Good evening, and welcome to Universal Music Group's third quarter earnings call for the period ended September 30, 2026. My name is Gavin, and I will be your conference operator today. Your speakers for today's call will be Lucian Grainge, Chairman and CEO of Universal Music Group, and Matt Ellis, Chief Financial Officer. They will be joined during Q&A by Michael Nash, Chief Digital Officer, and Boyd Muir, Chief Operating Officer. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star followed by the number one on your telephone keypad. If you would like to withdraw your question, please press star one again. As a reminder, this call is being recorded.
Please also let me remind you that management's commentary and responses to questions on today's call may include forward-looking statements, which by their nature are uncertain and outside of the company's control. Although these forward-looking statements are based on management's current expectations and beliefs, actual results may vary in a material way. For discussion of some of the factors that could cause actual results to differ from expected results, please see the Risk Factors section of UMG's 2025 annual report, which is available on the investor relations page of UMG's website at universalmusic.com. Management commentary will also refer to non-IFRS measures on today's call. Reconciliations are available in the press release on the investor relations page of UMG's website. Thank you. Lucian, you can now begin your conference.
Thank you, and good morning to everyone. If I am a little congested, I apologize. I got a cold this morning, but not important. Let me begin with the quarter. UMG's total revenue grew by 12%. Core revenue grew 5%, excluding Downtown. Our margins increased by 40 basis points, driving adjusted EBITDA growth of 7% to EUR 729 million. Subscription streaming grew 9.5% and 19.2%, including Downtown. Downtown is the most important acquisition UMG has made since EMI. This was a timely, strategic move, and we are only just beginning to realize the full benefits. The numbers are already proving why we made this deal, and we continue to capture upside in terms of additional capabilities, revenues, and synergies. While quarterly results provide a useful snapshot, our focus is to keep evolving how we serve our artists and songwriters, whilst delivering sustained growth and profitability.
Making great music is emotional, instinctive, as well as unpredictable. Building lasting careers takes discipline, investment, and a long-term view. Put another way, success takes a combination of EQ to understand artistry and IQ to make sound commercial decisions. That balance matters even more in the age of AI. Generating tracks may be easier, but great art and great music cannot be made to order. Building a career people care about matters more than ever. Artists need a partner who helps them realize their vision, grow a fan base, and deliver commercial opportunities. And the vast majority of the music ecosystem value comes from career artists. We are building from a position of strength. The industry is healthy and growing. New fans, new markets, and new ways to serve artists keep expanding the possibilities. That makes us more ambitious, not more comfortable.
Let me start with the foundation of everything we do, creative success. Achieving it consistently is the hardest thing in music. It is also a defining strength of UMG. Year to date in the U.S., the world's largest music market, UMG artists hold seven of the top 10 albums, including four of the top five. This quarter, Taylor Swift returned to number one with "The Life of a Showgirl: The Encore," taking her to a historic 100 career weeks at the top of the Billboard 200, making her the first solo artist to reach triple digits. Ariana Grande earned her seventh number one album with "petal," while Karol G secured her fifth number one on the top Latin album charts.
At the same time, we are building a new generation of global stars, including Olivia Rodrigo, Noah Kahan, Morgan Wallen , and Olivia Dean, all of whom delivered strong carryover performance this quarter, and great music as we know is evergreen. 15 years after its release, Katy Perry's "The One That Got Away" was named TikTok's Global Song of the Summer. Charts capture a moment. We help turn that moment into deeper fan engagement and greater value while building the catalog of tomorrow. That is why sustained investment in artist development is the lifeblood of our business. It drives future returns, powers the growth of streaming services, and gives UMG a stronger voice in how the economics around the music evolve. Streaming has many years of growth ahead.
The model is working, but it can work better. At our Capital Markets Day two years ago in 2024, we projected that global paid music subscriptions would hit 1 billion by the end of 2028. We are on track to exceed that target. Industry forecasts now point to 1.5 billion paid subscribers by 2035. Growth will come from both more subscribers and more value per subscriber. In India, for example, UMG launched premium early access this quarter, giving paying subscribers access to selective releases 72 hours before they reached ad-supported services. That is something we pioneered. We will pursue more market-specific innovation like this where music engagement is high but paid adoption remains low. There is also room to earn more by giving fans more differentiated subscriptions, better experiences, and benefits they value. That is one component of Streaming 2.0. Another is integrity.
