Bunny has over 100 billion streams, but it takes taste and real human judgment to spot an underground breakout in Puerto Rico or an emerging scene in Copenhagen before anybody's noticed. That's a proprietary foundation that took us 20 years to build. You can't shortcut your way to it. We're investing in bringing that same expertise to the surface so fans can experience the human side of Spotify. That instinct is what's bringing our playlist editors in front of the camera. Next month, New Music Friday, one of the most recognized brands in music discovery, becomes a video series. Every week, our editors will bring their taste and authenticity directly to fans, breaking down why the world is talking about the songs that just dropped.
That's what makes us different from anything an algorithm can produce. It's not just our editors. Artists are embracing video on Spotify, too. Here's something that might surprise you. Spotify is now one of the largest music video services in the world. Official music videos, live performances, covers, all of it lives on Spotify. It's seamless. You can flip from a track to a video with a single tap on your phone, tablet, TV, desktop. Today, more than two-thirds of all premium subscribers have watched music videos on Spotify. 65% say it makes their premium subscription more valuable. For our most engaged listeners, watching a video leads to an 85% increase in streams of that song in the following month. You might wonder, how much more value can we add to a premium subscription after 20 years?
The answer is a lot. As Nat mentioned, SongDNA and About the Song have already transformed how fans connect with the music they love. You add video in the mix, and we've added more to the core music experience in the last 12 months than in many of the previous years combined. We're moving faster than ever. That video experience spans a full spectrum, from official music videos and live performances to original programming like Countdown To, Spotify Live Room, and Billions Club Live. Let's talk about Billions Club Live for a second. Billions Club Live celebrates artists whose streams have crossed 1 billion. For each performance, we bring the artist's top fans into the room with hundreds of millions of fans on our platform and the data to know who listens the most.
It's something only we can do. Each show is captured on film and released globally. You've got The Weeknd in Santa Monica, Miley Cyrus in Paris, Ed Sheeran in Dublin, Bad Bunny in Tokyo, and our most recent, Olivia Rodrigo in Barcelona. Just wait until you see the final edit of that show. Let's just say our sound engineers earned their keep. In every clip I saw, you could barely hear Olivia's voice. The crowd was singing every word back at the top of their lungs. Now imagine if we could ensure that fans like those kids in Barcelona, the ones who truly deserve to be in that room, actually get there, not just for a handful of superstars, for artists and tours at scale. Rene is going to tell you more about how we're making that happen.
Thanks, Joe. We have all experienced it. Your favorite artist announces a tour. You are desperate for tickets. You clear your calendar. You set an alarm for the moment you can buy. The clock strikes. You are in, you are ready to go, somehow you're already behind thousands of people, likely also including bots and scalpers planning to buy tickets that they will personally never use. You refresh, you wait, you finally get through, the tickets you want are already gone. They are there, they are already being resold at three times the price by a scalper. It is one of the most frustrating experiences in music today. It's frustrating for fans. It's frustrating for artists, too, who look out at a crowd and wonder, "Are the fans who built my career actually here?"
The tragic irony here is that the most dedicated fans are too often the ones that don't make it into the room. That is deeply unfair. It has been unfair for far too long. Today, that changes. Spotify is proud to introduce Reserved. For the first time, an artist's most dedicated fans on Spotify Premium will have two tickets held just for them before they go on sale to the general public. No racing bots, no chasing around online for pre-sale codes, just two tickets held for you. Why? Because you've earned them. Here is how Reserved works. We don't just look at streams. We look at everything. Streams, yes, but also saves, how deep you go into an artist catalog, how you engage with an artist across the entire Spotify experience.
The full picture of what it means to be a true fan. If you're selected, Spotify holds two tickets for you. Actually puts them aside. You'll then have a dedicated window of time to buy before the general public rush. One tap and we guide you directly to our ticketing partner to complete your purchase. Only Spotify can pull this off. Here's why. First, verified humans. Our premium subscribers are real paying fans. Not bots, not brokers. Second, we know who the diehards are. Nearly two decades of listening data across hundreds of millions of users means that we can identify true fandom at a depth that no one else can come close to. Third, we have the scale to make this really meaningful for artists.
Spotify has millions of verified super fans ready, more than enough to fill every reserve seat on any given tour many times over. We've already proven it. We've been quietly building towards this moment for years. Having worked with more than 40 ticketing partners and driving more than EUR 1.5 billion in ticket sales to date. Reserved is the evolution of that work. This summer, Reserved arrives with Live Nation as our launch partner in a multi-year agreement. Spotify is the exclusive audio streaming service offering this type of Reserved access to Live Nation tickets. We will help fans get access to many of the most anticipated tours in the U.S., with more markets coming fast. Reserved is one of the most substantial improvements to Spotify Premium since the founding of the company. This is just the beginning.
We see clear opportunities over time to expand the experience in ways that creates even more value for fans, for artists, and for Spotify alike. Every streaming service has the same music, but Reserved is something that only Spotify can offer, and that changes what it means to be a subscriber. Up next, Roman and Maya dive into podcasts.
Thank you, Rene. Four years ago at Investor Day, there were some significant questions around the business model behind podcasts. Our podcast business was a highly negative business and a drag on overall company margin. Today, podcasts are on their second year of profitability, and growth is accelerating. From here, we see a path to 40% margins in the long term. Strengthening economics goes hand in hand with user value. In fact, podcast engagement has doubled since we last met. In our mature podcast markets, over 40% of Premium users listen and watch every month. This is the advantage of moving first. We made a bet on podcasts and the amazing creators behind them, and we did it years before most saw the opportunity. Today, Spotify is the R&D engine of podcasts.
We made the investments, we built the tools, and the outcome is clear. We've contributed more than EUR 10 billion to the podcast industry in the last five years. The reason we can drive that much value back to the industry is simple. We are the only platform that operates at three layers. Number one, we are a leading consumer platform for podcasts. By bringing podcasts into the core Spotify experience, we drive meaningful incremental value. We deepen habits, and that directly drives retention. Premium users who stream audio podcasts in addition to music are spending three more days on Spotify each month. Users who stream video podcasts spend one more day on top of that. Number two, as a premium publisher, we're scaling our advertising business by leveraging deep engagement.
Sponsorships, our fastest-growing format, are up over 100% year-over-year. We're unlocking a huge commercial opportunity with the ability to insert those sponsorships dynamically. Our new creator sponsorships product allows shows to monetize up to 50% more inventory by providing better tools to schedule, replace, and analyze their supply. Ultimately, this drives higher revenue while reducing the overall ad load, creating a better listening experience for users, and helping each advertising partner stand out more clearly. Number 3, we're building the most effective tools for creators. The Spotify Audience Network and the Spotify Partner Program create a unified foundation for monetization across platforms. Our Distribution API enables more than 1.5 million shows to bring video to Spotify while keeping their existing hosting providers.
Later this summer, we're taking the next step by introducing memberships, a set of tools that allow eligible creators to offer subscriptions directly to their most dedicated fans on Spotify. Alex and Gustav have talked about add-ons. This will bring the model to podcasts. Passionate fans unlock deeper experiences, and fandom becomes recurring revenue. What makes our approach different is that we're not sitting between creators and their audience. Creators own the relationship. They have direct access to their subscribers with the ability to import and export across platforms. For those who choose to manage subscriptions elsewhere, Spotify Open Access remains our open offering, allowing creators to distribute gated content on Spotify while using any of our partners. Everything we do is powered by world-class talent.
