The Multiple Fell a Third on a Story About AI While the P&L Did the Opposite: Upgrading Spotify to Outperform

Results vs. Consensus

Spotify reports under IFRS in euros. All actuals and comparatives below trace to the consolidated statement of operations furnished with the results and, for the segment split, to the audited segment note in the annual filing published the same day. Two comparisons are shown: against management's own point-estimate guidance, which is the bar the call grades, and against the Street.

Versus company guidance

Versus Street consensus

The same warning we attached to the Q3 EPS beat applies with more force here. Net income of €1,174M contains €320M of net finance income and a €153M income tax benefit, so €473M of it, 40.3%, sits below the operating line. The diluted-EPS calculation strips out €232M of fair-value gains on the Exchangeable Notes from the numerator, which is why diluted EPS of €4.43 is well below basic EPS of €5.71, but it leaves the tax benefit and the rest of finance income intact. A 55% beat against a €2.85 consensus tells you that the sell-side does not forecast these items, which is correct of them.

Year-over-Year Comparison (Q4 2025 vs. Q4 2024, € millions)

Full Year 2025 vs. 2024 (€ millions)

The four reported quarters sum to the full-year figures exactly: revenue of €4,190M, €4,193M, €4,272M and €4,531M totals €17,186M, and operating income of €509M, €406M, €582M and €701M totals €2,198M. Constant-currency growth rates are as disclosed by the company. Note that the CFO's spoken summary on the call ("revenue grew 13%, gross profit grew 20% and operating income grew in excess of 50%") is on a constant-currency basis for revenue and gross profit, which he flagged at the top of his remarks; the reported figures are 9.7% and 16.3%.

Sequential Comparison (Q4 2025 vs. Q3 2025, € millions)

Revenue. Reported growth of 6.8% is the weakest-looking headline of the year and the least informative. Currency cost roughly 580bp of year-over-year growth against the roughly 620bp assumed at the time of guidance, and on a constant-currency basis revenue grew 13%, an acceleration from 12% in Q3. The composition improved in both segments for the first time in a year: Premium accelerated to 14% constant currency from 13%, and advertising went from flat to positive 4%. Full-year revenue of €17,186M grew 9.7% reported and 13% in constant currency.

Margins. This is the line that changes the thesis. Consolidated gross margin of 33.1% is a record, up 83bp year over year and 146bp sequentially. More importantly, the segment attribution reversed. In Q3, Premium gross margin fell 34bp while the ad segment appeared to expand 525bp, and management confirmed a large part of that divergence was a cost reclassification rather than performance. In Q4, Premium gross margin rose 10bp to 34.8% and the ad segment rose 441bp to 19.5%, so both segments improved on their own terms. Operating margin of 15.5% is also a record, and the full-year operating margin of 12.8% is 408bp better than 2024's 8.7%.

Operating expense and the R&D line. Reported operating expense fell 10.4%, but the honest number is the one management disclosed: excluding currency and social-charge movements, operating expense rose about 13% year over year, driven by marketing and personnel. The exception inside that is research and development, which fell 22.9% to €290M in the quarter and 6.3% for the full year, while the company shipped more than 50 features in 2025. Some of that is currency and grant timing. Some of it, on management's own account, is agentic coding. We treat the R&D line as the most interesting cost item in the model and discuss it separately below.

Below the line, again. The mechanism we described in November ran in the same direction for a second consecutive quarter. The share price fell from $698.00 at the September close to $580.71 at the December close, a decline of 16.8%. That produced a €67M social-charge credit against forecast inside operating expense and, because the Exchangeable Notes are carried at fair value through profit and loss, €232M of fair-value gains inside finance income. A €153M tax benefit completes the picture. Investors should not extrapolate any of it. Note the corollary that matters for 2026: if the share price rises from here, social charges flip from a tailwind to a headwind, and the reported operating-income comparisons that looked so strong in the back half of 2025 get harder for reasons that have nothing to do with the business.

Segment Performance

Segment gross profit is derived from the audited full-year segment note less the nine-month figures previously reported, and it foots exactly to the reported Q4 gross profit of €1,499M.

Premium — the margin line finally turns

Premium added 9 million net subscribers to finish at 290 million, up 10.3% year over year and one million ahead of guidance. Revenue of €4,013M grew 8.3% reported and 14% in constant currency, and for the first time in several quarters the constant-currency revenue growth rate meaningfully exceeded the subscriber growth rate, which is the arithmetic signature of price working. Premium ARPU of €4.70 fell 3% reported but rose 2% in constant currency, against flat in Q3. Management guided Q1 2026 ARPU growth to 5% to 6% in constant currency, which is the US increase flowing through.

The gross margin line is the one we singled out in November as the bear case. Premium gross margin of 34.8% expanded 10bp year over year, reversing the 34bp contraction in Q3. Management's attribution is that revenue growth outpaced music costs net of marketplace programmes and audiobook costs, "largely offset by video podcast costs." So video is still consuming most of the underlying improvement. But "largely offset" is a materially better position than "more than offset," and it arrives in the first quarter after the full renewal round with the majors and the top five US publishers was concluded. That sequencing matters: whatever Spotify paid for the broader video rights, the segment margin went up in the quarter immediately after.

Assessment: The single most important datapoint in the quarter, and it is not the record consolidated margin. Premium is 88.6% of revenue and its margin was going the wrong way. It has stopped. With the US price increase landing in Q1 and management guiding ARPU acceleration, the setup for Premium margin in 2026 is the best it has been since audiobooks were introduced into the bundle.

