The 300 Millionth Subscriber, a Record Margin, and a User Guide Spotify Cut On Purpose: Maintaining Outperform

Results vs. Consensus

Spotify reports under IFRS in euros. All figures below trace to the interim condensed consolidated statement of operations furnished with the results; segment comparatives use the recast series published after the January 2026 reclassification of certain revenue-generating activities from Ad-Supported into Premium.

Versus company guidance

Versus Street consensus

Year-over-Year Comparison (Q2 2026 vs. Q2 2025, € millions)

Sequential Comparison (Q2 2026 vs. Q1 2026, € millions)

Revenue. Reported growth of 13.9% and constant-currency growth of 15%, an acceleration from 14% in Q1 and 13% in Q4. Currency cost about 70bp against the roughly 80bp assumed in the guide, so the drag that took 600bp off Q1 is essentially gone. The Q3 guide goes further and assumes a roughly 200bp FX tailwind, the first in the period we have covered. From here, reported growth exceeds constant-currency growth, which reverses the optical headwind that has made this business look slower than it is for four consecutive quarters.

Margins. Gross margin of 33.4% is an all-time record, 193bp above the prior year and 30bp above guidance. Both segments contributed: Premium expanded 174bp to 34.9% and Ad-Supported expanded 179bp to 19.1%. The CFO attributed the beat versus guidance primarily to quarterly timing shifts in growth investments plus a small one-time benefit from the cancellation of the Digital Service Tax in Canada, where a prior-year accrual was reversed. That last item is genuinely one-time and was disclosed unprompted, which is the right behaviour.

Operating expense. This is where the quarter looks worst and where the disclosure improved most. Operating expense of €941M rose 20.6% sequentially and 3.0% year over year, and operating margin fell 206bp sequentially to 13.7%. The spend management pre-announced in April arrived on schedule. What is new is that it now has a size and an end date attached.

Below the line. Net finance income of €65M is a fraction of the €222M in Q1 and the €320M in Q4, because the Exchangeable Notes were settled in cash in March and the fair-value line is gone. The tax rate was 24.3%. Net income of €545M and diluted EPS of €2.61 therefore mean something for the first time in a year, and the year-over-year comparison against a €86M loss is not informative because the prior-year quarter carried €447M of finance costs from the same notes.

Segment Performance

Segment gross profit is derived from the disclosed segment gross margins and foots to the reported consolidated gross profit of €1,596M. Comparatives use the recast series.

Premium — 300 million subscribers, and ARPU doing the heavy lifting

Premium added 7 million net subscribers against a 6 million guide to cross 300 million for the first time, up 8.7% year over year with growth in every region. Revenue of €4,331M grew 15.4% reported and 16% in constant currency against 8.7% subscriber growth, so more than a third of the growth came from price rather than volume. ARPU of €4.89 rose 7% reported and 7.4% in constant currency, inside the 7% to 7.5% management guided in April and up from 5.7% in Q1.

Premium gross margin of 34.9% expanded 174bp. The four-quarter arc of this line is the most important single trend in our coverage of this name: negative 32bp year over year in Q3 2025, negative 8bp in Q4, positive 129bp in Q1, positive 174bp now, all on the recast basis. The segment that made us initiate at Hold has become the reason to own the stock.

The add-on strategy also received its first hard number. Audiobooks+ has passed $100 million of annual recurring revenue, and management said separately that the figure has doubled since it was last discussed internally, from a handful of markets only.

"Audiobooks prove it first in just a handful of initial markets. We see that overall Audiobooks penetration among premium listeners has more than doubled this year. And Audiobooks+, the add-on, has passed $100 million in annual recurring revenue. It's subscriptions on top of subscriptions with more on the way." — Alex Norström, Co-Chief Executive Officer

Assessment: At initiation we wrote that the add-on model was the most credible answer to flat ARPU because it does not require a headline price increase and does not require the labels' permission, and that it was a directional signal rather than a modelable one because no figure was attached. A figure is now attached. On roughly €17.5B of trailing revenue, $100M of ARR is small; as proof that subscriptions can stack on subscriptions, it is the most important disclosure of the quarter.

Ad-Supported — positive year over year for the first time

Ad-supported revenue of €446M grew 1.4% year over year on recast figures and 3% in constant currency, and gross margin expanded 179bp to 19.1%. This is the first quarter in our coverage in which the reported line grew, and it comes with the first genuinely useful set of operating metrics the company has given on this segment: automated channels reached nearly 40% of ad-supported revenue from just over 30% in Q1, active advertisers reached 33,000 and grew 60% year over year, and the migration of ad inventory to an in-house ad server is complete.

Management also closed off the last piece of the transition narrative and sharpened the target.

