Hey, thanks for joining us. My name is Sean Diffley from the U.S. Media and Telecom Research Team at Morgan Stanley, and we are very excited to have the CFO of Spotify, Christian Luiga. Thank you for joining us, Christian.
Thank you.
Is this your second year at Spotify in Barcelona conference?
Yeah. It's not my second year here. I think it's my eighth year here.
I was going to say, Saab and Telia before that.
Yeah.
You are a regular.
Telco. Yeah, Telco was also here.
All right.
It's a good conference.
Thank you. We're very excited to get into talking about Spotify, and 2026 has been a big year for the company. You've been leaning into AI. You've had a record pace of product launches. We've seen the beginning signs of a turnaround in advertising. From the CFO seat, which you've been in for just over a year, how has the year compared to your expectations, and what are the things you're most focused on as we head into the year end?
Well, it's been a fantastic year. I mean, there's so many things happening in the world and so much things happening in Spotify. Starting out this year with two longtime persons, Gustav and Alex, they've been also co-presidents for some years and now are the co-CEOs. Raising the ambitions was the promise they made in the beginning of the year. We got to do, I think, a good way of explaining ourselves at an investor day in New York. We haven't had one for a long time. In that also, we have raised our ambitions, but also primarily delivered on what we promised. As you said, we promised already a year ago to actually make sure we do something about the ad revenue growth. We delivered on that. We talked about that we wanted to invest more in AI.
We are forerunner in AI, and we do have discipline and cost control, but we needed to do some step changes, and we did that, and we're going to moderate the cost going into the next quarter. We have built a platform for add-ons and moved that forward so we can actually have that going forward, which is going to be something we talk about soon, about the monetization and so on. Finally, I would like to say that we have delivered on our subs and MAU and growing our base in the way we want. On top of that, then being able to go with our books that we started at the bet two and a half years ago to another 160 countries. So it's been a fantastic year.
Right. Yeah, lots to look forward to.
I am excited for your Wrapped coming out in the next few weeks.
That is every year, but I have it in here, mine. Maybe you can see it tonight.
Oh, we got a little preview.
It's only mine.
All right. I want to see your top artist. I don't know if you're willing to share that with the group. Who have you listened to most this year?
My top artist was actually U2.
There you go.
which I was surprised about. It is quite interesting to see actually, and I am going to You can see what number of listeners you are in the world and what percentage.
Yeah.
One of the artists I listened to, a young lady from Maryland, U.S., I was number 20.
Wow.
Yeah.
Who was it? Which artist?
Maggie Rose.
Okay.
You should listen to her. She's a little bit country type, but she's super nice voice.
You're going to help her out by giving her a shout-out. Going back to the investor day in May, you outlined this mid-teens revenue growth target through 2030. I was curious if you could unpack for us the contribution from subscriber growth, core pricing, the paid add-ons that you've talked about, and advertising.
Yeah, I mean, durability and diversification is really important. We start in that and think about our subscribers and users. There's 3.5%-4% of the users in the world that of the people in the world that pays for Spotify. Of course, you won't get to maybe 50%, but if you can just imagine, could it be 6% or 8% or 10%, still low numbers, it's going to have a massive impact on our growth. We have consistently been able to deliver around 25 million subscribers every year. Some have thought, like, if you're building a bigger- and- bigger base, it's going to be harder to grow 25 million every year. But we're doing that every year. We do that because we have a model that we are trained on, first in the Western countries, then in LATAM, and then going further.
We have a model where we have this durability and starting with our free users that actually is our base for converting into paid users. That machine is working. We feel very confident about that, building on that and growing our subscriber base. Then we have the ARPU side, and there you get the help from the add-ons and the new products, like when podcasts and also books came into play. We continue to grow that based on that we're pushing in value, and the value we can price for over time. But on top of that, we can maybe do add-ons where we price them separately. That is really the two mechanisms for driving this as we go forward. That is really how we look at the power of revenue growth.
Great. I wanted to dig in on that pricing power point. As you pointed out, you have added considerable value to Premium and taken prices across markets. I think India and Indonesia are two markets where you seem to be turning the dial more towards monetization after you have had really strong MAU growth. Maybe just talk to us about some of the things you are doing to push that lever a bit.