Ad-Supported — the first positive constant-currency quarter of the transition

Ad-supported revenue of €518M fell 3.5% as reported but grew 4% in constant currency, positive constant-currency growth after a flat Q3. On a like-for-like basis, excluding the effects of the deliberate podcast inventory optimization, the CFO put growth at roughly 7%. Ad-supported MAU reached 476 million, up 12.0% year over year and up 30 million sequentially. Ad-supported gross margin of 19.5% expanded 441bp.

The CFO's framing of what the last 18 months were for is worth recording, because it is the clearest statement yet of why the segment shrank.

"It's now 1.5 years since we decided to reengineer Spotify's ad stack and really move off of a rented stack… It was a tough call back in that moment since it meant that — I knew it meant that we had to take some pain as this was going to be deep surgery for us. We now have — I'm happy to say that we now have record levels of advertisers on the platform, and that increased density means much better yields and as a result, more revenue growth for us." — Christian Luiga, CFO

The H2 2026 inflection target was reiterated without change, which after two consecutive quarters of that date moving out is itself a small positive. For the full year, advertising revenue of €1,836M fell 1.0% reported and grew 4% in constant currency, so the segment was roughly flat in a year in which it was supposed to be transitioning.

Assessment: We moved this from our sharpest bear point to a contained risk. Positive constant-currency growth, 7% like-for-like, record advertiser density and a 441bp margin expansion are consistent with a transition that is working rather than a business that is impaired. It is one quarter, the reported line still shrank, and we continue to model the segment flat in constant currency for 2026 with the H2 acceleration as upside. But the direction of travel changed.

Geographic mix

US revenue grew 5.4% for the full year in euro terms while the rest of the world grew 12.3%, and the gap is mostly currency rather than volume. This is the mix effect that produces the reported ARPU decline. It also frames why the January US increase matters disproportionately: 37.6% of revenue sits in the market with the lowest growth rate and the price lever that had not been pulled since mid-2024.

Key Operating Metrics

Two numbers in that table deserve a second look. Headcount was 7,323 at the end of both Q3 and Q4, so the company added 38 million users and shipped a quarter's worth of product without adding a person. And ad-supported MAU of 476 million against ad revenue of €518M works out to €1.09 per free user for the quarter, up from €1.00 in Q3. That is the first sequential improvement in free-user monetization we have seen, off a very low base.

Key Topics & Management Commentary

Overall Management Tone: Deliberately reframing. This was the first call run by the new leadership and roughly half of it was spent answering, in various forms, whether generative AI breaks Spotify. Management chose to pre-empt rather than react, devoting a large block of prepared remarks to the argument before a single question was asked, and the posture was confident to the point of being combative about the premise. On the numbers management was straightforward and volunteered the social-charge distortion unprompted for a second consecutive quarter. The one place the answers stayed thin was any forward disclosure ahead of the May investor day, which was invoked repeatedly as the venue for detail.

1. A 34% De-Rating That the Income Statement Does Not Explain

Since our November note the shares fell 34.9% over the trailing twelve months and 23.1% over the trailing thirty days into this print, closing at $414.84 the day before results. Over that same window the company delivered a record gross margin, record MAU additions, record operating income and a 61.0% increase in full-year operating profit. The gap between those two facts is the entire investment question, and an analyst put it to management directly.

"With the stock down approximately 1/3 over the last 3 months, the market appears to be implying Spotify will be negatively impacted from AI. What do you think the market is missing from how Spotify can benefit from AI?" — question read on the call

The CFO declined to comment on the share price and pivoted to the substance, and that is the right posture. But the framing of the question is the most useful piece of market information on the call: the sell-side itself characterised the drawdown as an AI de-rating, not a fundamentals de-rating. Nothing in the quarter, the guide or the full-year result supports a fundamental explanation.

Assessment: Multiple compression of this magnitude against improving fundamentals is the setup we look for. The AI risk is real and we treat it as a genuine pillar below, but the market has moved from pricing Spotify at roughly 51x trailing operating income in November to roughly 34x now, on a business whose operating income grew 61% in the year just reported. That is the upgrade.

2. The US Price Increase Lands, and Churn Behaves

In November we named the untaken US price increase as the largest unexercised lever on the income statement and made it an explicit upgrade trigger. In January, Spotify raised the US Premium price to $12.99 from $11.99. Management addressed the outcome directly.

"I'm really happy with the price increases we implemented back in January of this year. There have been really no surprises at all. Churn is low and came in according to our expectations. And just as a reminder, this $1 increase is the same magnitude as the U.S. price increase that we implemented back in — I think it was June of 2024." — Alex Norström, Co-Chief Executive Officer

The financial evidence is in the guide rather than the quarter, since the increase hits February billing cycles. Q1 2026 ARPU is guided to grow 5% to 6% in constant currency, against 2% delivered in Q4, and a co-CEO said the guide is only for "a portion of the quarter," which means the full run-rate benefit lands in Q2. The cost side is visible too: Q1 subscriber net additions are guided to 3 million, against 5 million in Q1 2025, and management explicitly said the pricing effects are in the forecast while characterising the level as within the historical range for a first quarter.