"Importantly, our price optimization work is now complete, and this channel is stabilizing. We have also completed the migration of our ad inventory to our in-house ad server, allowing us to further streamline our sales process and capture more demand. With respect to our outlook, nothing has changed. We continue to expect our ads business to inflect towards double-digit growth in the second half of 2026." — Christian Luiga, CFO

The margin ambition was also quantified for the first time, referencing the investor day: the segment moving from the 20% range today toward 40% over time, as automated and self-serve buying turns it into a scale business.

Assessment: We moved this pillar to contained in February, reverted it to emerging in April after the constant-currency deceleration and the 101bp margin reversal, and it now goes back to contained on better evidence than either prior time. The difference is that this quarter carries hard numbers rather than assurances: 40% automated, 33,000 advertisers, plus 179bp of margin and a positive reported line. The H2 2026 target has now been held for four consecutive quarters and has been upgraded from "the growth that we want" to "double-digit." That is a commitment we can grade in October.

Key Operating Metrics

Ad-supported MAU of 494 million against €446M of quarterly ad revenue is €0.90 per free user per quarter, against €1.00 a year ago in Q3 2025. Monetization per free user is still falling even as the segment turns. That is the arithmetic behind the emerging-markets strategy discussed below: Spotify is adding free users faster than it can monetize them, and has decided to stop.

Key Topics & Management Commentary

Overall Management Tone: The most confident and the most specific of the four calls, and notably the first structured around a published framework rather than a quarter. Management opened by grading itself against the May investor day, shipped a number against the AI spend that was conspicuously missing in April, disclosed the first revenue figure for an add-on product, and volunteered that a portion of the gross margin beat was a one-time tax reversal. The one place the posture outran the evidence was the scale argument, where a Munger citation and a five-year subscriber-addition statistic were used to make a case the numbers make adequately on their own.

1. The MAU Guide Is a Decision, Not a Warning

The single number the market reacted to was the Q3 MAU guide of 788 million, roughly 5.5 million below consensus. It is the second consecutive quarter in which the user line disappointed, and this time the quarter itself also came in 1 million short of the company's own guidance. The explanation offered was not demand weakness.

"To capitalize on this opportunity, we are adjusting elements like product optimization and ad load, among other things in select emerging markets. Now this strategy carefully increases friction in our free service with a goal of driving higher user conversion and revenue growth down the line. Yes, this will show itself in our Q3 MAU, but we believe it's well worth it." — Alex Norström, Co-Chief Executive Officer

Pressed on it in the Q&A, management enumerated the specific changes: tweaking sign-up to raise MAU quality, deprecating support for lower-end Android devices, and introducing friction in ad load and free-tier limitations. The framing was explicit about the trade.

"Gustav and I have a saying where we say sometimes we pull the growth lever and sometimes we pull the monetization lever. Now here, we're starting to pull the monetization lever." — Alex Norström, Co-Chief Executive Officer

The critical claim attached to it is that subscriber growth is unaffected in the near term. That is checkable: the Q3 subscriber guide of 305 million is essentially in line with consensus even as the MAU guide sits 5.5 million below it.

Assessment: This is the right decision and it was communicated poorly. At €0.90 of quarterly revenue per free user, marginal emerging-market MAU has been close to economically worthless, and a company that has spent four quarters being asked why engagement is not converting into revenue should be applauded for choosing conversion over the vanity metric. But management let a headline number miss consensus and then explained it, rather than reframing the metric first. The market marks what it sees.

2. The Investor Day Delivered the Framework We Said Was Missing

In April we wrote that management had pre-announced a two-quarter margin cost three weeks before an investor day at which it intended to explain the strategy, and that the May event would have to justify a spending decision the market had already marked down. It produced explicit long-range targets, restated on this call.

"That strength is why we set the 2030 targets we did in May, a mid-teens revenue CAGR, gross margin of 35% to 40%, operating margin above 20% and strong growth in free cash flow." — Alex Norström, Co-Chief Executive Officer

The base against which those targets are set was also given: revenue compounding at 18% a year since the 2022 investor day to €17 billion in 2025, gross margin from 25% to above 33%, and €2.9 billion of 2025 free cash flow.

Assessment: This changes how the stock should be valued. A mid-teens revenue CAGR from 2025's €17.2B compounds to roughly €34.6B by 2030, and above 20% operating margin on that base is more than €6.9B of operating income. Against today's enterprise value of roughly €76.8B, that is about 11x the 2030 operating income the company's own targets imply. Targets are not results and this management has a habit of spending upside. But the framework now exists, which is precisely what we said the shares needed.

3. The AI Spend Gets a Number, an End Date, and a Cost-Control Story

The largest single criticism in our April note was that management would not size the incremental spend or define success for it. Both were addressed.