Making sure compensation more fully reflects the value artists and their music bring to the platforms. Let me say more about what that means. When we launched our artist-centric initiative in 2023, we were the first company to speak out about the proliferation of lower-quality functional content that in some cases barely passes for music. We saw the wave of AI slop heading towards the music ecosystem. In early 2025, roughly 10,000 AI-generated tracks were arriving each day. Today, more than 90,000 fully AI-generated tracks are uploaded daily. The volume is growing faster than the safeguards. We have made real progress. We introduced first-of-their-kind anti-dilution protections in our platform agreements. We are helping establish AI labeling and chart standards for licensed human-led recordings. We helped launch the Streaming Integrity Initiative and have taken legal action against bad actors while supporting legislation that protects artists and their livelihoods.
Many of our DSP partners are also taking meaningful steps, and we welcome that progress, but the scale of the problem is still outpacing the guardrails. Let me illustrate this with an example. Look at the blues charts. The blues represents only a tiny fraction of total music consumption, but its cultural and historical importance is enormous. On some major DSPs blues charts, as many as five of the top 10 albums are attributed to AI-generated acts without clear labeling or disclosure. The blues is the North Star of authenticity in music, born from deeply personal human experience. Yet its charts are being polluted by AI content that synthetically imitates that emotional impact. Let me be clear. This isn't an argument against AI. It's an argument for licensed innovation, transparency, and fair economics because this is not harmless clutter.
AI slop degrades the fan experience and diverts value from real artists. It serves no one. Not artists, platforms, or consumers. We know how to build win-win agreements with our DSP partners, and the shared problem now demands shared action. We will enforce our contractual protections and insist on clearer labeling across the ecosystem. Our research is clear. Fans do not want undisclosed AI slop in their music experience. There is also plenty of evidence that some of that consumption is fraudulent. That raises a basic question. What does market share mean when AI slop and music by real artists are counted as the equivalent? The uniform application of pro-rata payments makes the problem worse. By treating AI slop as the same as the greatest songs in the world by some of the greatest artists, it incentivizes industrial-scale content production and dilutes the compensation earned by real talent.
Put plainly, unethically generated AI tracks should never be compensated on the same basis as music that creates real and lasting demand. And why would a music service even want to host them? Career artists and songwriters create the vast majority of the music ecosystem's value. They build lasting connections with fans. They bring subscribers to streaming services and give them a reason to stay. That demand extends beyond streams into concert tickets, vinyl CDs, merchandise, and committed fan communities. No company has done more to develop and sustain those careers than UMG. We've represented nearly 70% of the world's best-selling artists over the last decade. And UMG artists produced nearly half of the 900 music tracks streamed more than 100 million times in the U.S. just last year. Streaming services depend on those artists and the music that attracts subscribers and keeps them engaged.
That's why the structure of our VSP agreements has to change, so the compensation better reflects the value different content creates. The larger opportunity is still ahead. More subscribers, more value per subscriber, a better fan experience, and economics that reward the artists, and music that creates lasting demand. The opportunity expands beyond streaming. We also see opportunities for long-term growth by doing more for more artists. From expanding our menu of label services and growing our geographic reach, to building additional capabilities in direct-to-consumer sales and merchandise to create new live experiences and super fan products. These strategies are interconnected. They do build on one another. Virgin Music Group is already the industry's second-largest artist and label services company. Our ambition is to become the leading partner to independent labels, entrepreneurs, and DIY artists, giving them a way to work with Virgin on terms that fit their ambitions.
That broader reach gives us a sharper view of where talent and demand are emerging, especially in fast-growing markets, where large audiences and dynamic local music companies and cultures are not yet fully reflected in market value. As that network grows, so does our ability to deepen fan relationships. For example, our recent agreement with Apple Corps expands merchandise licensing and e-commerce around The Beatles. Our partnership with a tech startup, Hook, gives fans a licensed way to remix and share the music they love, with attribution and compensation built-in. Different partnerships, same goal, help artists turn attention into participation, community, and long-term value. AI can extend that further, giving artists and fans new ways to create and participate. Here again, two approaches strengthen each other, protect human artistry, and embrace innovation. Our partnership with ElevenLabs is the latest step in a broader artist-centric AI strategy.
ElevenLabs, Spotify, Udio, and Klay are developing new licensed forms of fan creativity. Stability AI and Splice are developing professional tools for artists, and NVIDIA is helping us advance music discovery and engagement. Across these partnerships, the standards are the same. Artists and songwriters opt in whenever their voice, image, name and likeness, lyrics or compositions are used and the creative community shares fairly in the value that is created. We are moving from principles to products and from products to revenue. We expect these partnerships to begin generating incremental revenue in FY 2027. More artists, more markets, deeper fandom, new tools for creativity. We are still early. The economics will compound as these opportunities connect. We will pursue them all with discipline. Each must address a real need for artists, provide a best-in-class solution, and earn an attractive return on the capital we commit.