Assessment: The trigger fired, and it fired cleanly. A dollar on 37.6% of revenue, delivered with churn management describes as in line, is worth more to 2026 gross margin than anything else discussed on this call. The three-million-subscriber Q1 guide is the price of it, and it is a price worth paying: Spotify is choosing revenue per user over user count for the first time in a while, which is what a maturing subscription business should do.

3. Record 33.1% Gross Margin, and This Time the Reason Is Content Costs

Consolidated gross margin of 33.1% is the highest in the company's history, 83bp better year over year and 146bp better sequentially. The composition of the beat versus guidance is what changed. In Q3 the CFO attributed the margin outperformance "primarily" to revisions of prior-period rights-holder liability estimates relating to the first half of 2025, which is a true-up rather than a result. This quarter the attribution was different and simpler.

"Our outperformance here was primarily driven by content cost favorability." — Christian Luiga, CFO

Asked what drives gross margin expansion in 2026 and whether the drivers shift, management named four: pricing outpacing net content cost growth, the gradual advertising improvement, marketplace programmes that added to both gross profit and margin in 2025, and vertical and geographic expansion. The CFO also committed to the first of those explicitly.

"What I said was that the price increases that we have done here is going to outpace the net content cost in 2026." — Christian Luiga, CFO

Assessment: That sentence is the most consequential forward statement on the call, and it is the first direct answer to the question we said in November could not be settled with the information disclosed. It does not quantify the renewal economics, but it does commit management to the outcome that matters: net content cost grows slower than price. If 2026 gross margin does not expand, that is now a broken commitment rather than an ambiguity.

4. Record MAU Net Additions and the Best Wrapped in the Company's History

Spotify added 38 million monthly active users in the quarter, the largest quarterly addition in its history, to reach 751 million, 6 million above guidance. Management attributed the outperformance to the global rollout of the enhanced free tier and to a Wrapped campaign that was itself a record: more than 300 million users engaged, up 20%, with more than 630 million social shares, up 42%. One disclosure connected the campaign directly to the subscriber line.

"Even more, day 1 of Wrapped marked the highest single day of subscriber intake in Spotify history." — Alex Norström, Co-Chief Executive Officer

Asked when the free-tier changes stop being a headwind to conversion and start being a tailwind, management used the leaky bucket framing: fix retention first and the water level rises faster. It is the same argument as last quarter, but the evidence behind it is now stronger. In Q3 the company added 17 million MAU and 5 million subscribers. In Q4 it added 38 million MAU and 9 million subscribers, so the ratio of subscriber adds to MAU adds did not collapse even as MAU adds more than doubled.

Assessment: In November we flagged that the Q4 guide implied 32 million MAU adds against only 8 million subscriber adds and said the gap was either the funnel filling or the funnel getting less efficient. The answer, at least for one quarter, is the former: both numbers beat. The Q1 guide of 8 million MAU and 3 million subscriber adds is a different shape, but that is the price increase, not the funnel.

5. The AI Question, Part One: Does the Business Model Survive?

Management pre-empted the AI debate in prepared remarks rather than waiting for Q&A, and the core argument was structural rather than product-led: disruption comes from new business models, not new technologies, and the consumer business model is not changing.

"My view is that new technology is seldom disruptive on its own. Significant disruption happens when new technologies enable new asymmetric business models… However, in the consumer space that we are in, we believe the dominant business model will continue to be ads plus subscription, both places where Spotify excels. This puts Spotify in an outstanding position because we already have the right business model." — Gustav Söderström, Co-Chief Executive Officer

The supporting evidence offered was that Spotify has been building for this for years rather than starting now: the Sonantic acquisition in 2022, the AI DJ launched in 2023 with about 90 million subscribers having used it and over 4 billion hours of time spent driven, and the recently launched Prompted Playlist, described as the "deep research mode" of Spotify against the DJ's chat mode. The strategic claim underneath is a data claim: Spotify is accumulating a language-to-music dataset that does not exist elsewhere, and unlike an encyclopedia it cannot be commoditized because there is no factual answer to what workout music is.

Assessment: The business-model argument is the strongest version of the bull case and we find it persuasive as far as it goes. Ads plus subscription is indeed what the AI labs themselves are converging on. The argument's weak point is that it addresses disintermediation of the monetization model without fully addressing disintermediation of the interface: if discovery migrates into a general assistant, Spotify becomes a licensed backend even if the subscription survives. Spotify's answer to that is the ChatGPT integration and the ubiquity strategy, which is a real answer, but it is a defensive one.

6. The AI Question, Part Two: Generated Music, Derivatives, and Whether Suno Becomes a DSP

The sharpest version of the bear case was put to management as a bear case: that the AI music generators become distribution platforms in their own right and take share while Spotify takes a cautious approach. The response leaned on industry alignment rather than technology.

"I've done the rounds and no rights holder is against our vision. We pretty much have the whole industry lined up behind us. And like Gustav mentioned before, we want to do it in a controlled way, respecting artists and the community, and we will not do deals that isn't — that aren't good for artists and ultimately Spotify." — Alex Norström, Co-Chief Executive Officer

On policy, Spotify declined to disclose what share of uploaded music is AI-generated and framed the tooling question as not its decision to make, while working with the industry on creation-method metadata that can be surfaced to listeners. The more interesting commercial framing was the split between net-new AI music and derivatives of existing catalogue, where management sees an unmonetized opportunity.