"We continue to expect these marketing and AI-related investments to drive approximately EUR 200 million in incremental operating expense for the full year. Quarter 4 is well positioned to see a moderation in the rate of year-over-year operating expense growth as we move past the heavy concentration of this year's marketing activities and calibrate on the other costs." — Christian Luiga, CFO

He added the structural point that matters for the terminal margin: headcount is flat for the year, so the investment cycle is not adding permanent payroll, and both marketing and compute are variable and controllable. Q3 operating expense growth excluding currency and social charges is expected to be roughly consistent with Q2, with moderation from Q4.

The cost-control mechanism was also disclosed, and it is more than a slogan. Alongside the Honk coding agent described in prior quarters, Spotify has built an underlying routing engine called Chirp.

"Chirp lets our engineers switch models mid-task and route every job to the best available price performance, including open source models that we host ourselves so that we are never locked in. It also shares the context across not just different models and different developers, but across the company. So we don't pay for the same reasoning twice, and we don't lose our own data. And it also shows us our inference spend down to the individual developers." — Gustav Söderström, Co-Chief Executive Officer

The framing management put around it is the one that matters for a market worried about AI capex: "being an AI beneficiary means winning on the cost side, too."

Assessment: A sized, dated, variable-cost investment programme with a named cost-control layer is a fundamentally different disclosure from the one we criticised in April. Two specifics carry weight. First, these are operating costs, not capitalized infrastructure, so the reversal is immediate if returns disappoint. Second, per-developer inference visibility means the company can actually enforce the discipline it claims. We remain unwilling to model a productivity dividend, and we no longer regard the spend as unquantified.

4. The Large Taste Model Shows Its First Measured Results

The proprietary personalization model that management described as the reason for the in-house training spend produced its first disclosed operating evidence this quarter, on the metric management has repeatedly said is hardest to move.

"In the first month, since we deployed our new Autoplay recommendation system powered by the Taste Model, active days have increased which is very hard to do at our already high engagement levels. Autoplay minutes and track saves have both grown significantly. And Autoplay drop-off has declined." — Gustav Söderström, Co-Chief Executive Officer

The model learns from what management described as 3.4 trillion events generated on the platform each day, and is built by continuing to pre-train an open-source base model on Spotify's proprietary data. Elsewhere, SongDNA has been used by more than 100 million subscribers, described as one of the fastest-adopted features the company has shipped, and DJ is used by roughly a quarter of active users.

Assessment: Active days is the leading indicator management has built its entire retention framework on, and it is the first time an AI investment at this company has been tied to a movement in it. One month is not a trend, and "significantly" is not a number. But the chain from spend to model to engagement metric is now visible end to end, which it was not in April, and that is the chain the bear case says does not exist.

5. Advertising Turns, With Numbers Attached

Beyond the segment result, the operating disclosure on advertising was the richest of the coverage period. Automated channels reached nearly 40% of ad-supported revenue from just over 30% in Q1; roughly 99% of impressions now serve on Spotify's own stack; active advertisers reached 33,000, up 60% year over year; and personalized ad load has launched as a new supply lever.

The structural argument for why the rebuild raises the ceiling rather than merely modernizing the plumbing was also made properly for the first time: the legacy model of fixed, guaranteed campaigns bought by phone and email capped both pricing and sell-through, because once inventory is sold it is sold. A biddable exchange lets advertisers who value the inventory bid the price up.

One incidental disclosure is worth flagging because it will not be in anyone's model: Spotify has launched plug-ins and Model Context Protocol connections to third-party AI assistants, and roughly 7,000 of its 33,000 active advertisers are using its AI audio-asset creation tool.

Assessment: The demand-side disclosure is what changes our view. Advertiser count growing 60% to 33,000 is a scale metric that cannot be manufactured by mix shift, and it is the first evidence that the new stack is expanding the addressable buyer base rather than just re-plumbing existing spend. Combined with the segment margin ambition of 20% toward 40% over time, advertising has moved from the thing that could break the thesis to a genuine option we are not paying much for.

6. Reserved: Ticketing Arrives, and Works

Reserved, the artist-fan ticketing product launched in the US in June with Live Nation, produced early results specific enough to be checkable.

"Since launching in the U.S. in June with Live Nation, Reserved has supported multiple tours with nearly 100,000 tickets reserved through Spotify. For some tours, we sold through 100% of our allocations. And Live Nation upsized them mid-run." — Alex Norström, Co-Chief Executive Officer

The strategic logic is the retention framework rather than a ticketing take rate: giving the most engaged fans first access makes the subscription more valuable, which raises retention, which raises lifetime value.

Assessment: A counterparty upsizing allocations mid-run is the strongest possible early signal, because it is Live Nation's money rather than Spotify's assertion. The revenue is immaterial in 2026 and the option is not: Spotify has 300 million subscribers with known artist affinities, which is the single most valuable targeting asset in live music. We would rather see this scaled than see another content vertical opened.