To capture those opportunities, UMG itself must keep evolving. We will continue to reduce costs, operate more efficiently, and improve margins. That is a clear financial objective. At the same time, we will sustain what drives this business. Investment in great artists, songwriters, and catalogs. An enhanced operating model will also give us more capacity to invest behind greatest success and the next wave of growth. When we announced our strategic organization redesign in February of 2024, we committed to delivering EUR 250 million in annual run rate savings by the end of 2026. We also committed to strengthening what makes UMG distinctive. Our leadership in artist development, the individual creative identities of our labels that help us attract the widest range of the world's most successful artists, and the entrepreneurial culture that drives that innovation. That program is now complete. You can see the results in our cost structure.
Excluding share-based compensation, SG&A as a share of revenue has declined by more than two percentage points since we began the realignment, even after reinvesting some of those savings in the business, as we discussed last quarter. Over the same period, we paid a growing share of revenue to artists, songwriters, and label partners. This increase reflects the broader ways in which we work with artists and the associated deal structures, as well as the growth of our artist and label services business, which we include in Virgin Music Group, which provides a different range of services and pays a higher share of revenue back to artists and label partners by design. Looking ahead, our continued transformation will allow us to expand our margins by leveraging technology to redesign processes and bring together non-creative functions currently spread across labels and regions. That work is already underway. Our approach is straightforward.
Strengthen what makes UMG distinctive. Simplify where scale improves service and lowers cost, and invest where we can create the most value and build durable advantage. That means keeping creative judgment local, leveraging UMG's global reach, and empowering our best executives. Frank Briegmann is one example. He's led our business across Central Europe and the Nordics for many years, helping artists build long-term careers through creative leadership, commercial judgment, technological expertise, and respect for each individual local culture. That track record is why I asked him to take on a broader group of markets in our overseas Canada, Australia, and New Zealand, along with the majority of Europe. His expanded remit will bring these markets closer together, speed collaboration, and help more artists reach audiences across borders while preserving what makes each market distinct. Scale matters when it makes artists, songwriters, and local teams stronger.
That is why we are adding company-wide expertise in data, AI, operations, and partnerships. Hannah Poferl, our Chief Data Officer, joined from the New York Times and recruited Òscar Celma, our Senior Vice President of Applied AI and Machine Learning. Òscar brings experience in music discovery and personalization, including at Spotify. Together, they are building a central team to sharpen audience insight and improve talent discovery tools. Tseyin Foo, our President of Global Supply Chain, brings experience from Nike and Skechers. She's strengthening execution across our digital, physical, direct-to-consumer, and merchandise operations. Libby Bush, another, joins us next month from CAA as President of Global Brands and Commercial Partnerships. Her remit includes improving economics and ad-supported music by connecting UMG's premium artist content, fan insight, and global brand relationships. The goal is simple: better insight, faster execution, and more opportunity.
Through all of it, music and artist development remain our center of gravity. I'm committed to extending our leadership in the years ahead by continuing to invest in developed career artists and securing win-win partnerships with DSPs that properly reward their work and ours. With that in mind, we've begun the next phase of our transformation, which will run through to the end of 2028. We'll build on the organizational redesign and look closely at how we negotiate price and execute platform agreements, how our central commercial organization operates, how our labels and international business is structured, and how we use data across the entire company. We'll share more about this scope and financial objectives when we report our full-year results. Thank you for that. Now, Matt, I'm going to turn over to you.
Thank you, Lucian. We're encouraged by the progress we've made in several areas of the business this quarter, while recognizing that there is still important work ahead of us. As Lucian discussed, we recently completed the actions of the past, about EUR 250 million organizational redesign. We know there are additional opportunities to improve efficiency and strengthen our financial performance, and the next phase of this work to build the UMG of the future will focus on both further improving our cost structure and accelerating growth across the business. We look forward to sharing more details with you soon. For now, let me walk you through the quarter. As in prior quarters, I'll discuss our results on a constant currency basis, and I'll also provide the underlying performance, excluding the contribution from Downtown.
The integration of Downtown is well underway, and it is contributing to our business as we expected it would when we agreed the deal. We passed an important milestone in the integration a few weeks ago, and our Virgin and Downtown employees are now working within the same people systems. We have already delivered the initial synergies target on a run-rate basis, and the team sees additional opportunities.
Our total revenue for the quarter grew 11.7% year-over-year to EUR 3.4 billion. Downtown contributed EUR 195 million of revenue during the quarter. Excluding Downtown, revenue grew 5.3%, driven by healthy growth in recorded music. Adjusted EBITDA grew 9.0% to EUR 729 million. Downtown contributed EUR 13 million of adjusted EBITDA during the quarter. Excluding Downtown, adjusted EBITDA grew 7.0%, with adjusted EBITDA margin improving 40 basis points to 22.4%, driven by margin expansion in recorded music. Let me turn to our segment results.