"In terms of the derivatives, as I said, we think this is an untapped opportunity for artists to make money off of their existing IP. We have the technology and capabilities that we need, and we're very excited about it. And we are ready for the partners that are hungry to seize this opportunity. We think the ones that move first will benefit the most." — Gustav Söderström, Co-Chief Executive Officer

The counter to the catalogue-explosion fear was also stated plainly: a bigger catalogue makes personalization more valuable, not less, and the catalogue has already grown roughly tenfold across Spotify's life without breaking the model.

Assessment: The net-new-music argument is strong, because the cultural moment and the royalty pool are both on Spotify regardless of which tool made the track. The derivatives opportunity is genuinely interesting and entirely unquantified. The AI-native services becoming DSPs is the real tail risk and management's answer to it is essentially that the labels will not license them on favourable terms, which is a bet on the industry structure holding rather than on Spotify's own moat. We keep this as a live risk and note that management would not put a number on AI-generated upload volume.

7. Honk, Agentic Coding, and Why the R&D Line Fell 22.9%

Research and development expense fell 22.9% year over year in the quarter and 6.3% for the full year, while the company shipped more than 50 features in 2025. Management described an internal system that puts a specific mechanism behind the number.

"An engineer at Spotify on their morning commute from Slack on their cell phone can tell Claude to fix a bug or add a new feature to the iOS app. And once Claude finishes that work, the engineer then gets a new version of the app pushed to them on Slack on their phone so that he can then merge it to production, all before they even arrived at the office. We call this system internally Honk." — Gustav Söderström, Co-Chief Executive Officer

Pressed on what this means for engineer productivity and R&D investment needs, the answer was more specific than the usual corporate AI commentary, including a claim about senior engineers that is checkable over time.

"When I speak to my most senior engineers, the best developers we had, they actually say that they haven't written a single line of code since December. They actually only generate code and supervise it. So it is a big change. It is real and it's happening fast." — Gustav Söderström, Co-Chief Executive Officer

Management explicitly declined to translate this into an R&D budget reduction, arguing instead that software companies will produce far more software and that the limiting factor becomes how much change consumers tolerate. That is the honest answer and also the one that does not promise margin.

Assessment: Headcount was flat sequentially at 7,323 while the company added 38 million users, and R&D fell 22.9% in a quarter with record feature output. We are not modelling an AI productivity dividend, because management has not promised one and currency and grant timing explain part of the decline. But this is the second consecutive quarter in which the R&D line fell while output rose, and it is the least-discussed potential margin lever in the story. Watch R&D as a percentage of revenue through 2026: it was 8.1% of full-year 2025 revenue against 9.5% in 2024.

8. The Co-CEO Structure in Practice

This was the first call run by the new leadership, and the question of what has actually changed produced a concrete operational answer rather than a platitude. The two chief executives chose not to split the organization.

"We decided not to. We decided to run this as a single direct-reporting group, something that we run weekly for 3 hours called eTeam… So we have the entire decision layer of Spotify, sort of the VP/SVP layer in this room 3 hours every week, deciding and running and unblocking the entire company. So there has been a shift in how we operate, and we focus even more on synchronization and planning." — Gustav Söderström, Co-Chief Executive Officer

In November we identified the specific risk of the co-CEO split as having two owners for the single most contested resource-allocation question at Spotify. Running a single combined leadership group with a weekly three-hour decision forum is a direct structural answer to that concern. Whether it holds under stress is unknowable from one quarter.

Assessment: Better than we expected. The risk is not eliminated, and the first genuinely hard trade-off has not yet arrived. But the design choice is the correct one and it was made deliberately rather than defaulted into. We downgrade this from a named concern to a monitoring item.

9. Capital Allocation: $510M Repurchased, and a €1.5B Note to Settle in Cash

Spotify repurchased $433M of shares in Q4, bringing full-year 2025 activity to $510M, against €9.5B of cash and short-term investments and €2.9B of full-year free cash flow. Asked about capital returns given the balance sheet, the CFO restated the reinvestment-first priority and then disclosed a use of cash that had not been prominent.

"And in addition to that, as you know, we have 1.5 billion fallen due or plus in convertible note now in March, which we will settle in cash." — Christian Luiga, CFO

The balance sheet carries the Exchangeable Notes at €1,458M as a current liability at year end, against €1,539M as a non-current liability a year earlier. Settling in cash rather than in shares removes the dilution and also removes the fair-value line that has been swinging reported net income by hundreds of millions of euros each quarter.

Assessment: Retiring the notes for cash is a quietly significant cleanup. It removes the notes' contribution to the diluted share count and, more usefully for analysts, removes the largest single source of noise in reported earnings. From Q2 2026 the net income line should start to mean something again. The buyback remains sized to offset employee-equity dilution rather than to return capital, which at €9.5B of cash and a business with €22M of quarterly capital expenditure is a conservative posture that will eventually need revisiting.

10. Audiobooks Expand, and Spotify Enters the Physical Book Market

Audiobooks in Premium launched in Sweden, Denmark, Finland, Iceland and Monaco, taking the catalogue past 500,000 titles across 14 global markets, with publishers crediting Spotify with driving double-digit category growth. The more unexpected development is a physical-book partnership, which management framed as consumer-led rather than commercial.

"The reason we want to be in the physical book market is because we think that it's not a separate market. It is the same book market… from the consumer, it's the same book, whether it's a physical book, it's on their Kindle or their audiobook." — Gustav Söderström, Co-Chief Executive Officer

Management was explicit that Spotify is not holding inventory. The stated goal is sync across formats rather than book retailing.