"The Ringer" continues to define the category with authority and taste. Bill Simmons remains one of the most iconic voices in sports and culture, whether it be his shows or sold-out live events. Amy Poehler's "Good Hang" won the first-ever Golden Globe for Best Podcast. In fact, for those of you in the room, you can check it out right there. The Joe Rogan Experience" tops Spotify Wrapped year after year. Wherever a creator builds, we give them the tools to grow. If creators win, we win. Next up, Maya will show you how the product drives the next phase.
Thank you, Roman. For us, the future of podcasts on Spotify is all about making them easier to find, easier to use, and more valuable to the user and creator. Let me show you how we're building that. Let's start with video. As you heard earlier, more than 500 million users have streamed a video podcast on Spotify, up nearly 50% year-over-year. This growth reflects how video expands the format. It gives creators a richer way to connect with their audiences and gives users a more immersive way to experience the shows they already love. Across both video and audio, we've been focused on making the podcast experience on Spotify more intuitive and more interactive. Transcripts make the catalog searchable. Automatic chapters take you straight to the moments that matter.
Now you can also ask questions about anything you hear and get answers in real time without leaving the experience. It's an amazing way to engage with podcasts, letting you go further on any topic the moment something sparks your curiosity. Imagine you're listening to your favorite podcast about music history. They mention something that catches your attention, but then they just move on. Now, you can just ask, "Wait, what music video did Spike Jonze direct?" Because Spotify understands what you're listening to, it doesn't just answer the question. It picks up the exact moment in the episode and gives you the answer in context. Then it goes a step further, linking you to other songs with videos he's directed so you can keep exploring without breaking your flow. That leads to the next piece, discovery.
My favorite example is Prompted Playlist, which we recently expanded to podcasts. You write a simple prompt like, "Find me interviews with authors whose books recently got made into movies," and you get a completely personalized list of episodes. What we're seeing is that more than 50% of listeners who use Prompted Playlist discover a new show. That's another signal of how powerful personalization can be. The next step is building a podcast experience made just for you, one that hasn't existed before and couldn't exist at scale until now. We're already seeing the demand for this. Some users have started creating custom audio using their own agents, and when we introduced the ability to save those episodes to Spotify, the response exceeded our expectations. Today, this is still a very manual and technical process.
We're making it easy for users to create personal podcasts right inside Spotify. Just like with Prompted Playlist, all you'll need to do is write a prompt, and we'll generate short, personalized, private audio based on your input. It draws on world knowledge, your Spotify Taste Profile, and any additional context you import to create something that's relevant, dynamic, and shaped just for you. Here's my daily briefing as an example. I live in the Hudson Valley. I'm interested in what's going on in my area and the top tech headlines. This morning, it opened with a quick update on the new AI release that everyone's talking about. It flagged an artist I've been listening to that's playing right near me this weekend.
It even pulled in a podcast episode from a show I hadn't discovered yet but was exactly in my wheelhouse. Another way I love to use this feature is to prepare me for the workday ahead. The other day, I had a meeting with a few of our podcast partners to get their feedback on a new feature. I uploaded a doc with details about the update and some notes from the last time we met, and Spotify made me a personal podcast summarizing their feedback from last time, predicting some questions they'd have this time, and it even gave me some headlines about their recent episodes, so I'd have some talking points.
This saved me a ton of time and let me head into the meeting with everything I needed with a really short, convenient summary to listen to on my commute. Just like the monthly allowance of hours we provide for audiobooks listening, all Premium users will have a monthly allocation of credits to generate podcasts included with their subscription. Power users who want to be able to do more with personal podcasts will be able to purchase additional credits for more inference. Speaking of power users, let's talk about fitness. This is a behavior that already lives across music and podcasts and video. Spotify has the audience, the personalization engine, and the daily contextual relationship to make that behavior more valuable over time.
We've been the soundtrack to people's workouts for almost 20 years. Now, through our Peloton partnership, Premium users have access to more than 1,400 workout videos. In just the first four weeks, we're already seeing strong engagement. It turns out people don't just want to follow workouts, they want to shape them. We're making that possible very soon with guided adaptive running sessions. You can prompt something like, "Give me a playlist for a 30-minute run at an eight-minute pace," and Spotify will select tracks in your taste, stretch them to the exact right tempo, seamlessly mix transitions between the tracks, and layer in coaching cues to guide you through your run. We see significant potential to build more experiences tailored to fitness enthusiasts, and that's how we think about the all-day user.
Someone who turns to Spotify across more moments, more use cases, and more parts of their daily life. Let's turn the page to the one audience at Spotify that's more engaged than podcast listeners. Here's Owen. Thank you.
Thank you, Maya. As Maya mentioned, when someone adds a book habit to their existing podcast and music listening, they become part of the most engaged audience on the whole of Spotify. This additional engagement, it is important because books are some of the most meaningful time well spent. Books, they inform us, they educate us, and of course, they entertain us as well. This makes them fit naturally on Spotify. Just two years ago, we entered a EUR 10 billion global audiobook market that was underserved by the existing players. Since then, we have expanded to 22 markets, bringing audiobooks to tens of millions of new listeners. What matters most is how we have approached this category.
A credit-based model does not allow every book to find an audience, so we introduced an hourly consumption model that unlocks discovery across a much broader range of titles, especially shorter formats such as poetry or children's books. As a result, our already sizable audience of younger readers and families continues to grow as more people explore the kids and family catalog that we keep investing in. As a new father, this is really top of mind for me. At the same time, we've scaled the catalog from 150,000 titles at launch to more than 700,000 titles today. It's a huge catalog to explore. We've opened up the ecosystem with many English language publishers moving away from exclusive deals to reach broader audiences through Spotify for the very first time.
We've introduced industry-first experiences like Page Match and audiobook recaps, redefining how people discover, navigate, and return to books. Just like in music, when we lead, the rest of the industry follows. In a short period of time, we've established ourselves as a major force, driving where this category is headed. What's become increasingly clear is that the industry has been waiting for a platform like Spotify, a platform that can connect formats, a platform that reduces friction, and that can bring entirely new audiences into books. We're just getting started. We see significant opportunity ahead, and we're going to keep growing the books experience for listeners, for authors, and for publishers all around the world. All of this is positioning Spotify to become the home for all books.
In April, we launched print book purchases through our bookshop.org partnership because we know that many of our users, they want to go back and forth between a physical book and an audiobook. Every sale directly supports independent bookstores. We've also invested in reimagining the reading experience itself. Our feature Page Match is now available in over 30 languages, and it's been a breakout hit since launching in February. As the first experience of its kind, it lets users move seamlessly between a print or e-book and the audiobook on Spotify, and it's something I absolutely love. I'll often rea d the book in the evening, and then I go for a run in the morning. I just switch to the point in the audiobook, and I can take the book with me. Users are telling us the same thing.
We're seeing strong repeat engagement with users returning week after week, even as the feature scales. It's quickly become one of our most consistently used and highest-performing product launches. We've seen an uplift in listening of up to 55% over a month when readers use Page Match, and it's particularly popular with younger listeners, with the average age of a Page Match user being six years younger than the average audiobook listener. What's even better, users are finishing books twice as fast. That means users are reading more books, generating more revenue for authors, and purchasing more listening time. This is exactly the kind of multi-format behavior that we believe will define the future of books on Spotify. Even best-selling authors agree with us, so thanks Harlan for helping me make that point.
We're also bringing entirely new audiences into the category. Almost half of Spotify audiobook listeners are under 35, which is meaningfully younger than the broader market. The audience is roughly 50/50 male/female because we're helping publishing unlock one of its hardest-to-reach young audiences, which is men. The great news is the publishing industry is growing, and audiobooks remain the fastest-growing segment, with roughly 20%-30% annual growth in the U.S. Spotify, well, we're growing even faster than that. Listening hours grew 60% from 2024 to 2025. Almost half of audiobook consumers started listening within the last 12 months. Our indie title catalog grew 50% year-over-year. As you heard earlier, the economics are working, too. I'm happy to announce that we are on track to reach EUR 100 million in annualized recurring revenue from Audiobooks+ only this July.