Recorded music revenue grew 14.7% for the quarter, with Downtown contributing EUR 162 million. Excluding Downtown, recorded music revenue increased 7.4% year-over-year. Recorded music adjusted EBITDA increased 14.9% and grew 12.7% excluding Downtown. Margin expanded 1.2 percentage points to 26.2%. The benefit of revenue growth and lower relative A&R, as well as improved cost discipline across marketing and overhead, more than offset pressure from revenue and repertoire mix. Looking further at recorded music, year-over-year subscription revenue growth accelerated compared to the first half of the year, up 19.2%, or 9.5% excluding Downtown. Growth continued to benefit from Streaming 2.0 pricing initiatives, which contributed slightly more than the 3.5 percentage points we saw in Q2, thanks to a partial benefit from Apple's recent retail price increase. Market share remained a headwind, although the drag moderated this quarter.
The remainder of the growth reflects the underlying healthy subscription performance of the industry and other recurring items such as minimum guarantees, audits, and accruals. Ad-supported streaming revenue increased 13.5%, or 4.7% excluding Downtown, with growth across several of our key video and digital partners. Better per-play rates more than offset the ongoing shift in consumption to short-form platforms. Physical revenue, which had no contribution from Downtown, continued to be strong, growing 17.4%. Performance was broad-based across North America, Japan, and Europe, and across frontline and catalog. Releases from Mrs. GREEN APPLE, Stray Kids, Olivia Rodrigo, The Beatles, Rolling Stones, and Ariana Grande all performed strongly and more than offset the pre-release sales of Taylor Swift's album last year. The strong level of physical sales we have delivered this year across both artists and geographies gives us continued optimism about the contribution from this category going forward.
Although the fourth quarter will be lapping exceptional volumes from last year. Licensing and other revenue excluding Downtown declined as growth in synchronization and audiovisual revenue is more than offset by lower live and related income. Turning to music publishing, revenue grew 7.5%, or 1.5% excluding Downtown. Excluding Downtown, digital revenue grew 3.9% as subscription growth more than offset softer ad-supported revenue. Synchronization revenue declined 4.7% as lower than anticipated film and television activity in the period more than offset improvements in advertising. Adjusted EBITDA declined 3.5%, or 4.3% excluding Downtown, impacted by the timing of expenses in the prior year period. As we have discussed in recent quarters, music publishing results can be affected by the cadence of industry distributions and the timing of synchronization activity, which creates variability in quarterly growth rates. That said, growth this quarter was below the level we expect from this business over time.
We continue to believe this will be a mid-single digit growth business in 2027. Turning to merchandising, revenue declined 4.6% year-over-year. Direct-to-consumer revenue grew as we broadened the rollout of our ship-to-home program for merchandise on certain tours. However, this was more than offset by lower touring income due to a larger number of high-profile tours in Q3 last year. Adjusted EBITDA declined EUR 2 million- EUR 11 million due to lower revenue, while margin contracted 60 basis points year-over-year, but was up nearly 10 percentage points compared to the first half results. As we've discussed, we continue to make structural changes to this business with a particular focus on strengthening our direct-to-consumer capabilities and improving profitability over time. Overall, we're encouraged by the progress we saw in several areas this quarter, including the acceleration in subscription revenue, continued strength in physical sales, and improved profitability.
At the same time, we recognize that there is more work ahead of us. As you heard from Lucian Grainge, that work is underway. We're focused on building on our progress, addressing the areas where there are opportunities to improve, and driving stronger growth and profitability over time. With that, Lucian, Michael Nash, Boyd Muir, and I will be happy to take your questions. Operator, please open the line for Q&A.
Thank you. We'll now open the floor for analyst questions. Each analyst is allotted two consecutive questions. To ask a question, please press star followed by number one on your telephone keypad. When preparing to ask your question, please ensure your phone is unmuted locally. Our first question comes from the line of Eddie Young from Morgan Stanley. Your line is open.
Thank you. First of all, struck by your commentary on Downtown, where you spoke about the beginning of synergy delivery as well as the major strategic benefit. If I annualize the Q3, it looks like you've already taken four turns off the acquisition multiple. Perhaps you could clarify if some of that is lumpiness. Moreover, would love to hear more on what you're seeing in the business and your hopes for it. Then second, encouraging to see the operational gearing on the back of the strong subscription and recorded music growth. Can you help us understand what are timing benefits versus what's more structurally right-sizing costs? Just to give us an idea of the underlying margin performance. Thank you.
Thanks for that question. I will maybe start with the second part of your question on the margin. Certainly very excited that we delivered the 40 basis points of year-over-year improvement, and an even larger amount on a sequential basis. That is really the result of strategic decision making about where we are investing in our business on a continued basis. We indicated that on our last call that we would be looking at a number of those areas. You started to see the impact in the third quarter.