Yeah. On the pricing, before we go to Indonesia and India, which we will talk about a little bit, the price is driven by value increase. I think that is sometimes what we think about more maybe than the outside. Sometimes we increase pricing faster and sometimes slower. It could be 12 months one time, and it could be 18 month- 24 months another time. We are very conscious of building value and then doing price increase. When we went out with Audiobooks, it took between six and 11 months to do price increases after that. It depends a little bit how we drive these different things. But the value is so important because when you lose that grip of actually delivering real good value, then the pricing becomes a much tougher monetization game over time.
You have that gap that we have always talked about, value to price, you keep that. In India and Indonesia, we have said earlier this year that we are having a conversion around just slightly below 10%. That is not different from the other countries that we have had going through this pattern, where we saw very early. Actually, it was people internally in Spotify, and I heard this story many times. They were sitting many years ago and said, "We should leave LATAM because we have so low conversion." Then they saw this, it was the same as it was in some European countries, but five years ago, and they said, "No, we are going to continue." Now LATAM is wonderful, both from a revenue point of view, engagement point of view, and Spotify brand point of view.
We have seen the same pattern in the other regions, but it is just some years later. So, we are going to catch up on that.
That makes a lot of sense. What would you say the key learnings are around willingness to pay and churn profiles through successive pricing rounds across different markets?
Well, when we do this as careful as we do, we see the same kind of churn pattern, which is very low, and then returning quite quickly after our price increases. It hasn't changed over the years. That's why we are so stuck to the model. You should also remember, every time we do a price increase, we have a learning from that. It's a little bit like our models train every day. This is fun to be in this company. It's actually, I think, a little bit better than when I was in Telco. In one way, there's things we're learning a little bit slower than in Telco, but actually, the thing that this company does so well is to work with the data all the time and see what did we learn when we increased pricing in Belgium?
What did we learn in applying that all the time? Thinking about both cadence and the level, and what value did we bring in before we did it.
No, that makes a lot of sense. Obviously, leveraging data to decide whether the next dollar of monetization comes from the base subscription or an add-on. Maybe just talk to us about the add-on strategy, since that's a fairly new endeavor for you.
Well, the add-on strategy is really trying to cater for the power law that is out there. There's a power law that Gustav has said it very clearly in many of his external discussions that there's no average user. The problem is if you have only 1 million users and you try to sort of diversify them and find pockets of monetization in them, you will not have a good base to work with, both from the profile typically, but also the size of differences to monetize enough on those that you can get scale. But when you have 800 million and 306 million paying, if you come out with a product in to have enough base to find enough people that will be attracted to that. Then you can actually reach those power users. There's power user in books.
Some people are reading every day, and I actually only read a fiction every month. So there's a difference there in power users. I'm probably paying less for books and willing to pay less for books than that other. Typically, it's about how do you then make sure that the Premium doesn't. Where do you pick to have in Premium and what do you pick to have outside? So, in books, we have said you can have some of it inside, but if you want more, you need to have it outside. So that is also going to be dependent on how the power user structure looks like, which we can see because we have 800 million people using things. We can see how that curve looks like. Secondly, also, we'll see what the cost is for something like that.
So you can't just give something away for free if it has a cost directed to it. These direct costs sort of momentum that we have in many things we do is something we have lived with for a long time. That's why we are so used to also. We're used to actually charging for something we have to pay a license for. We also use them to figure out how do we monetize on something where we may have inference or something else that needs to be involved, and then how that should be an add-on or part of the Premium value.
Yeah. That makes a lot of sense. I want to get into the AI Remix tool shortly, but I wanted to ask you about gross margins first. So at the investor day, you talked about 35%-40% gross margins for 2030. What would you say are the two to three biggest drivers from here to kind of get that gross margin from the low to mid-30s today?
You know, first of all, the music and in general, not only music, in the future, our marketplace is a fantastic lever for margin that will, on top of our agreements, then help us to drive margin. It still has room to improve the margin. The second thing is, of course, the non-music things. Both podcasts and audiobooks are typical examples of products that are non-music that are helping us to drive margin on top of music, higher margins. Typically, our non-music products will have higher margin, and the add-ons are also built to be neutral or margin accretive to everything we do.
In addition to that, of course, we have just, as I started out saying, we have turned the ads business around, even if it's just 10% of our revenue, going from a lower margin on that to a higher margin as we are gaining into double-digit growth will also help the margin improvement.
What would be the biggest swing factors between the low and high end of that 35%-40% range?