7. Music Remix: Universal in May, Merlin Today, and a Reinforcement-Learning Argument

Spotify announced a deal with Merlin, the digital licensing agency representing roughly 30,000 independent labels, following the Universal agreement in May. The framework is what management calls the three Cs: consent, credit and compensation, with artists opting in to make catalogue available for covers and remixes.

The differentiating argument was about post-training rather than licensing, and it is the sharpest thing management has said about why scale matters in generative music.

"So what we will do is we will allow people who are fans of a certain artist to start making remixes with songs from that artist. And what they do is they say, I think this remix was better than this remix. And then we get the preference data that actually makes the model better. And this is our unique advantage in this business. We have now 777 million people and music fans to do reinforcement learning with, which is why we think we're very well positioned in this business." — Gustav Söderström, Co-Chief Executive Officer

Management also drew the distinction it thinks the market misses: the product is about real artists with real voices, opted in, rather than synthetic ones, which is why rights holders are engaging. A research preview will go to some users without requiring the full catalogue.

Assessment: The preference-data argument is the first version of the AI moat case that is specific to generative music rather than to recommendation, and it is a good one: a remix model improves with human preference at scale, and Spotify has more music fans than anyone. Management was appropriately deflating about the timeline, saying explicitly that it will take time to scale into a material revenue driver. We assign no revenue to it and note that it converts the AI narrative from a threat into an owned asset, which is worth something to the multiple even before it is worth something to the P&L.

8. Capital Allocation Shifts Language for the First Time

Spotify has repurchased $662M of stock year to date through August 3, a 30% increase over 2025 levels, and has bought back nearly 2.2 million shares since resuming repurchases in 2025, roughly 1% of shares outstanding. Cash stands at €9.4B with no debt other than lease liabilities. For four quarters the buyback has been described purely as dilution offset. That changed.

"Given the stronger cash flow in the years ahead, even with M&A, we expect that we will also return cash to shareholders." — Christian Luiga, CFO

Assessment: Two things in one sentence. The first is the clearest signal yet that capital return will eventually exceed dilution offset, which we have flagged as the weakest part of the capital-allocation story since initiation. The second is the first mention of M&A from this management team in our coverage, delivered as an aside. Neither was picked up in the Q&A. We would treat an acquisition as a change in the risk profile of a company whose entire margin story rests on organic operating leverage, and we will ask about it next quarter.

9. Open Source Is Deflating the Cost Curve

Asked how Spotify weighs open-source deployment against frontier commercial models, management gave the most useful cost commentary of any call in the period, and disclosed that the internal routing layer may become a product.

"On a constant per feature quality level, it's pretty clear that the costs are coming down for a certain level of quality. That doesn't mean that we won't use more advanced models, but for a certain feature level of quality, the costs are coming down quite fast." — Gustav Söderström, Co-Chief Executive Officer

The Large Taste Model is itself built on an open-source base through continuous pre-training on proprietary data. Management also said Chirp will be offered to other companies, which is a small but genuine strategic tell.

Assessment: The bear framing of AI for a company like Spotify is that inference is a permanent new cost of goods. Management's counter is that unit costs at a given quality level are falling fast, that it self-hosts open-source models to avoid lock-in, and that credit-based add-ons cap inference per purchase by design. That last mechanism is the important one: usage, cost and monetization are structurally tied together, which is precisely what a consumer AI product needs and what most do not have.

10. Pricing: Category Leadership Restated After a Competitor Moves

Asked about the read-through from a competitor's July price increase, management declined to comment on it directly and instead restated its own position.

"But what it does show is that this is an example of just continued value in music streaming services, and we're happy about that for the ecosystem… And of course, the more value we deliver, the more pricing power we will have. And as of right now, we're happy to be the price leader in our category." — Alex Norström, Co-Chief Executive Officer

Assessment: "Happy to be the price leader" is new language and it is not idle. Spotify raised US pricing in January and its principal competitor followed in July, which is the textbook signature of a market with pricing discipline rather than price competition. For a business whose gross margin is set by content costs it does not control, an industry that raises prices together is worth more than any single increase.

Guidance & Outlook

The Q3 guide is the first in the coverage period that beats consensus on revenue, and it does so because the currency tailwind has arrived. A €5.0B guide against €4,272M a year ago is 17.0% reported growth, assisted by roughly 200bp of FX, against constant-currency growth management put at 14%. For four quarters this company has printed reported growth well below its underlying rate; from Q3 the optics flip.

The gross margin guide of 32.9% steps down 50bp sequentially, and management gave the reason rather than leaving it to be inferred: the annual charge Spotify takes in the third quarter for regulatory-fee exposure in one market, alongside continued reinvestment. Against the prior-year quarter it is still roughly 130bp higher.

The operating expense question, answered properly. Asked directly about the implied acceleration in Q3 operating expense, the CFO gave the clearest bridge management has offered.