That work will certainly continue as we go forward here. The next phase of our transformation and looking at the commercial organization structures, as Lucian Grainge and I both said, more details to come on that. But we continue to look at the right way to run the business most efficiently while delivering the maximum amounts of operational growth at the same time.
Certainly expect to see continuation of the trends that came through this quarter. Then on Downtown, I will start and Lucian Grainge pick up comments on there. Certainly, as we said, it is the most important acquisition that we have done since EMI. As the team has made very good progress on the integration plan. Most of the benefits already now in the run rate there. As we have gone through that, the team sees further opportunities, and it continues to be a very important part of the overall recorded music business. So we are very happy to be playing there on an even larger scale than we were previously.
Yeah. Look, developing artists in this space is substantially growing throughout the whole of the world. Latin America, obviously, the mature Anglo markets everywhere. Our strategic intention with the acquisition, as well as our Virgin Music Group strategy, is to partner with real people, with real artists, real labels, real entrepreneurs at every stage in their career. Strategically, it is an exciting leap forward in our ability to grow in this space. We really like what we see. There are some very, very strong people there, some great business opportunities. It is a great opportunity to add to a business that has propelled us to the number two position quickly in this fast-growing artist and services sector. We are very confident with what we see, and we know that it is just the beginning to unlock the value of that business within our entire global ecosystem.
Your next question comes from Peter Supino from Wolfe Research. Your line is open.
Hi. Good afternoon, everybody. I wanted to ask about your cost structure and your comments about improving it. I heard you loud and clear that you will provide more details soon. Without asking for any specificity, I wonder if you could talk about functions or areas or processes at UMG where you see incremental efficiency opportunities. As a second question, I wanted to ask about your market share results. They were much better than the first half of 2026. That is consistent with your messaging throughout the year. I just wondered, reflecting on the first half of 2026, if you would share any lessons learned or process improvements that we could consider in forecasting the recorded music business for 2027. Thank you.
Yeah. Thank you for the question, Peter. I will start with the cost structure and then move on to the market share piece. I think Lucian Grainge touched on a number of pieces in his comments. The key for us is finding the right balance that we are still taking the commercial decisions where they should be taken. Then we look across the whole business at how we best support those commercial decision-making, where there is opportunity to bring the scale of UMG on a global basis together, and improve both the efficiency of the work and also the effectiveness in supporting our frontline operators. We do that.
I would say you should expect to see changes to the cost structure across almost every facet of the company or operating areas of the company as we run the business the right way that it should be run with the tools available to us and the way the industry will demand of us going forward. More specifics on that, as we said, to come. In terms of forecasting market share results, we do not see new releases from artists in the same cadence as you see holiday season impact in retail, for example, where you know each year exactly where that is going to be. We certainly expect to see another strong slate of performances from our artists next year with the new releases that will come out there. Glad to see the improvement as the year has gone on, as we discussed it would.
We will keep doing what we do best. That is, Lucian Grainge touched on in his remarks, the remarkable streak of the percentage of the top streaming artists and albums that are coming from the artists that we work with. The timing of that is in the genius of the artists that we work with. For those of us who have tried to put models together around that, it is not always so easy to predict, but we know that we will see that genius come through.
But I'd add that obviously we watch market share closely. But we don't and we can't manage the business on one quarter's shared number, and I've been quite consistent about that. A company's creative performance remains strong across the genres, geographies, and artists at all stages in their careers. You'll break a new artist that will become huge. The frontline headwind moderated, and the deep catalog continued and was stable and continued to grow. Quarterly share moves with release cadence. Obviously, I suppose prior year comparisons, individual geographies, and the changing mix of consumption across those platforms throughout the world. We've got huge acts in Japan, and they stream less in Japan than the equivalent huge artists might stream in the mature Europe, in the larger, more significant Anglo markets, for example.
And this next phase, we've got two clear objectives: to improve the margins as well as evolve our organization. We're going to continue to invest behind artists and songwriter development. We're going to continue to extend our global reach while strengthening local market expertise. Use our scale, use the data. We've brought in new people. We've refreshed, and work on our own tech stack, where we've got the opportunity. We have to improve service, speed, and cost.
Your next question comes from the line of Stephen Laszczyk from Goldman Sachs. Your line is open.
Great. Thank you for taking the questions. Lucian, you spoke a good bit about the need to create guardrails around AI content. I was curious what tangible actions you would expect to see perhaps over the next year or so out of the DSPs, out of regulators, or perhaps even out of the labels themselves on this front. I'd be curious how much around this you think is within your control as the largest music label versus where you might feel like there's a need for greater buy-in from the community.
Michael, to actually handle that sort of thing. But we spend an enormous amount of time together.