Well, we've given ourselves room because of two things. We were discussing this in spring before we went to the investor meeting in May then. There's two things that, of course, can change. One is, we haven't given an exact number, of course. We've given a round number on revenue growth, and depending on revenue growth, it could actually do that. But also the mix in this. There's another thing is that we have given ourselves a little bit room because it's so hard to figure out how things are done, but we do a lot of things ourselves. The more we do ourselves, like a model or training, maybe that will end up in OPEX and R&D. But if you don't do that, you buy it from someone else.
We give us a little bit room to actually work a little bit between, depending on what we think is going to be best economic value on operating income in the end for our shareholders, and give us best control over time.
Can you remind us the quarterly seasonality from 4Q into 1Q for gross margins historically? I believe 1Q is usually down versus 4Q, I think primarily on the ads business, but just remind us the quarterly cadence there historically.
Yeah, quarter one is, when you take the ads business in particular, it is really, quarter four is actually the strong quarter where you have, now it is Black Week right now.
Black Friday.
Black Friday week. It's a week or a month, but Black Friday, then you have Thanksgiving, then you have Christmas, and you have all these things. They are driving, of course, enormous ads bump. That's a year-to-year thing, so we always have better margin in quarter four, as you say. In quarter one, then people spend less and ads are lower. So that is something. Then if you think about this year also, we just launched 160 countries with audiobooks. They are coming gradually into quarter four, but they will be fully in there in quarter one. Always when we launch something, it has a little bit drag in the beginning. So, that's also something we have right now that we have to live with. I've said it many times, and I will say it until I leave Spotify.
I mean, I'm not so focused on quarterly gross margin, but we know we need to educate our company. I refuse to be trying to find a pattern on a quarterly basis. We are working, and we have a good plan to find a pattern on a yearly basis of growing to the 20%+ that we promised in 2030. That's what we are focusing on. If one quarter is different from another one in the sense that maybe from the outside doesn't make sense without us explaining it, I'm not so worried about. I don't think people last year would have been thinking, "Okay, how would Spotify gross margin look like for this year?" For the year, we actually are improving our gross margin, and it's actually clear from as we've guided for the full year.
But we had three quarters that had the same pretty much gross margin this year, the first, second, and the third. No one probably would have expected that last year. If I would have said that last year, you wouldn't have believed me. You would say, "I can't get that through my spreadsheet." This is how we work. We work with a model that is more long-term, and we actually don't care so much about the quarter. Bryan has a tough job, our IR, because he comes and tell me, like, but I'll say, "We just have to think what's best for the business.
That makes a lot of sense. You mentioned the improvement in OPEX growth that you guided to in 4Q.
Yeah.
I was hoping you could maybe walk us through, you know, what are the drivers of that? I know you've talked about this year having kind of a EUR 200 million incremental investment this year. How should we think about how that kind of flows into next year? I think historically we've seen kind of high single-digit OPEX growth. This year is obviously going to be, you know, mid to high- teens. How should we think about how much of a unique year this was versus, you know, 2027 and the product release cadence?
Well, we see it a little bit as a unique year, but some things come and go. We had a good opportunity to do marketing this year on all the new features that we brought to market, including we actually spent a bit on books in general to push that. Books has been, I think it's one of my most favorite products right now internally, even if I listen more to music. Because there's so much opportunity there. To get that knowledge out and get usage is very important because if we don't get that, we don't get the engagement and the value perceived, because they can't see it, and then we can't increase pricing. It's very important to do things like this. We felt there was a good return on that marketing this year, and that's why we had to spend it.
The other thing was much maybe clearer and much more tangible in a sense, even if we know that the marketing is very tangible. But for you, it's like we entered this year with many things happening in AI. So many things happening in AI. We are transforming internally. We are putting products in place, and features in place that uses AI on the platform. At the same time, the whole sort of pricing and tokens and cloud is in a turbulent moment right now of understanding where it's going. As we are an AI company, a tech company, from that point of view, we felt that we can actually do this and invest and do it disciplined and have control of it.
We put to ourselves that we're going to spend more the first three quarters, and in quarter four, we should be able to spend less. What we have done during that time is actually to build our own product internally, Xirp, which is one for the productivity, where we actually both control how you work with agents, but also how you flow those agents to the cheapest model and work with them. At Spotify, you don't know which model is helping you soon, because we don't want you to be choosing that. Our AI is doing that for you instead. That is a product that actually even some customers have, business customers have taken and started to buy from us. That has been a productivity gain for us.