We're really excited about this product because Audiobooks+ users deliver multiples higher LTV than standard subscribers who listen to audiobooks. What makes this especially compelling is the user behavior underneath it. In the U.S., consumption amongst Audiobooks+ users increased 18% in the first 30 days of purchasing the additional hours. That tells us something important. This is not just another content vertical. We're creating healthy reading habits and high-value behavior that becomes even more powerful when it sits inside the broader Spotify ecosystem. The takeaway is not that th is category is mature. It's that it's been underserved, and we're showing the industry what serving it well looks like. That brings me to where we go next. The model, it's straightforward.
Whether you're a casual listener or whether you're devouring a book every single week, we want to offer a plan that fits your life. Audiobooks and Spotify Premium builds a healthy habit of listening to books by offering great value to our subscribers and meaningful revenue for authors and publishers. Audiobooks+ drives additional revenue by monetizing heavier users. People don't stop there. Some of these users go on to purchase even more top-up hours. That's why this summer, we're expanding Audiobooks+ with new add-on tiers designed to serve every single type of reader. Like Alex said earlier, you could think of it as Audiobooks++ or Audiobooks+++, I promise you it's not going to be called that, we're going to have a much better name.
We'll introduce a number of higher-hour plans, giving our most engaged listeners the flexibility to go deeper and to enjoy even more books. Then launching later this year, family and student plans for Audiobooks+ will bring the experience to entire households and to the next generation of readers. We're of course, building more great tools for those who fuel the books ecosystem. That is, of course, authors and publishers. Spotify for Authors provides direct access to audience data, to growth tools, and distribution. Today, we are expanding Spotify for Authors into 10 new languages, giving authors and publishers the insights and publishing tools to grow their readership now in the language that they work in.
For many self-published authors, converting their book to audio in the first place can be a challenge, whether that's a financial challenge or a technical one. That's why we've already invested in publishing programs like Spotify Selects, which provides financial support to make audiobooks for the very first time. We're also rolling out publishing funds in markets like France to ensure that all stories can be heard. Now, to help bring more books to market in audio for this vital group of authors, we're introducing new audiobook creation tools. Starting in June, we are opening an invite-only beta program providing access to best-in-class digital voice technology powered by ElevenLabs. We're building this directly into Spotify for Authors, enabling seamless generation and frictionless publishing. Authors are not locked into exclusive deals.
That means they get the widest possible audience for their books, which is incredibly important to them. I'm also excited to share that we're introducing two new features to assist with audiobook discovery. First up, if you're looking for new recommendations, you can ask a question to Spotify. You might ask, "Give me some Scandinavian noir for my upcoming work trip to Sweden." Maybe you're already deep into enjoying an audiobook, you'll then be able to ask, "What else has this author written?" "What should I listen to next?" This summer, we'll bring Prompted Playlists to audiobooks, too, giving users more control while expanding our natural language discovery experience to books for the very first time. If any of you in the room or online run a book club, this could be great for you.
I'll give you a flavor of how it might work. You could say, "My book club loves murder mysteries by Agatha Christie or Arthur Conan Doyle," both Brits, by the way, "but we'd like a fresher take on the genre. Give me a list of modern murder mysteries similar to those classics that my book club would enjoy." We'll then create a playlist of audiobooks to fit the prompt. These features and our new plans, they all serve the same goal, which is to help people engage with books more by making the experience feel more natural, more connected, and more integrated into everyday life, all in the Spotify app that hundreds of millions of people already use every single day. Now I'll pass the mic to Katie to walk through how we've rebuilt the Spotify ads machine.
Thanks, Owen. Our ads business is built on a simple belief: advertising should respect the user experience. It should feel relevant to the moment, and at its best, become a part of culture rather than an interruption from it. Last year at Advance, we previewed our new tools and tech, which make it simpler to buy, create, and measure. Now, the theme was automation, but the bigger idea was that our ads platform should be built for Spotify, not bolted on Spotify. Today, for the first time, it is. Our tech is now built for Spotify and the 483 million people who use our free tier. It powers advertising across music, podcasts, and video, bringing targeting, measurement, and ad delivery together in one system.
Because it was custom-built for the Spotify experience, it's designed to drive outcomes for advertisers while respecting the user. Every day, we use advertising to support the free experience, introducing users to new products and features, and helping free users understand the value of Premium. We've built one of the world's largest subscription businesses by converting the users who start on our fr ee tier. In a way, Spotify is the ultimate proof point for what our ads business can do, and we know we can drive those same results for advertisers. A few years ago, our advertising business was not where we wanted it to be. Simply put, Spotify was not set up for where the market was going. We were too dependent on direct buying, too concentrated in the U.S., and too centered on audio-only budgets.
At the same time, marketers were moving toward automated performance-based buying. Advertisers wanted measurable proof that their investments were paying off. We're the first to admit it took us too long to make the shift. The good news is we understood the assignment. We accelerated the pivot and stopped treating culture and performance as a trade-off. Instead, we doubled down on building for both. Now, we have high-impact sponsorships where the world's biggest brands tap into Spotify's role at the heart of culture and fandom. We also have scaled biddable, where advertisers can access Spotify through self-serve and programmatic channels with the automation, measurement, and performance tools they expect. Together, these engines unlock TAM by attracting ad partners who want custom partnerships, outcome-focused performance, or ideally, both. We've dramatically expanded our potential upside.
Let me explain why we're so confident in this plan. We're already seeing the early results of our investments. Our biddable channels have grown to more than a third of our ads business. We now operate one of the world's largest global audio ad exchanges. Spotify Ads Manager gives advertisers a first-party path into Spotify, while the Ad Exchange gives programmatic buyers access to our audience through the platforms they already use. By the numbers, active advertisers grew 68% year-over-year in Q1, including through the holidays and into the spring, a strong trend that we expect to continue. The growth is also becoming more global. In Q1, EMEA grew nearly 10% year-over-year. LATAM grew 25% year-over-year. That matters because Spotify is not just a global audience.
We are becoming a global ads business. That's crucial for two key reasons. First, more advertisers give us the opportunity to drive price improvements over time. Second, it also means that we are able to more effectively deliver personalized ads to our users. Now, our sponsorship business is also growing. As Roman shared earlier, podcast sponsorship revenue more than doubled year-over-year, showing the power of connecting top creators, deeply engaged audiences, and brands in the moments that matter most. Our direct business is becoming higher impact while our biddable business is scaling demand. The heavy platform rebuild is now behind us, and our focus forward is on accelerating revenue growth and continuing to improve innovation. The first priority is further growing the market for Spotify.
We're a category leader in audio for advertisers, and we know that ears can be just as powerful as eyes when it comes to driving successful results through advertising. Our new measurement capabilities allow us to clearly demonstrate that. As we further diversify our video offerings, we will capitalize on growing those opportunities. The new engine of sponsorships and biddable is now running. The second priority is leveraging AI to make our tools even more powerful, taking Spotify's Ads offering to the next level. AI reduces one of the biggest barriers to audio advertising, which is the audio, the creative itself. Roughly 70% of the ads we create for our partners in Spotify Ads Manager are built using AI tools, and we're already seeing the economic benefit of reducing the costs of creative production.