Stephen, thank you for your question there. With respect to DSP protections, we continue to negotiate and secure some very important protections in our agreements with the services. They vary, of course, according to the partnership. They include four key elements. First of all, AI models can't be trained on our artists' work without consent. Second of all, AI recordings trained on our content without authorization have to be removed. Third, AI content that misappropriates the identities of our artists and infringes on their right of publicity will also be removed. Perhaps most importantly, you've heard about this a lot from us, it's our aim that our contracts fully. In our contracts, excuse me, that fully AI-generated music doesn't dilute our artist royalties. Various monitoring requirements are also included in the agreements. We negotiated and secured key protections, in our YouTube deal, recent Spotify agreement.
We've highlighted AI protections and announcements with Roblox and NetEase. This indicates the successful efforts to secure AI protections across multiple channels in the music ecosystem. Regarding labeling, AI detection technology is now available, and research indicates that 80%- 90% of consumers insist that AI content should be appropriately labeled. Keep in mind, this is a problem for the services too. Questions about individual platform plans to manage AI music should also be taken up with them. They're incentivized to address these developments which undermine the user experience and the integrity of their services, as Lucian Grainge pointed out. For example, Spotify recently announced that they've taken more than 50 million tracks identified as AI slop and spam out of their recommendations, effectively partitioning this content off from consumers.
This follows their announcement earlier in the year about removal of 75 million spammy AI-generated tracks from their service entirely. Deezer is launching a kill switch feature that allows users to exclude fully AI-generated content. I would note there that 74% of U.S. consumers have expressed interest in having that type of interaction with the service to be able to exclude AI content. Consumers don't want to be deceived. They don't want to have to guess if they're being force-fed AI content. They want a connection with real artists. Finally, Apple reported earlier this year about its efforts to address the flood of AI uploads, that included the exclusion of two billion fraudulent streams last year. So we welcome these steps and further innovation to counteract harm that's posed by unethical AI on these platforms. But this is not sufficient. I want to underscore this.
We are facing a situation right now where, as Lucian Grainge pointed out, first quarter of 2025, 10,000 AI tracks uploaded a day. Less than two years later, that has increased by an order of magnitude to almost 100,000 AI tracks per day. This is ecosystem pollution at a massive scale. Clearly, the ecosystem is damaged by this flood of infringing content. It is completely detached from real, meaningful artist-fan relationships that drive the music economy. While most of the value in the ecosystem is being created by career artists, this is the antithesis of career artistry, this flooding of the platforms, with this glut of pure AI content. It is very important in terms of the integrity of the user experience, and especially music discovery, that we work with our partners to better address this problem.
I would add there that the reality is it is a new technology, and it is developing activity in real-time, and it has exploded over the last several months. It has got to stop. We expect our partners to address it.
Your next question comes to the line of Jason Bazinet from Citi. The line is open.
I just had two quick questions. You mentioned the share headwinds moderated in the quarter. I assume that is a sequential comment versus, I think it was 1.5 points headwind last quarter. Just clarifying that. Then second, you mentioned that you do have to enter new agreements with the DSP partners to reflect AI. I guess the question is, today, would you say that AI has impacted your financials in terms of market share erosion, or it is just that we are seeing an order of magnitude increase in uploads, but not enough consumption given your guardrails to impact market share? Thanks.
Thanks, Jason. I will take the first part of that, it is probably a shorter answer, then hand it over to Michael. On the market share in the third quarter, as we said, the headwind reduced. That was a sequential comment. A smaller headwind this quarter than we saw in the second quarter, as we saw the continued success of the releases that we had in the second quarter then into the third quarter. You are thinking about that the right way. Michael, I will hand it over to you for the question around the DSP agreements.
I think that, Jason, the question kind of also went to how we are thinking about the status of our efforts to address AI and what impact that has had. To be clear, our previous comments were concerned with what we could measure in our streaming results at that point. We had presented data that was focused on the developments in FY 2025 on earlier earnings calls. Our comments today describe what is happening to the ecosystem around us three quarters later. As everyone understands, and as Lucian Grainge articulated very well in his opening comments, the development, dissemination, and impact of AI continues rapidly, and the scale of the problem is accelerating.
The volume of purely AI-generated content and fraudulent activity continues to proliferate at a significant rate, and we believe that this is increasingly destroying traditional measures of market share and can divert economics away from legitimate artists and content. I gave you some of the metrics associated with that Lucian Grainge has highlighted, in my earlier answer to the previous question. The dilution threats are growing, fraud threats are growing, and these harms are not confined just to the platforms. They extend to artists, songwriters, and rights holders whose works have been taken without authorization or compensation to train competing AI products. This glut of AI content on these platforms does further risk devaluing the rich human-created catalogs that sustain them. The impacts from AI content are not limited to just streaming platforms, but can affect a variety of products and revenue streams.