The other thing, and that has had a lot of improvement internally, but we've also gone through everything that happens in the app and how that is using AI to do something. Like, do you have to have the most expensive to do a Prompted Playlist, or can you have something else, and when do you use what? We also worked very hard on that modeling to get the cost down on that. Thirdly, we have done a lot of new sourcing, but also platform work on how we store and how often do you need to be able to get something down to the platform and where do you store it and how much. So, is it in the back warehouse, in the middle warehouse, or in the front warehouse, and things like that.
Those things are very natural for us, but it's been extra much work on that this year, and that is going to improve our cost base. Even if the cost base is going up, we have taken down sort of the relative cost on that. So that's the way to keep track of the cost as we grow, because we grow fast, you would expect it also to go up. On top of that, we have worked on our own model for training
Yeah
and different kinds of things that we want to train on
and that we may want to use for different things.
That makes sense
for consumers.
Yeah. Can you remind us which AI costs flow through which part of the P&L?
If you use something, when we do R&D, when we do an R&D work, we do a new feature, like the Remix feature,
that is worked on. That is going to be in the R&D cost.
We don't capitalize anything, for good and bad maybe, but we don't capitalize, so our income statement is very similar to our revenue. The other thing then is that when that feature is done and maybe we're going to train a model on it, we want to train a model to give you better personalization. That model will be that first training will be on R&D as well. But then when you use the model and when you use inference, for example, for Prompted Playlist, that AI cost will be in cost of goods sold.
That's how it works. And then part of the platform, the whole storage and everything, we have like a model where the majority, of course, ends up in cost of goods sold because that's used by us when we use Spotify, while we also use that platform internal R&D building, and that part of the platform will then be allocated to R&D. So that's how we-
Very helpful. And one of the products I'm excited about is Reserved by Spotify to get concert tickets.
Yeah. Have you gotten any?
I have not. I was going to say, if there's anything you could do to help me out there.
Unfortunately, Maggie Rose is also on our list of tickets, because otherwise probably I would have gotten one.
You would have been top of the list.
Yeah.
Top 2% of listeners. Live Nation had their investor day last week, and they talked about how excited they were about the partnership, and it sounds like you are expanding that feature beyond the U.S. and into Europe and Australia next year. Maybe you could discuss some of the early learnings from this unique and differentiated feature.
I mean, Alex and Gustav were so excited when we were at the investor day when we launched this, and we Reserved by Spotify the tickets for you. We have some famous people that have been on that. I am not a good person to remember names, but Miley Cyrus and Metallica.
Metallica, yeah. I was hoping I was going to get the Metallica.
We launched it in the U.S. and then in Canada, and we are going to continue to roll it out, as you say. There has been hundreds of thousands of Reserved by Spotify tickets out there. They were so excited, and I was as maybe as a CFO-ish, I was a little bit like, "Can it be that exciting?" This is one of the most positive things when we get feedback from our consumers that we have launched in a long time. I just think it is the feeling of that you do not have to worry, but you also do not have to pay extra, because tickets are a little bit weird thing out there, because if you want to get tickets, I mean, you have probably done like me, like calling someone, "Do you know where to get tickets?
Have you bought You know someone that can sell to you?" it becomes this weird thing. you're still a fan and want to So it brings the fandom, the simplicity of not being robbed together with a artist that also loves to find a way to get to fans in a good way. So we find the fans for them also. it's just a lovely product, it's been so much appreciated both by our users, but also by the artists. So, I was a little bit too negative there, Alex and Gustav was exactly what it should be. This is top product, actually.
That's great.
it's going to continue rolling out, this is just the beginning of that.
That makes a lot of sense. I-
We do sort of handle sales of tickets also, if you've seen that on Spotify. There we take a commission, but Live Nation sells the ticket through us. It's also a way to help the consumer or the user to actually find tickets when they sit and listen and see where it is.
That makes a lot of sense. I wanted to ask, what other new product enhancements are you most excited about? Obviously, fitness and custom podcasts are two that come up often.
Yeah.
Maybe elaborate on those, or if there's anything else we should be thinking about.
No, there's a lot of features coming out, and we're all different. I'm actually, as a CFO, I'm most excited that it's going to be just so many broad ones, so it takes the diversification and the risk actually profile becomes very good. It gives you a little bit more time also to be patient on getting things to work out. You're not depending on one smash hit, like maybe some other subscription companies are. So we're more like finding a way to meet these 800 users out there and picking up, as I said, with 1 million there, 1 million there, and growing slowly on this. So that's what I'm actually most excited about. I have a colleague that is most excited by DJ. I'm not myself, but I may change over time. Who knows? But it takes a lot of time. That's a hobby.