To date, we've generated more than 20,000 ads for over 7,000 advertisers globally, and that number is growing. Advertisers can now create scripts, voiceovers, and fully mix with background music without giving up any creative control. AI is also helping advertisers move more quickly from objective to campaign by improving audience discovery, optimizing delivery in real time, and helping us better personalize ad load for each individual user. In practice, that means customizing the volume of ads that each user receives based on their unique behavior on Spotify. The third priority is building new ad experiences that just weren't possible before. Brands can now sponsor playlists, be in the conversation with creators, and as Spotify introduces more AI-powered premium experiences, brands can also sponsor access to those moments. These are not your standard ad impressions.
They are deeper, more organic brand integrations that connect advertisers to culture and turn love for Spotify's brand into advertiser value. That's exactly why we expect higher growth in the second half of 2026. Looking ahead, double-digit growth beyond that, and a much larger advertising business over time. When you step back, advertisers have always loved Spotify for our passionate audience, our deep engagement, and the role we play in culture. Now, we have the ads platform to create advertising experiences which fully harness that power. This strategy will drive better outcomes for advertisers, better experiences for users, and better monetization for Spotify. With that, I'll hand it to Gustav to take you under the hood of the intelligence system that powers what we're building next.
Thanks, Katie. 3.4 trillion data points from our hundreds of millions of engaged users every day tell us where we should go next. That's how Spotify has evolved into a product that others try to replicate but often struggle to match. Let me take you under the hood to see what drives that advantage, our intelligence. Just a few reminders. As I said in my earlier remarks, we know that we need general LLM reasoning across facts and domains like math, coding, et cetera. Today, we're buying that and benefiting from that price performance cost curve. We are instead spending our resources on building and training a Large Taste Model, an LTM, using our proprietary data. Let me explain a bit further for you.
A traditional language model learns patterns in text by predicting the next word at massive scale. Our LTM actually goes a bit further. It understands not just text, but also every historical user interaction and every piece of content on Spotify. Our advantage isn't just in architecture or knowledge and keeping secrets. It's in the data, both historical but also ongoing. Training these models requires billions of real user behavior sequences over long periods of time, literally years. Unlike the static factual knowledge in an LLM, like the capital of Texas, taste is constantly evolving, so the model also needs to continuously learn from processing trillions of new signals. You simply need an insanely active user base every single day to stay relevant in taste.
There isn't just one data set, actually. There are several that we need to combine into the LTM. First is that behavior layer that I just mentioned, the billions of historical user sequences and trillions of daily data points. Second is the metadata layer. That's all the licensed and acquired foundation that maps hundreds of millions of pieces of content across our three verticals, along with all the rich metadata and engagement signals that power the Spotify experience. Third is the creator layer. Less visible maybe to you, but equally important. In fact, you may not know this, but we have proprietary tools developed by us, used by millions of artists, authors, and podcasters along with their publishers, labels, and distributors.
These applications generate unique data that powers capabilities like SongDNA, where they correct their own catalog, et cetera, and further strengthens our recommendations and discovery. Fourth is the cultural layer. We're using systems that understand what is happening in the wider world, going outside of Spotify every day to inform features like About the Song or Prompted Playlist about what happens in the world in real time. It doesn't actually stop there. Combined with modern reasoning capabilities, this understanding is actually interactive, not just static. Users can ask for example, as Maya said, a workout playlist tailored to their specific habits or discover books, as Owen said, that fit their exact taste. This result is not just prediction, but the ability to actually shape and generate experiences in real time.
That's how we think about the LTM. Not a standalone system, but a fusion of proprietary taste understanding with advanced reasoning. This is where the economics come into focus. AI isn't just a cost layer for Spotify. It's a product and monetization layer. Our early model deployments are already improving the quality and effectiveness of our personalization, driving meaningful gains across the platform. This includes an almost 10% growth in auto-play song saves, almost 10% improvement in podcast discovery from home, and close to 20% increase in users interacting with DJ suggestions. This is a really big deal that has real implications for retention and lifetime value across our entire user base at scale.
Beyond improving the core experience, our Large Taste Model is also enabling entirely new kinds of products, including Prompted Playlist, the interactive DJ, Taste Profile, and many of the generative experiences that you've seen today. Already, these newer AI-powered experiences are being used monthly by nearly a quarter of our U.S. Premium users. Large adoption. When cost per interaction continue to decline and usage expand, users get more value from Spotify. We gain more opportunities to capture that value over time. As Maya and Charlie showed, with personal podcasts and new music fan tools, we believe experiences like these will naturally expand users' willingness to pay. This allows us to scale a tiered model and turn intelligence from a cost driver to a revenue driver.
There is intelligence in free, there's more of it in premium, and higher intensity experiences available as add-ons. As costs come down, we can expand these offerings while improving the value. As we said earlier, many times by now, willingness to pay is not an average. It only looks that way when you happen to have a single price point. In reality, it follows a power law with usage intensity and willingness to pay scaling together. I'm not just making that up. AI has already proven this. If you look at the leading LLM products, they are aggressively priced as a power law across usage tiers because their heaviest users get meaningfully more value and are willing to pay much more.
That is why we see AI as an opportunity, not a drag. Better product, more value captured, strong margins. That's the model. That's what Niklas will show next, how we're putting this into practice to deliver more value faster and more efficiently. Niklas.
Thanks, Gustav. Now that Gustav has walked you through the Large Taste Model, I'll show you how we leverage AI to build and ship, and why it's so difficult to replicate. Let us start with how we build. As Nat and Nicole shared, this is not just about productivity. It's about what becomes possible when capacity is no longer a constraint for us. Our engineers are spending more time on high impact problems. AI is embedded across the entire workflow, and we've extended its utility with our own internal systems, making it secure and scalable. Honk, our background coding agent, is a strong example. It builds on leading coding agents, but it's tailored for Spotify, running persistent cloud sessions so engineers can seemingly switch devices, collaborate in real time, and automate routine work.
We're seeing up to 90% reduction in the time spent on common maintenance tasks, freeing developers to focus on user-facing innovation. Tasks like dependency upgrades and migrations now run fully autonomously, allowing teams to move even faster with less friction. I've been building software for over 30 years, The AI transition that we're going through now is like nothing I've ever seen before. Today, more than 99% of our developers use AI weekly. Coding productivity is up 76% since last year, from what was already a very high level. More than 73% of our code contributions are AI assisted across more than 4,500 production changes that we ship every day. At the same time, quality remains stable.
In fact, as I've been preparing this presentation over the last few weeks, we've had to change these numbers multiple times because they keep going up. We're no longer primarily scaling through headcount. We're scaling by increasing the impact of the people that we already have. You can also see this in our economics. Revenue per engineer is rising. Each engineer's work reaches more users and supports more use cases. Our ability to scale our R&D has changed with AI, and once these systems are built, the cost to serve each additional user will further improve. In practical terms, one engineer now supports more output, more features, more services, and more user value. The same team can power a broader and more complex set of experiences. Next, let's take a look at how we ship.
As Nat mentioned, we moved beyond traditional prototyping. Today, anyone at Spotify, not just our engineers, but also our product managers, our designers, and business leaders, has in fact all the way up to our executives like Gustav, can and do build working prototypes right on top of our production code base. Our real data, our real design system, all running inside our actual app. We have an internal app store where these prototypes can be discovered, installed, and tested, all kept in sync with our live code base. Ideas are now tested in real conditions and with far fewer resources than before. What used to take weeks can now be validated in minutes. For most of the streaming era, platforms had a wide downlink and a very narrow uplink.