Assessment: Strategically coherent and financially immaterial in 2026. The risk worth naming is scope creep: a company that has just been rewarded for cost discipline is opening a new vertical with structurally worse economics than software. The mitigation is that it is a partnership rather than an owned operation. We would rather see the energy go into the ad stack, but the sums involved do not move the model.

11. 2026 Framing: "The Year of Raising Ambition" and a May Investor Day

Management labelled 2026 the Year of Raising Ambition, following 2025's Year of Accelerated Execution, and scheduled an investor day for May 21 in New York, with the company's twentieth anniversary marked at South by Southwest in March. The ambition itself was framed in the prepared remarks as converting a much larger share of the world's population.

"We proudly count 3.5% of the world as subscribers, and there's still lots of room to grow. It's not impossible to imagine us converting 10% or even 15% of the world's population to subscribers." — Alex Norström, Co-Chief Executive Officer

Several answers deferred detail to the investor day, including the product roadmap and the long-term vision.

Assessment: The May investor day is now the most important scheduled catalyst in the story, and it is where we expect the long-term margin framework to be reset. A company that has just delivered 408bp of operating margin expansion in a single year and refuses to guide the full year is holding its framework back for a venue. Between now and then the model runs on the Q1 guide and the pricing flow-through.

Guidance & Outlook

The Q1 revenue guide of €4.5B is flat sequentially in euros and represents roughly 15% constant-currency growth, an acceleration from the 13% just delivered, with an assumed FX headwind of about 670bp based on a USD to EUR rate of 0.8514 at the December close. The CFO also sized the incremental currency drag versus prior-quarter rates at €35M, which is a level of specificity the company has not previously offered and is useful for anyone bridging the guide.

The gross margin guide of 32.8% deserves attention because of what it is not. In November the CFO pre-announced that Q1 gross margin "typically sees a sequential step down from the fourth quarter from advertising seasonality." A 30bp step-down from 33.1% is a much smaller decline than the seasonal pattern would imply, and it is 120bp above the year-ago quarter. An analyst made exactly this point and management confirmed the read without extending it to the full year.

"While we do not give full year guidance for gross margin and operating margin, we are expecting both to improve in 2026." — Christian Luiga, CFO

The subscriber guide is the one soft spot. Three million net additions in Q1 against five million in the year-ago quarter is a real deceleration, and management named the cause as the pricing implementation while calling the level "within our historical range for quarter 1." Both statements can be true. We read it as an accepted trade: two million subscribers of near-term growth exchanged for a dollar of monthly price on the largest revenue base in the portfolio. On roughly 120 million US-and-equivalent subscribers, a dollar a month is worth considerably more than two million subscribers at €4.70.

Implied full-year 2026. Management declined to guide the year beyond the direction of margins. Taking the Q1 guide as the base and assuming the pricing benefit annualizes while the FX drag persists at the currently implied level, we model full-year 2026 revenue of €18.6B to €19.1B (8% to 11% reported, low-to-mid teens constant currency) and operating income of €2.9B to €3.2B. That range embeds one important caveat: 2025's operating income was helped by social-charge credits in both Q3 and Q4 totalling €108M against forecast, created by a falling share price. If the shares recover, that reverses.

Guidance style. Unchanged. Point estimates, conservatively set, beaten on four of five metrics with revenue in line. Spotify has now beaten or met every guided metric in each of the last two quarters.

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 the Market Is Missing on AI, With the Stock Down a Third

The most direct question of the call named the de-rating explicitly and asked management to argue against it. The answer conceded that AI has been hard for the market to grasp, declined to comment on the share price as a matter of policy, and rested the case on having started early rather than on any new disclosure.

Q: "With the stock down approximately 1/3 over the last 3 months, the market appears to be implying Spotify will be negatively impacted from AI. What do you think the market is missing from how Spotify can benefit from AI? And what are your top priorities so you don't fall behind within this new industry landscape?"
— Steven Cahall, Wells Fargo

A: "I think it's been notable listening today's discussion and also seeing the last quarter, of course, that AI has been something that has been hard to grasp for many people. We don't comment on our share price when it changes like in this short term and so on, and we will not do that going forward. But it's obvious from the recent months, but also from the discussion today, I would say, and all the questions we get that AI is something that is interesting and will have an impact. And I think, hopefully, we have discussed and explained why this is a great opportunity for us. And as Gustav said before, we didn't start now. We started many years ago. And if you haven't, you probably will have a tougher time."
— Christian Luiga, CFO

Assessment: The answer is honest and thin, and the thinness is instructive. Management has no way to disprove a narrative about the future with a quarter of results, and it declined to try. What it did instead was let the numbers argue: record margin, record users, 61% operating income growth. For an investor, the useful observation is that the entire de-rating rests on a claim with no supporting evidence in the disclosed financials, while the counter-case has four quarters of it. That asymmetry is what we are buying.

Whether the AI Music Generators Become Distribution Platforms

A question laid out the bear thesis in its strongest form: that the generative music services enable consumers to create music and then become streaming services themselves, taking share while Spotify moves cautiously. Management's answer did not engage the technology and instead argued from industry structure, that no rights holder opposes Spotify's approach and that a working monetization model is the scarce asset.