"Excluding the currency and social charges we see on our expenses, our expense growth in quarter 3, we expect it to be roughly consistent with quarter 2. So it's going to be pretty much the same growth… So we're well positioned to actually moderate this into quarter 4 and to keep it consistent into quarter 3." — Christian Luiga, CFO

Implied full-year 2026. First-half operating income is €1,370M actual, and the Q3 guide adds €670M for €2,040M through nine months. With Q4 moderating on operating expense growth and carrying the seasonal revenue peak, we model full-year 2026 revenue of €19.5B to €19.9B and operating income of €2.8B to €2.9B, narrowed from the €2.8B to €3.0B we carried in April. Full-year gross margin and operating margin improvement were reiterated for a fourth consecutive quarter.

Guidance style. Spotify has now beaten or met its own guided revenue, gross margin and operating income in every quarter of our coverage, and has guided the forward operating line marginally below consensus three times running. The pattern is consistent enough to be a policy: set a number you will clear, and let the Street's own model be the thing that disappoints. It costs the stock something every quarter and it costs credibility nothing.

Analyst Q&A Highlights

Spotify runs its Q&A through a moderated queue in which submitted analyst questions are read aloud and voted on by participants, so the exchanges below are reproduced as they were put to management and answered on the call.

What Is Behind the Slower Third-Quarter User Guide

The question that most needed answering asked directly whether the Q3 MAU guide reflected a change in the competitive environment or in campaign intake. The answer was neither, and it laid out a deliberate strategy shift in emerging markets with an explicit claim about what it does not affect.

Q: "What's informing your guidance for slower MAU growth in the third quarter? Do you see a change in the competitive environment or the general intake for your campaigns?"
— Batya Levi, UBS

A: "We've made changes like tweaking the sign-up to get higher-quality MAU throughput. We've deprecated a lower-end Android device support, which builds the business and makes it more efficient for us. We've carefully introduced some friction in both ad load and some limitations in our free tier. All of these things are positioning us for more monetization… And I think also a consequential point here is that this will not affect subs growth in the near term."
— Alex Norström, Co-Chief Executive Officer

Assessment: This is the exchange that should have set the tone of the session and did not, because the guide had already been marked. The specificity is what makes it credible: sign-up quality, device deprecation and ad-load friction are three distinct, verifiable actions, not a euphemism for weak demand. The claim to hold management to is the last sentence. If Q3 subscribers come in at or above 305 million while MAU lands near 788 million, this was a good trade and the market mispriced it.

Why Advertising Growth Has Stayed Subdued, and What Changes It

A question noted that ad growth has been subdued for a year despite programmatic scaling, and asked for a multi-year outlook and for what can drive momentum in a high-margin business sitting on a growing engaged base. The answer produced the demand-side numbers the segment had never previously been given.

Q: "Advertising growth in the past year has been subdued despite programmatic now comprising over 25% of ad-supported revenue. Can you provide an outlook for the coming year or years? And what can you do to drive momentum in what is a very high-margin business that monetizes your engaged and growing base of users?"
— Jessica Reif Ehrlich, Bank of America

A: "The bottom line is really that this has led to not just existing advertisers moving over to this new way of buying, but it's also that we have increased the active advertisers. We are now at 33,000 active advertisers. This is 60% up from last year, which is great… when the inventory is bought, it's bought. And it's also capping us, obviously, in the ways that people want to buy… now people that find our ad inventory valuable can go in and bid up the pricing."
— Alex Norström, Co-Chief Executive Officer

Assessment: The pricing-cap argument is the piece that had been missing from six quarters of ad commentary. A guaranteed-campaign model cannot capture upside demand because the price is set before the auction that never happens; a biddable exchange can. That is a structural reason to expect a higher revenue ceiling, not just a faster sales process, and it is the first time management has explained the rebuild in terms of yield rather than convenience.

The Drivers Behind the Implied Third-Quarter Expense Acceleration

A question asked management to explain the implied acceleration in operating expense embedded in the Q3 guide. The answer reframed the guide as arithmetic rounding rather than acceleration, and restated the full-year size and the Q4 moderation.

Q: "Can you talk about the drivers of the implied acceleration in operating expenses in Q3?"
— Doug Anmuth, JPMorgan

A: "We have said we're going to elevate it with EUR 200 million for this year… excluding the currency and social charges we see on our expenses, our expense growth in quarter 3, we expect it to be roughly consistent with quarter 2… I think it's important to be careful with the rounding nature of our guidance. We talk about EUR 5 billion. We also add, I mean, growth, 14% growth and ARPU consistency between the quarters… And this is a structural — nonstructural cost change, personnel at in the same level."
— Christian Luiga, CFO

Assessment: The rounding point is worth taking seriously and is easy to miss. A guide of "approximately €5.0 billion" against a stated 14% growth rate spans a range wide enough that the implied operating expense figure depends heavily on where in the range revenue lands. Anyone deriving a precise Q3 expense number from a rounded revenue guide is manufacturing precision. The substantive answer, that ex-currency ex-social-charge expense growth is flat sequentially and moderates in Q4, is the one to model.