We could deliver millions of pieces of content, user could only respond back to us with a skip or save or like. We then had to infer their intent. Gen AI changes that dynamic, making the system interactive. We move from guessing to understanding and from one-way consumption to real-time collaboration, with users telling us what they want in natural language. Today, you've heard about DJ, Prompted Playlist, and Taste Profile. Together, they represent a shift towards a more interactive and user-controlled Spotify. Today, I'm very excited to tell you about how this shift is coming to life in Studio by Spotify Labs, a standalone desktop app that takes personal podcasts further. It will be available soon as a research preview for premium users in more than 20 markets.
Studio understands your Spotify taste across music, podcasts, and audiobooks. It can draw on world knowledge, and with your permission, connect to services like your calendar, inbox, notes, and feeds. You simply ask for what you need, like a briefing for your day or an overview of your weekend plans, and Studio creates it for you. Because it's built for Spotify, the output can be saved directly to your library, where your personal private content lives alongside the music, audiobooks, and podcasts you already love. By understanding, building, and shipping faster than ever, Spotify is becoming the home for personal media. Not just what already exists, but what users create and carry throughout their day. Finally, let's dig into what sits underneath all of this.
Gustav described our Large Taste Model and the data advantage behind it. Let me show you what's actually inside. The model is trained through continued pre-training on our proprietary data set. Every track, artist, and podcast in our catalog is mapped to a learnable token that captures both what the content is and how users engage with it. Like traditional recommendation systems that predict from historical patterns, the model generates candidates directly, much like a language model generates text. This is the architectural shift that makes the generation era Gustav described technically possible. That is also what makes our agent fundamentally different from a generic AI assistant. Most start from zero every session. Spotify reasons natively over taste so that conversation reflects years of real listening behavior, not just a prompt and a search query.
What we've built is not just a model. It's a system where data, models, agents, and our engineering platform work together to deliver personalized experiences at scale. Because it runs on what we believe is the world's largest licensed audio and taste data set, it compounds. Every interaction improves the model, and every improvement can reach hundreds of millions of users through a single platform. Because this system is built on licensed data, trust is foundational. Spotify's role is at the center of the ecosystem, strengthening the connection between creators and users, rather than abstracting away from it. This is what we mean when we say that Spotify does not rent the future. We build it. Now over to Christian.
Hello, everyone. Nice to see you all here today. Before I start, let me tell you that all growth rates that I will be referencing are on a constant currency basis, unless otherwise said or noted. I wish I had cool guys telling that instead for me, but next time maybe. Differentiated experiences creates differentiated economics. Here's what ours looks like. Four years ago, we made a commitment: turn Spotify into great business by growing music, scaling our podcasting into profitability, and launching accretive new verticals like audiobooks. We delivered, and today we are stronger than ever, proving that we can scale efficiently and we can monetize effectively. As you heard, monetization starts on the top of the funnel, where we delivered a healthy 17% CAGR in users over the past four years.
Growth is coming from every region, taking us to more than 760 million users, MAU, globally as we continue progressing towards 1 billion users and beyond. That scale has been a key driver of our healthy subscription business. It's grown at 13% CAGR to nearly 300 million subscribers. Engagement remains strong and continues to deepen. The number of subscribers using Spotify more than 28 days per month across multiple devices and verticals is growing. That breadth and frequency of use is a powerful drive of retention. Those fundamentals, scale, engagement, and retention, that drives our financial model. Here's what that means for our performance. Alex talked you through targets we set forth in 2022. The results speak for themselves.
18% CAGR, more than 7 percentage points of gross margin expansion, and nearly EUR 3 billion in free cash flow generated in 2025 alone. It was the result of a disciplined financial model. Targeted investments into music business, audiobooks, video podcasting, and the beginnings of our AI capabilities that we heard about today. As the impact built over time, we stayed patient. Today, Spotify generates meaningful cash flow, and we are entering this next phase from a position of real strength. Our advertising business grew at a 10% CAGR, even as we were re-platforming the ad stack and absorbed the short-term pressure that came with it. That work is important because it positioned the business for a larger opportunity.
Gross profit grew faster than revenue at 23% CAGR, with gross margin reaching 32% in 2025. About one-third of that came from the music business, and with the rest driven by podcasting and audiobooks. Today, both music and non-music verticals together operate above 30% margin. In 2025, the gross profit we generated from Marketplace was four times higher than 2021. Our offering of artist tools and licensing initiatives continues to create value for both artists and for Spotify. At the same time, we maintained a tight control of cost. Operating expense growth has remained essentially flat since 2022, as the inflationary increase in R&D and sales and marketing were more than offset by the 23% reduction in full-time headcount. That discipline delivered over 18 percentage points of operating margin expansion since 2022.
Combined with our revenue growth and gross margin expansion, this moved us from a loss-making position to generating EUR 2.2 billion in operating income 2025 at 12.8% operating margin. All of this fundamentally has reshaped our free cash flow profile. In 2025, free cash flow margin reached approximately 17%. That strengthened our balance sheet, and it gives us the flexibility to invest in what comes next. Before we look ahead, let me briefly outline how we think about return on investment. It shapes our financial strategy. At Spotify, we invest where we can generate returns above our cost of capital, and we stay disciplined everywhere else. What makes our business unique is that the most reliable leading indicator of a return above WACC isn't always IRR on a spreadsheet.
It's an increase in customer lifetime value. You heard the mechanics throughout today. The KPI we underwrite are centered on engagement, revenue, efficiency, and retention. Our bets from Audiobooks+ to DJ to Reserve have clearly quantified targets tied to those drivers. It's the way these bets build on each other over time that drives lasting improvements in LTV. This creates a direct chain to free cash flow growth. In the U.S. alone, our largest market, customer LTV has increased by more than 70% since 2022. We see similar trends across markets of every stage of maturity, from developed markets like Sweden to faster-growing markets like Brazil and India. AI accelerates our ability to make these bets with even greater velocity. Our bar for returns stays the same, and the direction is clear.
Unique economics is keeping improving. Let's look at what this framework will produce by 2030. First, we are targeting a mid-teens revenue CAGR. The drivers are clear. We believe we're on our way to 1 billion MAU. We have emerging and developing markets as twin engine fueling growth in subscribers. Subscriber retention is already best in class. We see opportunities to improve it even further. ARPU will become a more meaningful growth driver, supported by thoughtful price adjustments, more defined product tiers, and an expanding mix of add-ons and à la carte offerings. This applies for both Premium and add-on-supported users. Our clear success with Audiobooks+ shows what's possible when we deliver for super users. The add-on opportunity is significant across music, podcast, audiobooks, and beyond.
AI will enhance this opportunity and our already best-in-class personalization. We plan to build on this momentum over time, unlocking new, scalable ARPU streams across the platform. We also expect our advertising business to re-accelerate growth into the double-digit range. We expect that to start ramping towards these levels in the second half of 2026. More users, more advertisers, better monetization per impression. Let's talk about where we're going with profitability. We expect gross margin of 35%-40% by 2030. Within Premium, music has further room to run, supported by Marketplace and add-ons. Our higher-margin non-music offerings expand to more markets and monetization deepens. The favorable revenue mix shift will drive further improvements. In ad-supported, the story is automation at scale. We expect music margins to continue to improve, particularly in emerging markets and through Marketplace.
Podcasting has made significant progress moving from a deeply negative in 2021 to what we believe will be more than 20% in 2026. From there, we expect it to grow. Long-term, we see a path for 40% driven by both first and third-party content. At Spotify, we manage cost with rigor and with constant focus on efficiency. That discipline underpins our ability to scale financially. It gives us both speed and flexibility to invest where we see strong returns. We see strong returns potential, we will go there. In 2026, we are investing in areas where we see clear returns. As we mentioned on our last earnings call, the OPEX will step up. We have told you that for quarter 2. It will step up in quarter 2 and quarter 3 before moderating into quarter 4 and 2027.