But there's many people out there that have that as a hobby. Gustav thinks it's a smash hit, and maybe he's more right again than I am on that thing.
That's great. I did want to turn to the AI Remix opportunity. We had UMG here earlier, and they were very excited about the potential for this product. Obviously, Warner Music had their results earlier, and they were talking positively about kind of the product and customer interest. I'd say, what about consumer behavior that you know gives you confidence that people are going to be excited for something like this?
Well, I mean, I just think that we have always loved to do remixes. We haven't tried this product on any of our users yet, but remixes and covers have been tested for a long, long time. There's even TV programs where artists come together and play each other's songs. So, there's been a long history of doing variants of music that have engaged people. So that's why we think this is something cool. The other thing we like with this is that it actually goes back then to the artist. So, if I do a cover on you, I'll be another person. But this is someone doing a cover on you, and you keep the rights. This is actually something that I think is also interesting. But the engagement, I think, will just come from it's a very cool product at first sight.
That makes sense. I wanted to ask a bit about the economics and launch timing. Obviously, press has speculated maybe in 1Q. What still needs to come together for the remix tool to actually come to market? Is it catalog breadth? Do we need other labels like WMG to come on? Then how do you, as the CFO, think about the economics of this product relative to the existing?
The first thing to think about is that we actually have a role in society and among users because we came in to try to protect the IP and take away piracy. I think that's just, it's actually worth thinking about. What we're trying to do is to One of the strengths of Spotify is patience. Maybe some thought it was too long to have 17 years without profit before we were profitable, but we were patient to make sure we had the best product and the best sort of global reach. Here, it is actually important to remember that the consent and the compensation goes to the right person, and sort of the consent from the artist and then the compensation goes back to the right, and they get the credit.
Therefore, we're trying to find a model that actually, A, can drive revenue for the labels and artists and us, but also is licensed in the proper way that works over time. We're not just rushing to get something out. We're rushing to get something out that can hold for many, many years. We're not so worried in the end by other things coming out because we have 800 million users. We have the biggest revenue platform and the biggest streaming, and that is a good foundation. Like in podcasts, you probably don't want to be on YouTube only. You probably don't want to be on Spotify only. You want to be in a place where you have enough users to bring you an audience that can pay for your ads and so on. But in this case, it's also the same.
We want to find a good platform for this, but they also need to then chip in. We have signed a lot of agreements with UMG and Believe and Kobalt and so on. But it's also an artist opt-in. We have a product that we think is ready, but we also need that consent from the artist. That's actually something the labels should answer and not us.
Yeah.
We think it's going to be something that will be launched.
We think we have a model that is good for everyone. The reality, it's from an economics point of view that the listening of these remixes will be as any song listening to. They're to come from the pool.
Yeah.
Then you will pay for driving some of this as a user because you find a good engagement in it and find that valuable for you, therefore, you will help the artist to get more covers out there, you put it that way. We will also make sure we monetize on that together in a way that actually keeps the margin for us and for them.
Got it. It shouldn't be negative to margin.
No.
Then we did have UMG, as you mentioned, they said they're working hard at attacking, trying to get artists on board and opting in, and they said they have a list of thousands of artists that they're working towards. It feels like there's progress being made on that front.
Yeah. I think it's also good to remember that we have a lot of agreements now, but as you say, you need artists to opt in. You want to have enough artists opt in because you want to have enough TAM, because if you go out only with Maggie Rose, which I talked about before, maybe I'm the only one that will do the remix and no one else. That's not a good product to come out with. But if you go out with thousands of very popular artists, it's going to be much easier to reach a good TAM to start with. You want to have a good platform when you start this. We see that they are working, but it's a new thing for artists also. It's a long-term thing.
It's not a thing that, "Oh, should we do this next year?" It's a long-term change in how an artist thinks, and artists think differently from finance people, I can tell you.
That makes sense. I did want to turn to audiobooks going global. You mentioned 22 markets going to 180. We're basically going fully global with the audiobooks. You mentioned before kind of six-plus months from a price hike, but how should we think about how big of an opportunity that is? How much of the Premium revenue base already has access to audiobooks, and how incremental are these additional markets?