Open Source Versus Frontier Models, and What It Does to Costs

A question asked how Spotify weighs deploying open-source models more broadly against its use of commercial frontier models. The answer disclosed that the proprietary personalization model is itself open-source-derived and described the internal routing layer as something the company may commercialize.

Q: "Over the course of 2026, we've seen meaningful momentum in open source. How are you thinking about the costs and benefits of deploying open source more broadly versus how you're using Anthropic, Claude and other models today?"
— Justin Patterson, KeyBanc

A: "The Large Taste model is based on open source. It is using what is called CPT continuous pretraining, where you take a model that was trained an open source model and you continue to train it on your proprietary data… we've built this internal product called Chirp that we will also offer to other companies actually because they're asking for the same thing, where you can seamlessly switch between one of the paid models and a hosted open source model, keeping the context of the project that you are in."
— Gustav Söderström, Co-Chief Executive Officer

Assessment: Two things worth extracting. The proprietary moat asset is built on a commodity base, which means the defensibility is entirely in the data and the continuous training loop rather than in the model architecture, exactly as management has argued. And a company planning to sell its internal developer tooling to others is signalling unusual confidence in it. Neither is in anyone's model and the second should not be, but it tells you where management thinks its advantage sits.

Why More Artists Have Not Signed On to the Remix Product

A pointed question asked whether management is surprised that more artists have not joined the AI music product, and what is holding them back. The answer reframed the premise around the two label agreements already signed and then explained why a full catalogue is not required to launch.

Q: "Are you surprised more artists have not signed on to the AI music tier to enable you to launch? What is stopping artists from participating?"
— Rich Greenfield, LightShed Partners

A: "What we're doing is something different and artist see that. Our products are about real artists, not fake artists. And in the case of remix, real artists with real voices, right?… you should expect to see a research preview coming out of our model where we start to give this to some users. And for that, we do not require, as we said before, a full catalog. Why is this important for us? So this is not a test of the product. It is the thing that guarantees that the product will be very good."
— Gustav Söderström, Co-Chief Executive Officer

Assessment: The answer converts a licensing question into a machine-learning one, and does so legitimately. A limited research preview generates the human preference data that post-training requires, so a partial catalogue is not a compromise but a starting condition. It also means the product will improve fastest exactly where fan engagement is highest, which is where the compensation case to rights holders is easiest to make. This is a well-designed sequence and it is years from mattering to revenue.

Read-Through From a Competitor's Price Increase

A question asked whether a competitor's July price increase carried any read-through for Spotify's own pricing plans. Management declined to discuss the competitor and used the answer to restate its own pricing posture in language it has not used before.

Q: "Is there any read-through from the Apple Music price increase that was announced in July? And how does this impact your thinking about price increases going forward?"
— Jason Helfstein, Oppenheimer

A: "We don't comment on other companies' price increases. But what it does show is that this is an example of just continued value in music streaming services… And of course, the more value we deliver, the more pricing power we will have. And as of right now, we're happy to be the price leader in our category."
— Alex Norström, Co-Chief Executive Officer

Assessment: The most valuable thing here is the sequence rather than the words. Spotify raised US pricing in January and the largest platform competitor followed six months later. In a market where content costs are effectively an industry-wide input, competitors moving prices in the same direction is worth more to the gross margin thesis than any single increase, because it means the category is not competing on price. This is quiet but genuinely thesis-relevant.

Whether Industry Subscriber Growth Is Slowing

The closing question set Spotify's strong net adds against label commentary that industry premium growth slowed in the second quarter, and asked how confident management is in long-term subscriber growth for both. The answer was brief and declined to hedge.

Q: "You had solid premium net adds this quarter, but the record labels suggested the industry's premium growth slowed in the second quarter. How confident are you in the long-term growth of premium subscribers for the industry and Spotify?"
— Jason Bazinet, Citi

A: "I'm very confident. We don't give guidance for long term, but what we're optimizing for is a healthy funnel. This not only builds our business, Jason, it's, of course, aimed at helping our subscriber growth and not slowing it down. So very confident about the long-term growth."
— Alex Norström, Co-Chief Executive Officer

Assessment: The question is the more informative half. If the labels are reporting decelerating industry premium growth while Spotify adds 7 million net subscribers against a 6 million guide and crosses 300 million, Spotify is taking share. That is a better read on competitive position than anything management could have said, and it sits awkwardly alongside a market that spent the session marking the stock down on a user metric the company suppressed on purpose.