This represents a temporary increase of roughly EUR 200 million on marketing and R&D. We will naturally calibrate these costs as we move through this short-term increase. As you have seen from us over the last four years, we always strike a balance between investments and ensuring sustainable margin expansion. This will not change going forward. On the marketing front, we are investing opportunistically to drive conversion and increase LTV featuring key product enhancement. Our focus on differentiation, conversion, and retention is why we continue to add value to Premium. This enables monetization over time. The customer acquisition cost is incurred upfront. We know that. The payoff comes from higher LTV, growing revenue at scale, in addition to improving sales and marketing leverage. Our AI R&D spend has two pieces.
The first one is somewhat front-loaded, and that is primarily the large taste model like Gustav just talked about. This will drive improved personalization and scale the new add-ons. The second one is ongoing, recurring developer usage of AI like Claude Code. This will lead to shorter build and ship cycles. Our global scale makes distribution and monetization of every new product more efficient. As these two forces compound, we will see accelerating LTV expansion, shortened payback periods, and improved operating leverage. All of this points to an operating margin of at least 20% 2030. Before we get into free cash flow, just a couple of words on our taxes. We estimate our normalized P&L tax rate to be around 22% and foresee that we will become a full taxpayer by 2027.
On average, over time, we estimate our cash tax to be around 90% of our P&L tax rate. That leads us into the free cash flow. We have a capital-light model, favorable working capital dynamics from upfront customer payments, and one of the world's largest free paid subscription business. Our operating improvements directly translates into cash flow. When gross margin expands, it flows. When LTV improves, it compounds. We expect free cash flow to show strong growth through 2030. Our free cash flow per share will be part of our scorecard going forward. Free cash flow per share is the metric that tells us without ambiguity whether we are allocating capital well, executing the model, and drive value creation. It's gone from around 0 in 2021 to approximately EUR 15 per share today.
Now let's talk about capital allocation. You all know we carry cash and short-term investment balance of EUR 8.8 billion today. With a free cash flow development, that means that this balance could grow considerably over the next few years. Simply put, capital is not a constraint. That said, we will remain disciplined in how we allocate capital guided by a clear ranking. We will keep a strong balance sheet that gives us flexibility and supports the execution of our strategy. This enables us to move on the right opportunities, even in a constrained market condition. First, we will continue to invest organically for growth that drives long-term value into R&D, content, marketing, and new verticals. Wherever we have a clear line of sight on returns above cost of capital.
The LTV framework is our real-time signal, not a projected spreadsheet. Our bias remains towards build first as that remains our core strength. Second, we will allocate capital to acquisitions when they can accelerate our strategy. The bar is consistent, returns comfortably above our cost of capital with a clear path to cash flow accretion. Third, capital to shareholders. We will of course continue to repurchase the equity compensation we grant employees annually. This protects against dilution. It is currently below 1 million shares per year. Given the stronger cash flow in the years ahead, even with M&A, we expect that we will also return excess cash to shareholders. This will be over and above the anti-dilution baseline. The thesis is straightforward.
Deploy capital where returns are compelling, stay disciplined where they aren't, and allow free cash flow to build over time. Four years ago, Spotify made a commitments. We delivered. What you heard today is how we run the same machine over the next five years, not as an aspiration, but as an extension. It's an extension of the framework we have been operating and refining for many years. Now back to Alex and Gustav.
Okay, this brings us to the close of our presentation. Hopefully, after today's update, you have a clear picture of Spotify's future. The scale that we have built unlocks vast opportunities for value creation. The world is moving in our direction towards more personalized experiences, more interactivity, and more control. These are areas that we know we can lead, because living in the future has been part of our DNA for a long time. Never getting comfortable. Yeah. Looking to see where technology, behavior, and culture are heading before it becomes obvious to everyone else. That instinct has shaped some of our biggest decisions over the years and enabled us to define and build for what comes next. What matters most in this next chapter, taste, trust, culture, has always mattered to Spotify.
It is why we exist. It is where we succeed, and it is what we will continue building for.
That's what helped us build lasting differentiation, increase our engagement, and evolve the business over time to not only meet, but actually exceed the expectations of our hundreds of millions of users. Remember, we started with access, we moved to personalization, and now we're going into generation. Each of these transitions has expanded what Spotify can become. Every time, our advantage has compounded because the generative era rewards scale, data, and deep user understanding more than any era before it. What excites us most about this journey isn't any single feature. It's the entire foundation underneath it all. We're very proud of what we built, and we're very proud of the teams that you've met today. We're even more excited about the next 20 years.
W e will continue to raise our ambitions. Thank you. Now let's take some of your questions.
All right. Thanks, Alex and Gustav. We are going to start the Q&A session. We're going to take questions from the room. We'll call on you. We'll have mics circulating. Please introduce yourself and your firm name. We ask that you please limit yourself to one question so we can get in as many as possible. Why don't we start with the first one here from Steven Cahall.
Do I need a microphone?
Microphone's coming over here.
Thank you. Steven Cahall from Wells Fargo. I know there's going to be a bunch of questions around the UMG deal that you announced and the ability to create AI music. Let me see if I can just cover all of those in one giant question.
Thank you. Let's try.
Yeah. Wondering how you're thinking about pricing for this add-on? How should we think about how it impacts your gross margins? Will this be rolled out globally? Do you see this as paving the way to sort of a fast follow with the other major labels? Maybe lastly, will we be able to share it on our TikTok feeds? Thanks.
I'll start, Steve. It's a good question. Basically, you had probably 20 questions in there packed into one. I'll start, and maybe you can jump in as well. Today, this marks a landmark deal for us with UMG. I'm very pleased with where we are. We're not going to talk about the specific details of it, I'll break it down to you in at least how we're thinking about the whole sort of ambition here going forward. The first thing that we've done is b asically we have created the first really legal alternative here to create remixes for fans and users to create remixes and covers. The second thing that we did here was actually to create a licensed and controlled medium for artists to actually participate in artificial intelligence. Now, we all love covers and remixes, that's proven.
It's largely an untapped opportunity, because there isn't really a scaled way to make money on it. What we did today was basically to unlock that opportunity. I think this is going to be additive both to artists, songwriters, the community at large, Spotify, rights holders, and so on. We're very pleased with this. You can talk a little bit about how we plan to launch the product and so on. I want to say also that this is also the first time where Spotify does not need to have everyone to launch a product like this. That said, we don't want to leave artists and people outside of this. Of course, we're welcoming everyone and we want everyone to be part of this because we believe this is one of the greatest products we've built.
It's pretty unique for us to not actually need everyone to get started. We're excited about the opportunity. A couple of more things. We don't want to go into the economics, but you should expect this to be at least moderately neutral to accretive to us. We just don't do deals that are bad for any of us, which is one of the reasons why we spend so much time on getting these deals right. A couple of things that I think are important is the incentives here. So far, AI music has mostly been about net new music, right? Meaning when you create a song that didn't exist before. Those can be new creators, those can be existing creators using those tools.
Largely, existing catalogs and creators with most existing catalogs have been left out completely. It's really only about replacement for them. We think net new music is happening. There are many companies doing this. This music gets uploaded to Spotify. That is happening. This other part was just not going to happen without us. It would not happen by itself because, as Alex said, there was no legal framework. We think the opportunity is very large. If you look at other media types, existing IP in TV and movies is usually considered the most valuable rat her than the l east valuable. As Alex said, you can see artists doing remixes and covers today manually because it is a way for them to get rediscovered and get more stream share.