I mean, from the revenue base and where we already launched, I mean, we've already sort of passed the half of the revenue base in a sense, around that number. But we are moving up, but we are moving up also in additional revenue. You have to remember, there's one part is the pricing here, but many of these countries, they have maybe so low ARPU in music, so you need to have it as an Audiobooks+ or an ร la carte sale. Therefore, there's an additional revenue coming from that. That is a very good margin business in general as well. It's accretive to our margin. We just see also, this is coming back. The audiobooks is a so good example of how we start with a bet.
It's a bet two and a half years ago, one bet that maybe people didn't think we would handle. We do bets that doesn't work, and we do bets that works well. The thing is that we have control of these journeys all the time. This is back to the discipline. We're working very tight in our leadership team on all the financial metrics together with the operational metrics. Probably the company I've been in where that is most visible in the leadership team on the top. It helps because we have one platform we all work on together. In that, in the books side, you have a journey of seeing the margin improve, the value improve, and we can bring pricing on board, but also then the add-ons, and you can also see the engagement.
Engagement is very important part of the whole pricing and the journey for us, because there's three parts of engagement that we have found drive the retention and the value creation that we can monetize on. It is really, first of all, how many days you use and how many hours you use on Spotify. Those who use books and music and use all the products, they typically, they are 28 days or more today. They pretty much open Spotify every day. That is such a good thing from a retention and also a monetization potential point of view. The other thing that is very important for, and now I'm stepping outside a little engagement, the other thing that we also is very important for us is, of course, how many devices you use.
If you use your Spotify in your car and on your Sonos, but also at the country house, another kind of system for speakers, but also use it with your AirPods and you use it on United Airlines maybe, the screens where we install. You use Spotify in more devices, and we are AI in ChatGPT. The ubiquity strategy for us is very important. We have found that the more platforms you use Spotify on, the more engaged and the more retention we have. Finally, the third one is actually the simplest one that we talked about, is that you use books, music, podcasts. That also increases the value of our customers. That's also, if you bring in more add-ons and different features like the fitness and other things, we will bring that engagement platform up.
That's where we start all the time, and then we figure out how pricing can happen afterwards, because we are so super cool that it can happen as long as we fix this. The worst thing is coming to lose that value to price. Then you're sort of stuck in a negative spiral where you've been going too fast. That's a very deliberate strategy from our side. That was a long answer to-
No, it is a great answer. It reminds me of the no regrets that you guys talk about all the time.
Yeah.
Maybe just elaborate on that a bit. You do a bunch of survey work, and your users are so passionate and so engaged. Maybe just flesh that out.
Yeah, we tried. Something that came up actually as late as two years ago, we started to think, what do people, when they leave Spotify, what do they think about being on Spotify compared to a TikTok or a Facebook or something else, or Netflix? And we found that we were actually best in class when it comes to when you leave Spotify and get the simple question like, do you regret that you spent time, and do you regret how much time you spent on this platform? We have a very low number that is negative, while some of the other platforms have a very high number. That is something we have brought with us in our strategy. We do not want to dilute the experience with less well-spent time content or features. So, that is, we found, a very good way of thinking strategically.
Fitness is something that typically goes very hand-in-hand with music, and it goes also very hand-in-hand with that you feel it was time well spent. So, that is a good example of that.
That makes a lot of sense. Just turning back to Audiobooks briefly. So across different markets, how will the offering differ between it being included in Premium versus add-ons versus individual purchases? Then how should we think about kind of the returns and the economics across those?
So, actually, when you see how some countries will have only ร la carte or Audiobooks+, and some others will have it included in Premium, it will be because we are very sensitive to our margin and realize that if you do it in certain countries, it could be a risk on the margin. Because if the ARPU is very low, but the licenses for books are much higher, this is a product that comes with more expensive, maybe hourly costs, then you need to build that into the model. So you will see low ARPU, as I said before, low ARPU countries having actually that more pay-as-you-go model compared to countries where it's possible to bring it into the pricing of the Premium product.
That makes sense. Then I wanted to turn to advertising. We referenced earlier you're seeing some green shoots there. Last quarter you accelerated to 9% constant currency. What still needs to happen for advertising to become a more meaningful contributor to your mid-teens growth algorithm?