What They're NOT Saying

1. What M&A means. The CFO said the company expects to return cash to shareholders "even with M&A." That is the first acknowledgment of acquisition intent in four quarters of calls, it was delivered as a subordinate clause, and nobody followed up. For a business whose entire margin thesis rests on organic operating leverage, this is the most under-examined sentence of the quarter.
2. How much of the Q3 gross margin step-down is the regulatory charge. Management named the annual third-quarter regulatory-fee charge as a driver of the 50bp sequential decline but did not size it, so the underlying trend cannot be separated from the known one-off.
3. Free-to-paid conversion, for a fifth consecutive quarter. The entire emerging-markets friction strategy is a bet on raising conversion, and the conversion rate itself remains undisclosed. Management is asking the market to accept a lower user number in exchange for a metric it will not publish.
4. The Audiobooks+ base. $100 million of annual recurring revenue is a genuine first disclosure. What is not given is the subscriber count behind it, the attach rate, or which markets it covers beyond "a select few," so the runway cannot be sized.
5. Any timing on add-on tier rollouts. Asked directly about the timing of add-on tiers across verticals and the ARPU impact into 2027, management declined on both.
6. What the remix product costs to run. Management has described inference-driven products as carrying a marginal cost per use and credit purchases as carrying "a defined amount of inference." Neither the unit economics nor the pricing was disclosed.
7. Any 2026 full-year guidance, for a fourth consecutive quarter. The 2030 targets are now public, which is a real improvement, and the current year is still directional only.
8. Whether Chirp becomes a business. Management said the internal developer tooling "will also offer to other companies actually because they're asking for the same thing," then moved on. No model, no timing, no framing of whether a consumer subscription company intends to sell developer infrastructure.

Market Reaction

• Pre-print setup: SPOT closed at $486.33 on August 3, down 16.3% year to date against an 11.0% gain for the S&P 500, and down 26.2% over the trailing twelve months, but flat over the trailing thirty days at plus 0.1% from $485.97. The 52-week closing range was $412.75 to $738.53.
• Reaction session (August 4, before-the-open print): Opened at $466.95, a 4.0% gap down, fell to an intraday low of $462.00 (down 5.0%), then reversed to a high of $505.32 (up 3.9%) before closing at $478.17, down 1.7% or $8.16.
• Volume: 4.9 million shares against a 1.7 million thirty-day average, or 2.8x normal.
• Context: The S&P 500 rose 1.8% on the same session, so the stock underperformed the market by roughly 3.5 percentage points despite closing well off its low.

The intraday reversal is the story. The stock gapped down 4.0%, traded 5.0% below the prior close within the first hour, and then recovered the entire decline and more before fading into the close. The $43.32 intraday range on 2.8x volume is the signature of a genuine disagreement being resolved in real time, and the resolution went most of the way toward the buyers. Same-day coverage written against the opening leg described a 6% decline; the session closed down 1.7%.

What moved it down. Two forward numbers and one backward one. The Q3 MAU guide of 788 million against roughly 793.5 million of consensus, the Q3 operating income guide of €670M against roughly €678M, and Q2 adjusted EPS of €2.61 against €2.76. The MAU line is the one the headlines led with, which is unfortunate given that it is the one management deliberately chose.

What brought it back. Our read is that the recovery tracked the call rather than the release. The release contains the MAU guide; the call contains the explanation, the €200M spend sizing with Q4 moderation, the advertiser count, the Audiobooks+ figure and the Merlin deal. A market that sells the print and buys the call is a market that found the disclosure better than the numbers.

What it does not settle. The stock closed 35.3% below its 52-week high and has underperformed the index by a wide margin over twelve months. Two consecutive quarters of below-consensus user guidance, whatever the reason, have cost the company the benefit of the doubt on the metric that most investors still use as shorthand for the franchise.

Street Perspective

Debate: Is Trading MAU for Monetization the Right Call?

Bull view: The constructive case is that marginal emerging-market free users have been close to worthless, at roughly €0.90 of quarterly revenue per ad-supported user and falling, and that a company criticized for four quarters for failing to convert engagement into revenue is now doing exactly what critics asked. Sign-up quality filtering, low-end device deprecation and ad-load friction are surgical rather than blunt, and management states subscriber growth is unaffected, which the in-line Q3 subscriber guide supports.

Bear view: The skeptical camp holds that MAU is the top of the only funnel Spotify has, that management has spent years arguing the free tier is the single best leading indicator of everything else, and that deliberately shrinking it contradicts the company's own framework. It also arrives immediately after a quarter in which MAU missed the company's own guidance, which raises the question of whether the strategy is a choice or a rationalization.

Our take: The strategy is right and the timing of the disclosure invites the bear reading. The tell is the subscriber line: guiding MAU 5.5 million below consensus while guiding subscribers in line is not what a company with a demand problem does, because a demand problem shows up in both. We would treat Q3 as dispositive. Subscribers at or above 305 million with MAU near 788 million vindicates it; subscribers below 305 million means the friction cost more than management said.