Typically, one song can become maybe three remixes or four remixes and maybe even five covers. With this, one song becomes 10,000 songs, 100,000 songs, paying tribute you know, being playful with the original.
I think that's important because if you look, why would people want to do this on Spotify? If it's about getting a share of the revenue pool, you want to get the largest share of the largest revenue pool. We are by far the largest revenue pool. We think this is what we should do and what we're going to do better than anyone else. To answer some of your other questions, creation will be an add-on. You will be able to do some creation in Premium and so forth. As you know, Alex here has worked for, what is it, 17 years on finding the perfect conversion points between free to Premium to the next. We're going to find the right conversion points, but creation is a Premium add-on.
Consumption is not a Premium add-on. Anyone on Spotify, Premium and free, can consume. I think it's very important that if you take the time to create a cover or re mix that you love, you can share it with anyone. Because the free tier is free, literally anyone can consume. All you have to do is to install Spotify Free to hear it.
You think about the catalog as one that's going to not just expand but multiply. With that comes usage, with that comes monetization and further unlocks.
Right. Did you get some answers at least to the 20 questions? Good. All right. Thanks, Steve. Let's go next one right here in the front. Rich?
Since that was 17 parts, I'm going to ask two. OpenAI believes in two to three years, we're going to go to an agentic world. No one's going to use apps. Apps will fade away. I don't really understand why I would use OpenAI to listen or watch music, given everything you just showed on stage here today. They obviously have a biased view, I'm just curious, not just OpenAI, but just the LLMs and foundational models in general. What is that tension, who ow ns the customer, and how do you think about playing in that world of LLMs over time? Is that a risk or a tailwind, how do you think about it? Just two, because I've really been thinking about it this whole meeting.
The ticket thing is a really interesting take on giving fans tickets. Obviously, a small number of tickets, but if I am a free user, I upgrade to Premium, and then I find out I don't get my Taylor Swift ticket. How do you not get pissed off Premium subscribers? Thank you.
I can start with the second one first, actually, if you want to. People usually only remember the second one. The way we think about tickets and Premium is there's never going to be enough tickets for everyone in Premium because the arenas have a finite amount of space. We are talking about us holding millions of tickets. T his is not a small lottery or something. We'll make a lot of people very happy. There was never enough space. The problem was there was never enough space, and it was deeply unfair who got that space, right? It was the scalpers. It was the one with the most money. What we are doing is we're bringing fairness to this, right? We're taking the actual fans to get those tickets. That's how we're thinking about it.
It's also important to remember that you are not paying for this. You just get it in Premium, right? I think if you were paying extra for this and you haven't got it, that's very different. You are not paying extra.
You're paying for the tickets, but you're not paying for the reservation.
We think people will be very happy about this because it's something they just get, and two, we bring fairness to this system.
Frankly, this is one of the most lovely improvements to Spotify Premium since our founding. It truly is, right? We know this because we've done our research around this and tested it with some people. Because of the scale, we can get very, very precise on what people enjoy or not. Even you asking that question straight out and, you know, is a reaction to how good this is.
Obviously, it's not just the fans that get happy. The creators, they get very upset when their biggest fans cannot come to their shows. Everyone there gets happy. Now, I was not trying to avoid the other question. Let's go back to the other question which was, the LLM saying that all apps disappear and so forth. We obviously don't believe that all apps disappear. We do think that it's getting very cheap to write software, which is a tailwind for us that we're riding. The way we think about this, and we always have, is our ubiquity strategy. Spotify needs to be where users are. If you look at our history, I think one of our biggest success recipes has been that we always partnered. Interestingly, we always partnered with our biggest competitors.
An early landmark partnership was Facebook, now called Meta. We also partner with Google, who's one of our biggest competitors in YouTube. We partner with Apple. We partner with TikTok. We partner with everyone. The whole idea is to be where users are, right? Because there's a lot of discovery happening there. Now, to answer your question, there's also a lot of creation where people are saying, "Couldn't I make a playlist of this?" OpenAI knows that they're into this. Can they get a podcast from Spotify on that? We just see this as the next Google search. It was very important for us to rank highly on Google. It's very important for us to rank highly in these LLMs. For us, this is just more ubiquity.
Hopefully, after seeing what we did today, you understand that we feel quite comfortable about the unique advantages we have. Yes, we are great at building software and product, but we also have an enormous amount of personalization data, where the biggest companies in the world have tried to beat our personalization for soon 13 years at least, and have not. We also have enormous amounts of license data. That is the Spotify value, the personalization, and we're not giving that data away. Last, I would say Spotify was always in the background, right? We were always soundtracking all the other experiences. We were never reliant on actually being the foreground app. To soundtrack your OpenAI or Claude session, we think is actually fantastic for us. That's how we think about it.
All right. Thank you. We'll let this mic circulate. Let's go to the back here. Michael Morris?
Good afternoon. Thank you. I'm Michael Morris with Guggenheim. I wanted to ask you about the freemium funnel and the conversion. A couple questions there. First, you did enhance the functionality of the free tier last year. I'd love an update on how that's progressing with respect to what you wanted to see in the conversion to premium. The second, you made a comment earlier that AI is accelerating with faster localization and personalized services. My question there is, when we think about that conversion from free to premium, I think generally we've thought about the economic sensitivity, but apparently you see something in addition. I'd love to hear how the investments that you're making and that AI's helping with, and that localization can drive even more conversion if that's in fact what you're expecting.
I'll start with the first one. Then maybe you can jump in on the second. You heard us talk about growth, and many people have asked me throughout the years, "How do you grow Spotify? How come you guys have hundreds of millions of users and hundreds of millions of subscribers? How do you actually acquire users? What's the secret?" That second part of the question is a misunderstanding of how growth works, right? The way super scale growth works is that you actually start with engagement. You start with engagement in the way that we talked about engagement. We're not in the business of just maximizing number of minutes. We think about days in a month, which is really a proxy for contexts, right? Different contexts you're in.
If we can be in as many contexts as possible, being as many days as possible in a month, then that's the right engagement. Any product we build or develop starts with us thinking about how can we get you to interact with us more days in a meaningful way. The new free tier addresses exactly that. What you're seeing basically is not just, in the years prior, where we have consistently increased engagement across these axes of days and devices and verticals. What you're seeing basically is us taking a step change. That's why you've seen us in some earnings calls, we've been talking about how we've been beating the MAU growth, versus even our own expectations really. That's the result of improving the free tier.
Once we have people on the Free tier, you know exactly as I, that what we focus on is actually to build out the differentiated proposition of Premium to get them to come on board, right, and come on board the paid experience. That is also why I think you've seen some of the stuff today, like Reserves for instance, that someone here asked about and so on. Those are all investments into making Premium even more sticky and the conversion from Free happen in a much more sort of pre- scaled way.
Can I ask you on the other question you had about AI and free, can you repeat that? I'm not sure I understand exactly.
Sure. I'd like to understand the conversion of free to premium and the potential for that conversion to become stronger as a result of what you described as the faster localization and personalization by AI.
Sure. Sure. As Alex said, Alex comes from the gaming business.
That was many years ago. That was 20 years ago.
Did I?
Yes.
What you know there is you need to start with engagement, as you said. What we've said for now at least the 17 or maybe 18 years I've been here, is that the only thing that was very predictive of how much you're going to pay is how much you play on the free tier. More engagement drives more conversion, which is a shock to some people because they think you should cap the experience and push people over. That's the wrong way to think about it. You're not going to start paying for something you don't use. You need to first use it all the time, and then you need to find good pay gates. Like for example, you don't want the ads anymore, you want on-demand, you want offline and so forth.