Q: "Is Spotify playing to win in AI? The bear thesis on Spotify is that Udio, Suno, Klay and Stability not only enable consumers to create AI music, but also become DSPs that take share from Spotify with Spotify taking a more cautious approach. Any comments on that?"
— Rich Greenfield, LightShed Partners

A: "At Spotify, we provide a service to rights holders and artists and songwriters, a service to distribute and monetize their art. And the key point here is this is a scaled service with a working business model. This is where you go to put your new songs, whichever technology or instrument or tool you use to create it, and I've done the rounds and no rights holder is against our vision. We pretty much have the whole industry lined up behind us."
— Alex Norström, Co-Chief Executive Officer

Assessment: This is a bet on industry structure, not on product. It is a reasonable bet, because a generative service that wants to become a general-purpose DSP needs catalogue licences on terms the majors have no incentive to grant a competitor to their own distribution economics. But it is worth being precise about what management did and did not claim: they claimed alignment with rights holders, not technical or product superiority over the generators. If the majors' incentives ever change, this answer does not hold.

Price-to-Value After the January US Increase, and How Churn Compares

A question asked how the price-to-value relationship looks after the US increase relative to competitors and how churn compares to prior rounds. The answer is the cleanest confirmation on the call that the largest lever in the model was pulled without damage, and it anchored the magnitude against a known precedent.

Q: "Following the recent U.S. price increases, how do you see the price-to-value relationship of the service relative to your competitors? And how do you expect churn to play out versus prior rounds of price increases?"
— Batya Levi, UBS

A: "I'm really happy with the price increases we implemented back in January of this year. There have been really no surprises at all. Churn is low and came in according to our expectations… The one important thing to point out, though, is that price increases, as you know, is one of several levers we pull for growth. And when we adjust price, we do it from a position of strength… ultimately, what we strive to do is to always create more value than price."
— Alex Norström, Co-Chief Executive Officer

Assessment: Three questions on US pricing last quarter produced three non-answers about frameworks. One question this quarter produced a completed action and a churn characterisation. The gap between the two calls is the difference between a Hold and an Outperform. The number to hold management to is the guided 5% to 6% constant-currency ARPU growth in Q1, which is checkable in eleven weeks.

What Actually Drives Gross Margin Expansion in 2026

A question asked for the drivers of 2026 gross margin expansion and whether they differ from recent years. The answer named four and, unusually for this management team, made a falsifiable commitment on the first of them.

Q: "What are the drivers of gross margin expansion in 2026? And do they shift at all from recent years?"
— Doug Anmuth, JPMorgan

A: "What I said was that the price increases that we have done here is going to outpace the net content cost in 2026. Remembering also that we are improving our ads business slowly as we go forward, and we feel that, that will pick up in the second half of 2026. We have a marketplace that added both to gross income and margin in 2025. That is also a good tool for us. And finally, as we expand new verticals within the countries that we are in and also to new countries, that is also a good support for our margin development."
— Christian Luiga, CFO

Assessment: The renewal economics are still undisclosed, but this is the functional answer to the question we said in November could not be settled. Price growth exceeding net content cost growth is precisely the condition under which the label deals are not a problem, and management has now put it on record. The other three drivers are second-order. We would still prefer the rate disclosure, and we no longer think its absence is disqualifying.

Eighteen Months Into the Ad Stack Rebuild, Where Does It Stand

A question asked for a progress report on the advertising platform rebuild, where the most progress has been made, where work remains, and whether a step change arrives later this year. The answer was the most candid account yet of why revenue went backwards.

Q: "You've spent the last 2 years building out your ad tech platform. Can you provide a progress report? Where are you seeing the most progress? And where do you have more work to do? And will there be a step change in advertising growth later this year?"
— Jessica Reif Ehrlich, Bank of America

A: "It's now 1.5 years since we decided to reengineer Spotify's ad stack and really move off of a rented stack. And we did this primarily to better match what our clients ask of us, the way they would like to buy on Spotify… It was a tough call back in that moment since it meant that — I knew it meant that we had to take some pain as this was going to be deep surgery for us. We now have — I'm happy to say that we now have record levels of advertisers on the platform, and that increased density means much better yields and as a result, more revenue growth for us. We are positive on ads. We still have work to do, but we're definitely making good progress and seeing very positive signs."
— Christian Luiga, CFO

Assessment: "Deep surgery" is a better description of the last six quarters than anything management offered while it was happening, and the retrospective candour is welcome. The disclosure gap we flagged in November remains: advertiser count is up to a record, but programmatic share of revenue, the metric that would let an investor track the crossover, is still not given. The difference is that this quarter the constant-currency line finally turned positive, so the claim now has a number behind it.

What Changed Operationally After the Founder Stepped Back

A question asked what has changed at Spotify in the early days after the founder moved out of the chief executive role. The answer described a deliberate structural choice that speaks directly to the standard objection to co-CEO arrangements.

Q: "Curious, what's changed at Spotify in the early days following Daniel stepping back from the CEO role?"
— Rich Greenfield, LightShed Partners

A: "Structurally, some things have changed because first and foremost, Alex and I are 2 people. So we had 2 direct-reporting teams. And we thought long and hard about how we were going to do that. Were we going to sort of split the thing down the middle, manage our own teams, have our own meetings? We decided not to. We decided to run this as a single direct-reporting group, something that we run weekly for 3 hours called eTeam… So there has been a shift in how we operate, and we focus even more on synchronization and planning."
— Gustav Söderström, Co-Chief Executive Officer

Assessment: This answers the specific risk we named last quarter. A single combined decision forum meeting weekly is the structural mitigation for a divided-ownership problem, and the fact that the alternative was explicitly considered and rejected suggests the risk was understood internally. It has not been stress-tested by a genuinely contested decision. But as first-quarter evidence goes, it is the right design.