Debate: Does the Advertising Inflection Actually Arrive in the Second Half?

Bull view: The bull argument now has numbers rather than assurances: automated channels at nearly 40% of segment revenue from just over 30% one quarter earlier, 33,000 active advertisers up 60%, 99% of impressions on the owned stack, price optimization complete, the direct channel described as stabilizing, and the segment already positive year over year with 179bp of margin expansion. Management upgraded the target from vague improvement to explicit double-digit growth in the second half.

Bear view: The bear argument is that the same second-half target has been repeated for four quarters, that ad revenue per free user is still declining, and that constant-currency growth of 3% is a long way from double digits with one quarter to go. Reaching double-digit growth in Q3 requires an acceleration of seven percentage points in a single quarter, which nothing in the trajectory supports.

Our take: The bear arithmetic on Q3 is right and the target is for the second half, not for Q3 specifically, so Q4 is where it has to land. We move this pillar back to contained on the strength of the demand-side disclosure, which is the first evidence that the buyer base is expanding rather than migrating. The advertiser count is the metric to track; if 33,000 is 40,000 by year end, the revenue follows.

Debate: What Is the Business Worth Against Its Own 2030 Targets?

Bull view: The bull case is now anchored rather than asserted. A mid-teens revenue CAGR from €17.2B in 2025 compounds to roughly €34.6B by 2030, and operating margin above 20% puts operating income above €6.9B. Enterprise value of roughly €76.8B is about 11x that figure, for a debt-free business with €9.4B of cash, 300 million subscribers, demonstrated pricing power, an advertising option approaching free and a gross margin trending toward a 35% to 40% target from 33.4% today.

Bear view: The bear framing is that 2030 targets set by a management team that has guided the forward operating line below consensus three quarters running are aspiration, not a plan; that the mid-teens CAGR assumes a subscriber engine the company is now deliberately throttling at the top of the funnel; and that a 35% to 40% gross margin requires either label concessions never yet obtained or an advertising business that has missed for six quarters to reach 40% segment margins.

Our take: Both are fair and the discount is already large. At 28.9x trailing operating income the shares require the near-term compounding to continue but do not require the 2030 targets to be hit; at 11x the target-implied 2030 operating income there is room for management to miss its own framework by a wide margin and still leave the shares cheap. That asymmetry is the position. Our fair-value range rises to $560 to $650.

Model Update Needed

Valuation framework. At the August 4 close of $478.17 on 205,788,241 basic weighted-average shares, market capitalization is approximately $98.4B. Converted at the 0.8756 USD to EUR rate the company used for its own Q3 guidance, that is roughly €86.2B; net of €9.4B of cash and short-term investments against no debt other than lease liabilities, enterprise value is approximately €76.8B. Against trailing-twelve-month revenue of €18,113M that is 4.2x, against trailing-twelve-month operating income of €2,653M it is 28.9x, and against trailing-twelve-month free cash flow of €3.3B it is 23.3x. Against the €2.8B to €2.9B we model for full-year 2026 operating income, it is 26x to 27x.

The four-quarter progression of enterprise value against trailing operating income is the clearest summary of what has happened to this stock: roughly €101B at 51x in November, €75B at 34x in February, €68.5B at 28.5x in April, and €76.8B at 28.9x now. The multiple stopped compressing this quarter, and trailing operating income has risen 34.4% over the period, from €1,974M to €2,653M.

The 2030 anchor. A mid-teens revenue CAGR applied to 2025 revenue of €17,186M produces roughly €34.6B by 2030, and operating margin above 20% on that base produces more than €6.9B of operating income. Today's enterprise value of €76.8B is about 11.1x that figure. We treat this as a sanity check on the terminal value rather than as a forecast, and note that even a substantial miss against the framework leaves the current price undemanding.

Valuation impact: Raising the fair-value range to $560 to $650 from $530 to $620, on the published long-range framework and the improved quality of the operating result. The midpoint of $605 implies 26.5% upside to the $478.17 close.

Thesis Scorecard Post-Earnings

The pillars below are the standing thesis carried since our November initiation, graded against this quarter.

Overall: Thesis strengthened. Three of the five bear pillars were challenged this quarter and the two bull pillars that carry the valuation were confirmed with harder evidence than at any point in our coverage. The one pillar that weakened, the funnel, weakened because management chose to narrow it, and the subscriber line held. Across four quarters the arc is unusually clean: the business improved in every quarter we covered while the multiple fell by nearly half.

Action: Maintaining Outperform with the fair-value range raised to $560 to $650. We would move to Hold if Q3 subscribers come in below the guided 305 million, which would mean the emerging-market friction cost more than management said, or if full-year 2026 operating margin fails to improve after four consecutive reiterations. We would move to Underperform on evidence that the advertising business is not reaching double-digit growth by the fourth quarter, or on a large acquisition that changes the organic-leverage character of the story.