In terms of our protections, we do have multifaceted protections in place, I enumerated them, with many of our partners with the goal of mitigating the impact of this type of activity, including specifically addressing royalty dilution. Part of the solution is ensuring that those provisions are properly enforced and that artist royalties are not diluted. This is not enough. As Lucian Grainge discussed, we also believe the industry needs to go much further. Our artists create enormous value for the platforms, and partnership economics should reflect that value rather than reward industrial-scale content fabrication or fraudulent consumption.
Yeah. It's not real music. It's not real art. It's not real listening. As I said, it has to stop.
Your next question comes to the line of Will Packer from BNP Paribas. Your line is open.
Hi there, many thanks for taking my questions. Firstly, just following up on the previous question. Would it be fair to think that perhaps as part of this push to reform DSP deals to reflect the true value of the content you bring to the music ecosystem, that we could have almost a supercharged version of the artist-centric model? So where perhaps elevated royalties are paid for the major artists who drive it. Is that something that you're pushing for, or have I misunderstood? Secondly, we've had some mixed feedback from industry stakeholders on the willingness of major artists to participate in the AI tier at launch or Spotify's AI tier. Is that fair? How have your conversations gone, and should we expect a decent portion of your, let's say, top 20 or top 50 artists to participate in such a tier at launch?
Thanks very much.
Will, thank you for your question. With respect to the DSP deals, I liked your characterization of supercharged artist-centric, because our entire focus is on ensuring that our artists are properly compensated for the value that they bring to the platforms. You are referencing a number of specific reforms and innovations that we have worked with our partners to introduce in terms of adjusting the pro-rata model. In terms of how we envision the partnerships changing, Lucian Grainge talked how most of the value in the music ecosystem being created by career artists. We have previously presented what we roughly generalize as the 80/80 rule. 80% of the subscriber acquisition and 80% of the subscriber retention is based on fans' engagement on the platforms with real artists, career artists, classic catalog, emerging new talent. That is what is driving the value in the ecosystem.
We have got to evolve our partnerships so the models more fairly reflect that value. I want to double down on a stat that Lucian Grainge mentioned, because I think it is very important here to consider what we are talking about in terms of the value creation that UMG brings. UMG artists were responsible for almost half of the 900 tracks that streamed more than 100 million times in the U.S. last year. At that meaningful threshold of artist content that fans are seeking on the platforms, UMG was responsible for almost half of it in the U.S. Volume-based market share metrics do not fully account for platform value creation by artists that matter. We intend to address that with all of our partners. Again, I liked your characterization of a supercharged artist-centric model.
With respect then to your question regarding artists opting in of the Spotify AI tier, and of course this relates to other deals recently announced with ElevenLabs and deals that we have previously announced with Klay and Udio, where we are working to develop a new super fan tier that enables fans to interact with artists' content. We obviously do not disclose in-progress details like that, but our engagement with the creative community is going extremely well. We have engaged with thousands of our artists talking about this product model. We have already converted a significant percentage of those discussions into opt-in agreements. We continue to make excellent progress. We are building our portfolio of partnerships in this category, which is certainly contributing to that, and we expect to achieve a critical mass of artist support to help launch these services. It is going well. We continue to press ahead.
We expect successful launches of these tiers.
One thing I would like to add to it. The answer is obviously yes. Real artists drive the value of the platforms. They drive it to the platforms. And they deserve to be more highly compensated for how critical they are. And we will work with our partners to make sure that that happens.
Your next question comes from the line of Julien Roch from Barclays. Your line is open.
Yes. Thank you very much for taking my question. Matt, I apologize to ask for precise number, but if we could come back to the breakdown of the 9.5% between price increase, market share, and minimum guarantee audits and accrual. So price increase slightly better than the 3.5%. Are we talking below 4%? Market share, you say sequential improvement, but by how much? Are we talking - 0.5%, - 1%? And then on the minimum guarantee audits and accrual that you mentioned, what is the benefit? Are we talking 50 on the basis point? Sorry for asking for precise number. That is my first question. And then on the EUR 250 million of cost cutting, Lucian said it was done.
But at the first half, you told us that you had delivered EUR 150 million, and that the EUR 60 million in 2026 and the EUR 40 million in 2027 that was expected out of the EUR 250 million was basically completely reinvested. So can we have an idea of the remaining EUR 100 million? How much is going to the bottom line and how much is reinvested? So for the EUR 250 million, before the next round of cost cutting. Thank you.