We need to get people deeply engaged. First, you need to get them to care deeply, then you ask them to pay. The key driver to that engagement is personalization. We see it clear as day. I showed you the metrics. We increase personalization, you listen more. Your propensity to say, "This is actually worth EUR 6.99," goes up. It is a very direct relationship. The more you play, the more you pay. We still need to have good pay gates and as you know, we pay gate on functionality, not on content. You have the same content in Free and Premium, and that's so that we can get you as engaged as possible.
The localization bit is obviously a balance that we strike between a number of different things. One is language, of course, which has never been very costly, but something that's just simply is much faster and much better. Right? You no longer have broken sentences in Finnish and in Hindi and so on. That's immediately addressed and something that is hygiene. Number two is culture, of course. There we have actually great signals from our teams around the world, the editorial teams that we also talked about, how that kind of feeds back into the model and helps us personalize even better. Really the content signals and the cultural signals that are coming in from the outside feeds our Large Taste Model even further, which then gives you that upside on personalization which then gives you that additional engagement, and then gives you the conversion.
There's one more concrete example that I think is important. I just said that we need to get you engaged, and to get you engaged, we need to personalize, but then we need to understand what you're into. For a long time, we worked with onboarding flows. You would click, "I'm into these artists," and so forth, which had some efficiency, but mostly you just had to use it. You had to use it a lot before we started to understand who you were, because we only had a few skips. In the previous free tier, we didn't even have on-demand, so it was a pretty weak signal. We suggested something, and maybe you skipped. It took a long time to understand who you were.
What is interesting with the localization of AI is now we can ask you in English, just tell us, just spend five minutes in English writing who you are and what you like. Now we can deeply personalize to you very quickly. That's another clear example of the fact that we can now act in many languages means that we can personalize much, much faster.
All right. We've got time for one or two more. Why don't we go here in the middle with Jessica?
Thanks. Jessica Reif Ehrlich from BofA Securities. There were so many products introduced today, which is a bit of a surprise obviously all about driving engagement and monetization. Just wondering, you didn't say anything about pricing, if you could say anything about how you're pricing these tiers, what's your philosophy? Will you have multiple super premium tiers at different price points rolling? How do you think about it globally? On the engagement side, you talked about it's basically every day people are on. Almost every day. What can you say about the daily use engagement? Is it two hours? It just drives everything, not just staying on the platform, it just drives advertising as well.
Yeah. Okay. Let me see there. You had a couple of different questions in there. I'll bring you back to something that we've said many times before. The one thing that we tracked very closely is what we call value to price. All the way from basically the founding of Spotify until 15, 16 years later, we just focused on the value, not so much the price. Value over price just went up and up and up and up. Spotify became the greatest music product in the world as a result of that. Now, something else actually happened, and that's that we created these plans, these multi-account plans. That actually pushed price down slightly. What you had was basically a ratio that just skyrocketed. Both the numerator was going up, and the denominator was going down.
It's a good recipe for increasing the ratio of something. What then happened, if we take ourselves back three years, we started raising price. That value to price came down slightly without really a big loss of subscribers. Rather the opposite. Even us, we were surprised that the churn was so minimal. Then we went back, we invested more into value, we did another price increase, it dropped again, we increased more in value. What used to be just a straight line of value to price increasing has now become a more moderated, sort of almost like a up and down sine curve that's sort of pointing into the future. The way we think about it in the end is that the consumer or the subscriber user should always win. That is our focus.
We will monetize when we need to so that we can invest back into differentiation for the whole ecosystem, not just Spotify Premium, but also free. Now, when it comes to different add-ons, you can actually look at Spotify Premium as almost like a mosaic of different niche audiences. 300 million of them. We talked about the power law usage curve, demand curve earlier today. That has opened up our eyes tremendously as we've scaled. What used to be a niche audience that we had, which was thousands of people, we were never interested in building a product to harvest that or to convert that person. It's too small. Even the smallest niches are now in the millions for us.
For us, it makes a lot of sense to look at this demand, this specific niche demand, and then we create specific products for them to engage more and then monetize more. That really is how we think about it. This next era is going to be about the segmentation of our 300 million strong and growing subscriber base.
Instead of thinking about it as one add-on tier, like the sub-premium idea, just think of it as audiobooks as a sub-premium for super audiobooks fans. We just presented a personal podcast add-on. We also presented a creator add-on. It's just many of them. The whole point of super fan engagement is that eve ryone is not a super fan. There are going to be multiple add-on products that are priced, as you can see with Audiobooks Plus, you kind of gave away, there are going to be Audiobooks Plus, Plus, because the engagement in there differs, right? You're going to see multiple of them priced for those segments. Some more expensive because there's more willingness to pay, some slightly less.
That's right. Next.
Great. Okay. We've got time for one last question. Let's go over here to Justin. There you go. Justin.
All right. Thank you. Justin Patterson from KeyBanc. I wanted to go back to just touching on taste a little bit more. You mentioned Studio by Spotify Labs. That seems like something that really broadens your corpus of information about the user and feeds into the broader AI personalization. Just would love to hear more about how you're really positioning the app going forward to ingest more of this data and take people through what sounds like it's going to be a more complex user experience. Just making sure you're taking that good signal from the user and then surfacing the right information at the right time to keep growing more of those moments each day. Thank you.
Yeah. Yes, we are very excited about the fact that it's so easy and cheap for us now to learn much more about the user. This is why I said one way to think about generative era is that computers understand English. I've been doing deep user research interviews with 10 users for the last 17 years trying to understand them deeply and then saying, based on these 10 users, let's guess that the other 599 million are similar, which is not true. Now we can do a deep user research interview with every user, every day, because they're talking to an LLM. That makes me very excited. This is truly deep personalization. What is interesting, though, that we see with the DJ and from the playlist, people are telling us enormous amounts of things about their lives, right?
When they're asking for podcasts, when they're going for a run, when they're asking for books, we learn so much more about them. All of a sudden, this data that maybe only Google had, we're getting an enormous amount of this. We're starting to learn a lot more about our users. Now, as you said, with Studio by Spotify and also what Maya showed, personal podcast, people are now starting to upload PDFs and personal content. We only had public content before. We are very quickly learning much, much, much more about our user base. That makes us very excited. We think we have a very big potential there. I want to touch on the other question you have of complexity.
This was a contrarian bet that we made some years ago and went into podcasts where everyone said, "Clearly you're going to build a separate podcast app," and then sort of clearly you should have had a separate audiobooks app maybe. By then people had understood the strategy. Our intuition then, which I feel has been very vindicated now with the question you had, Rich, about software doing more, was that it's just software. Shouldn't the software adapt to the user instead of the user adapting by switching software? Shouldn't the skip button turn into a 15-second scrub button? Shouldn't it turn into a chapter button when you listen to books? Software is getting more malleable and now we're even entering the era of generative UI, not less malleable.
My job of keeping the complexity down and serving many different use cases actually gets easier and easier, not harder. I think we're on the right strategy. If you think about what these others, Anthropic and so forth, they're also super apps. They're trying to serve many different use cases, right? This is the opportunity that we have as well, to just be able to serve much more. The whole point is to leverage your existing distribution. You want to bring something to 760 million users day one, not zero users in the App Store. We still think about that strategy, but we actually think it's more vindicated now and it's easier to do than before.
Great. All right, thanks, Justin. Thank you, Alex and Gustav. That concludes our Q&A session.
Thank you, everyone.
Thanks everyone.
Thanks for coming.
For spending your morning with us. We appreciate your interest in Spotify. Thank you.