Well, we started a couple of things, and we need to continue to do some of them. I mean, we have Last year, we expanded our funnel by actually bringing in a much better free product. So that's more available now, so we need to work more on that. We moved more of the direct into biddable and increased the biddable channels value. We also found that we had to train and educate our customers on that. Spotify came When you come from a weak position, you still have some kind of training and work to do with your customers to understand why Spotify is a good place to be on. It's not that we don't have a good situation in the sense that we're getting to now the double- digits, but we have higher ambitions than that.
With video and other products coming in, I think it is interesting. In podcast, it is very clear that we have been able to work with sponsorships this year, and that has helped. There is a lot of things we can do more there. On sponsorships and podcast, we think that the margin also can go from the 20% today up to closer to 40% over the coming years. There is room for opportunity. We just changed leadership because we had this journey of fixing things, and we did it exactly like we wanted to. Maybe it could have been done faster, who knows? But we did it, and that was the most important thing. Now we need to have a stable growth, double- digits, and grow from there. That is the next journey for our new leader to take on.
Maybe just, you referenced going to biddable, maybe just talk about programmatic and how important that is in kind of getting the whole flywheel spinning.
Yeah, I mean, the growth is good. The number of customers are adding, and the spend is also going up. This is where we actually need to just continue to drive that same flywheel and continue to educate. It is a little bit tougher to come from an audio standpoint than from a video standpoint, and that is why video has been also important to get into. It is not only the main reason is that people want it, and it creates engagement. Even people use it when they do not look at it for some reason. But it needs to be there, and it drives engagement. But that will also help advertising because you get more paid on videos than you get on audio.
That makes sense. Great. I wanted to ask about capital return and the balance sheets. You have bought back over $1 billion of stock year- to- date. You are basically neutralizing the equity employee programs. But free cash flow is really starting to ramp, over EUR 3 billion in the trailing 12 months. So I wanted to hear how you are thinking about your willingness and ability to maybe get more aggressive on the buyback.
Yeah. I want to be a little bit careful. A little bit like the gross margin discussion. I am going to repeat myself. There is a couple of things. We are still that I have been two years in the company, but we just got profitable three years ago or something, and cash flow positive. We are still a young company in that journey compared to the others out there. If you look at Facebook, I think it took seven years before they started doing something at all. We shouldn't compare with them. We are a young company. We believe it is an important strength to have a strong balance sheet. That is the starting point for us.
But of course, we have also said clearly with the cash flow we have and the balance sheet we have, it is not going to be possible to not return more cash to the shareholders. We stated that in May, and we started to do more buybacks in quarter four. That said, we do have the priority of organic growth and inorganic growth. We want to be a growth company. We work very hard to continue to be a growth company. If we find good investments there, that is where the balance sheet should be used for first. If that is not something we can do or we have done that already, then we will look to shareholder returns. As I said, we came back to that now in quarter three with a little bit more buyback than we had done before.
Great. Closing out, as we think about the 2030 plan, when you look across it, which areas do you think require the most focus on execution, and where do you think we could see surprises to the upside? Then maybe two or three milestones that investors should be watching over the next 12 month- 18 months to see how we are tracking towards those 2030 targets?
I will start backwards. I mean, what we should track on, I mean, if you go for the small numbers, it is still small numbers. We have said ads should be double- digit. We should continue to have that in our targets and deliver. More importantly, the long-term goals we have given to the market. As you said, we are steadily going to move into our 20% operating margin over the next four years, and that is something we should do. Don't be concerned about the quarters, but look at the years and see that we are moving on as we should. We have also been very clear that we have a good opportunity on books and on add-ons in general. You should be able to see more add-ons and that the book business continues to grow.
That I think is good things that we have talked about that should obviously come out. I am very focused on, it sounds like it is nothing to be sort of concerned about, but it is like we focus on certain things. We focus on our financial model, and we focus on the and I focus a lot on the model around AI cost and the structure of our cost base. So much things are happening in the world, and what we knew two years ago is not what we have today. We work with a lot of scenarios. We work very tightly, Alex and Gustav and myself with the E-team on how this can model out and what it means for our choices and what we need to think about. We proactively think about that.
Either we go for cloud cost or for inference cost or for license cost or for consumer products, packaging, et cetera. We really think about that in a good way. That is really a strong focus while the rest of the companies are building everything and producing everything to keep track of that. This is back to that we want to deliver on those goals, really, and we feel that we have a good opportunity to do that. That is really why we do that.
That is a perfect place to end. Christian, thank you so much for joining.