Thank you, Julien, for the questions. You highlighted the key components of the 9.5% subscriber revenue growth. We saw a modest benefit of Apple's price increase come through in 3Q, much more of it to come through in the fourth quarter. That is why we said we were slightly more than the 3.5 percentage points we grew in the second quarter. Market share continued to move in the direction you expect with the strength of the artist releases that we had, so less of a drag. I would reiterate that when you think about that as a year-over-year drag, as we have discussed previously, 2025 was our strongest market share in over a decade. So the market share performance this year is still at very strong and healthy levels.
We have the items that always give us a little bit of variability because of the various deal structure elements. Those were a net positive contribution year-over-year in 3Q this year versus last year, but not at a massively material level or anything. In terms of your comments on the cost cutting, I am not sure I fully understood or necessarily agreed with some of the part of your question.
We have certainly delivered significant amounts of the EUR 250 million through mid-year, more than the EUR 150 million that you suggested. I think the important thing to take away is that with the actions that we have continued to take into 3Q and into the start of the fourth quarter, and you will have seen a number of the announcements over the past few weeks around some of those actions, that when you take those actions on a run rate basis,
they deliver the EUR 250 million of cost savings since the program was begun. Obviously, a full impact of the EUR 250 million this year is not there because some of the actions were taken, as I said, just a few weeks ago. As you think about the incremental benefit next year over this year of getting the full benefit of those actions, as we estimated previously, we still expect a EUR 35 million-EUR 45 million year-over-year benefit as you get a full year impact next year of those actions that we have taken. So pleased with the actions we have taken. As you mentioned, some of them are reinvested in the business. Lucian Grainge touched on this in his prepared remarks. We have had some overhead that has obviously come in from acquisitions that we have made since the program began.
Whether that be PIAS, Mavin, or the whole host of the other acquisitions we have discussed over the last two to three years. We have strengthened some part of our businesses. You heard some of the individual names around that, again, from the prepared remarks, as we continue to look at those capabilities that we need to strengthen to deliver the growth in the areas where we said the growth will come from. So taking the actions we have taken to reorganize the business gives us the cost capacity to then invest in the capabilities necessary to build the business to deliver on the strategic objectives going forward. Then you have the ongoing, call them inflationary items in the business such as annual salary, et cetera, those types of things.
Some of that has been reinvested, but it's also allowed us to restructure the business to have the tools necessary to implement the strategic areas that we said we're focused on and we're excited now about moving on to this next stage of organizational design, as we look at the opportunities ahead of us and we'll get into more specific details on those next time we speak.
Your next question comes from Adrien de Saint Hilaire of Bank of America. Your line is open.
Thank you. Thank you very much for taking the question. Two of them, if you wouldn't mind. There were some headlines in Bloomberg about the potential discussion around a take private. Just wondering if you could maybe comment on that. Then secondly, should we expect that Streaming 2.0 deals will have an impact in your 2027 goals? I think previously you said that there shouldn't be any escalators really, but I just wanted to double-check on that. Thank you so much.
Adrien, thank you for the questions. With respect to your first question, as always, we don't comment on press speculation. We never have. We're not doing so now. In terms of your second question, I'll let Michael kind of jump in as well here, but certainly, as we start thinking about the cadence of the price increases. In the fourth quarter, we'll start lapping a little bit some of the price increases in certain markets from last year. That will obviously build as you go through the year. We'll start to see more of a benefit of the Apple retail price increase, not a 2.0 agreement there yet. There's a number of factors that will build into pricing next year. Michael, I'll let you go into some more specifics.
Yeah. Happy to take the baton pass there, Adrien. Thank you for your second question. With respect to 2027, of course, we do not provide quarterly or annual guidance on any of our metrics, including subscription revenue. With respect to how we are thinking about next year, the growth rate is still going to largely be tied to underlying volume growth in subscribers and then the geographic mix of those new subscribers. Matt talked about the pricing benefits and how you should consider those moving forward. I think it is also important, in the subscription category, to note that Spotify has the AI tier that they are developing. There are other licensed service plan launches. We are in discussions with other platforms about similar offers. These tiers are going to drive new revenue. They are going to drive higher ARPU on existing DSPs.
Our monetization is obviously going to depend on the timing of the ramp-up of these tiers, but we have confidence there is meaningful demand for these types of products. It is too early to have a view on market share for 2027, but stepping back for a broader perspective, the macro trends are very positive. Third-party research and industry data strongly support our projection that we gave at Capital Markets Day 2024 that we will hit a 1 billion-subscriber mark by 2028. As Lucian Grainge said, we are optimistic that we may exceed that projection. New industry projections, 1.5 billion subscribers by the middle of the next decade, give you a sense of the runway moving from that point forward. We are confident in the macro trends with respect to the volume growth, which will be the most significant driver of subscription revenue growth in 2027.
That concludes the question and answer session as well as today's call. Thank you for joining. You may now disconnect your lines.