Agentic Coding, Product Velocity, and What It Means for R&D Spend

A question asked how agentic coding is changing product velocity and what generative AI could mean for engineer productivity and R&D investment needs. The answer contained the most specific productivity claim any consumer-internet management team has made on a call, and pointedly declined to convert it into a spending forecast.

Q: "How is agentic coding changing product velocity? What do you believe GenAI could mean for engineer productivity and R&D investment needs?"
— Justin Patterson, KeyBanc

A: "When I speak to my most senior engineers, the best developers we had, they actually say that they haven't written a single line of code since December. They actually only generate code and supervise it. So it is a big change. It is real and it's happening fast… I think what will happen is something more like what happened with the Internet. When the Internet came along, everyone thought that we would all have our own web pages. What actually happened was there ended up being very few web pages. In times of lower friction, things actually tend to aggregate, not disaggregate."
— Gustav Söderström, Co-Chief Executive Officer

Assessment: The aggregation argument is the most under-appreciated point on the call and it cuts directly against the AI bear case: lower production costs historically concentrate demand into a few destinations rather than fragmenting it. Management deliberately did not promise an R&D reduction, which is the credible choice, but R&D fell to 8.1% of revenue in 2025 from 9.5% while feature output rose. We do not model a productivity dividend; we do watch for one.

What They're NOT Saying

1. Any full-year 2026 guidance. Direction of margins only, with everything else deferred to the May investor day. After a year in which operating margin expanded 408bp, refusing to frame the next one is a choice about venue, not about visibility.
2. The label and publisher renewal economics, still. Second consecutive quarter with all deals concluded and no rate, term or step-up disclosed. The CFO's commitment that price growth outpaces net content cost growth in 2026 is a useful substitute, and it is a promise rather than a number.
3. What share of uploaded music is AI-generated. Asked directly, with a follow-up on daily upload volume, management declined on both. In the middle of a de-rating driven by exactly this question, choosing not to disclose a number that would presumably help if it were small is notable.
4. Programmatic share of advertising revenue. Third consecutive quarter without it. Record advertiser counts and "better yields" are directionally useful and do not let anyone track the crossover the H2 2026 target depends on.
5. Free-to-paid conversion rate. Asked again, answered again with engagement and active days. Q4's data makes the inference easier, but the metric itself remains undisclosed.
6. The size of the derivatives opportunity. Management describes AI derivatives of existing catalogue as untapped, says it has the technology ready and that first movers benefit most, and offers no market size, no timeline and no product.
7. Any quantification of the Honk productivity gain. Senior engineers reportedly no longer write code, R&D fell 22.9%, and management explicitly would not connect the two into a forward R&D expectation.
8. What happens to social charges if the stock recovers. The company correctly refuses to forecast its share price, and the consequence goes unmentioned: €108M of favourable social-charge variance across Q3 and Q4 reverses into a headwind if the shares rally, making 2026 operating-income comparisons harder for non-operational reasons.

Market Reaction

• Pre-print setup: SPOT closed at $414.84 on February 9, down 28.6% year to date against a 1.7% gain for the S&P 500, down 34.9% over the trailing twelve months and down 23.1% over the trailing thirty days from $539.37. The stock entered the print just above its 52-week closing low of $412.75, against a 52-week high of $775.90.
• Reaction session (February 10, before-the-open print): Opened at $462.78, an 11.6% gap up, traded a range of $462.78 to $495.94, and closed at $476.02, up 14.7% or $61.18. The open was the low of the day.
• Volume: 10.9 million shares against a 2.7 million thirty-day average, or 4.0x normal, the heaviest participation of the period.
• Context: The S&P 500 fell 0.3% on the same session, so the entire move was idiosyncratic.

A 14.7% single-day gain on a quarter that delivered revenue in line with guidance and an operating beat that was two-thirds social charges is not a reaction to the print. It is a reaction to the positioning. The stock came in two dollars off a 52-week low having lost roughly a third of its value in three months on a thesis about AI, and the quarter did not merely fail to confirm that thesis, it produced record margins and record user additions in the same window. When expectations are set that low, in line is a beat.

What actually moved it. Three things, in our reading. The record 33.1% gross margin and the €701M operating income gave the numbers-driven case something concrete. The confirmation that the January US price increase produced low, in-line churn removed the single largest execution risk in the 2026 model. And management spent a large share of the call making a structured argument that AI is a tailwind, which the market had been assuming without evidence was a headwind.

What it does not settle. The stock closed the session 38.7% below its 52-week high, so the AI de-rating has been dented rather than reversed. Volume at 4.0x average with the open at the day's low suggests genuine repositioning rather than a short squeeze, but one quarter of margin evidence does not resolve a multi-year question about whether generative models change how people find music.

Street Perspective

Debate: Is Generative AI a Threat to Spotify or an Accelerant?

Bull view: The constructive case on the Street is that AI improves the two things Spotify monetizes, personalization and production velocity, without touching the business model. A larger catalogue makes recommendation more valuable rather than less, the language-to-taste dataset Spotify is accumulating is not replicable from public data, and the historical pattern in lower-friction media is aggregation into a few destinations rather than fragmentation. The internal evidence supports it: R&D fell while output rose, and 90 million subscribers have used the AI DJ.

Bear view: The skeptical camp holds that music becomes commoditized content when anyone can generate it, that the cultural role of the DSP erodes as discovery migrates into general assistants, and that the AI-native music services eventually secure catalogue rights and become distributors. On this reading Spotify's ChatGPT integration is not a moat but an admission that the interface layer is moving somewhere else, and the multiple should compress toward a licensed-backend business.

Our take: The bear case is a real risk on a five-year view and is not a 2026 or 2027 earnings risk, which is the distinction the market has stopped making. Management's structural argument (disruption requires a new business model, and ads-plus-subscription is not being replaced) is the strongest formulation available and we accept it. What we do not accept is that a company disclosing record margins, record users and 61% operating income growth should carry a lower multiple than it did when Premium gross margin was contracting. The market has priced a narrative and ignored a print.

Debate: Does the 2026 Gross Margin Expansion Actually Arrive This Time?

Bull view: The bull argument is that all four drivers now point the same way for the first time: the US and international price increases flow through from Q1, the Spotify Partner Program cost reclassification annualizes out, marketplace programmes are additive, and advertising stops being a drag. Q4 delivered a record 33.1% and the Q1 guide of 32.8% is a far shallower seasonal step-down than the pattern, which implies the underlying trend is stronger than the optics.

Bear view: The bear argument is that Premium gross margin expanded only 10bp year over year even in a record quarter, that video podcast costs "largely offset" the underlying improvement and are still growing, that the renewal rates remain undisclosed a full year after the negotiations began, and that management explicitly reserves the right to spend the margin on "disciplined investments" whose timing makes quarterly progression variable. On this reading, 2026 gross margin lands near 33% and the operating leverage is once again deferred.

Our take: We have moved to the bull side of this debate, which is the substance of the upgrade. Last quarter the disclosed facts supported gradual improvement at best and Premium margin was going backwards. This quarter Premium margin turned positive, the beat came from content costs rather than an accrual true-up, and the CFO committed on the record that price growth outpaces net content cost growth in 2026. We model full-year 2026 gross margin of 33.0% to 33.6%.

Debate: What Should a Business Like This Be Worth After a One-Third De-Rating?

Bull view: At an enterprise value of roughly €75B against €2,198M of 2025 operating income and a business compounding revenue in the low teens in constant currency with 408bp of annual margin expansion, negligible capital intensity and €8.0B of net cash, the shares are near 34x trailing and mid-twenties forward operating income. For a subscription franchise with 290 million paying users, a demonstrated price lever and an advertising option approaching free, that is a reasonable price for quality rather than a premium.

Bear view: The bear framing is that the multiple was 51x three months ago and is 34x now for a reason, that terminal-value uncertainty in a business whose content is about to become abundant justifies a permanently lower multiple, and that the operating income figure being capitalized is inflated by €108M of social-charge credits that reverse if the stock recovers. Adjusting for that and for a normalizing tax rate, the real forward multiple is higher than it looks.

Our take: The social-charge point is fair and we adjust for it: excluding the Q3 and Q4 variances, 2025 operating income was closer to €2.09B, which puts the trailing multiple nearer 36x than 34x. That does not change the conclusion. The de-rating has taken the shares from a price that required both the ad inflection and the US price increase to a price that requires only one of them, and the one it requires has already happened. Our fair-value range is $560 to $640.

Model Update Needed

Valuation framework. At the February 10 close of $476.02 on 205,771,982 basic weighted-average shares, market capitalization is approximately $97.9B. Converted at the 0.8514 USD to EUR rate the company used for its own guidance, that is roughly €83.4B; net of €9.5B of cash and short-term investments less the €1,458M of Exchangeable Notes settling in cash, enterprise value is approximately €75.4B. Against full-year 2025 revenue of €17,186M that is 4.4x, against full-year 2025 operating income of €2,198M it is 34x, and against the €2.9B to €3.2B we model for 2026 it is 24x to 26x. Against €2.9B of 2025 free cash flow it is 26x. For comparison, the same calculation at our November note put enterprise value at roughly €101B and 51x trailing operating income.

Valuation impact: Raising the fair-value range to $560 to $640 from $560 to $620. The midpoint of $600 implies 26.0% upside to the $476.02 close, which supports the upgrade even after a 14.7% single-day move.

Thesis Scorecard Post-Earnings

The pillars below are the ones established in our November initiation, graded against what this quarter revealed. Status tags are carried forward from the standing thesis.

Overall: Thesis strengthened materially. Both bear points that drove the November Hold were challenged by this quarter's evidence, the single largest upside lever was exercised with the outcome management predicted, and the multiple compressed by roughly a third while all of that happened. A new bear point replaced them, and it is a narrative risk rather than a demonstrated one.

Action: Upgrading to Outperform from Hold. In November we committed to upgrading on either a confirmed US price increase or two consecutive quarters of positive constant-currency advertising growth; the first has fired and the second is one quarter in. We would move back to Hold if full-year 2026 gross margin fails to expand despite the pricing benefit, or if Q1 constant-currency ARPU growth comes in materially below the guided 5% to 6%. We would move to Underperform on evidence that AI-native services are securing major-label distribution rights on competitive terms.