Spotify Crosses 300 Million Subscribers as Management Reveals AI Cost Engine and Deliberately Slows Emerging-Market Growth to Force Monetization
Q2 2026 earnings call, August 4, 2026
Spotify's second quarter results confirmed what has become a familiar pattern for the streaming giant: accelerating revenue growth, record gross margins, and a subscriber base that just crossed 300 million for the first time. But the more consequential disclosures came from how the company is engineering its cost structure around AI and why it is intentionally introducing friction into its free tier in emerging markets, a decision that will suppress near-term user growth in exchange for what management believes is a much larger monetization opportunity.
Spotify Built Its Own AI Routing Layer, and Now It's Selling It
The most substantive new disclosure was about Chirp, an internal engine Co-CEO Gustav Söderström described as the infrastructure sitting beneath Spotify's coding agent, Honk. Chirp lets engineers switch between AI models mid-task, routing every job to whichever model offers the best price-performance, including open-source models Spotify hosts itself. "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," Söderström said. The system shares context across models and developers company-wide, which means Spotify avoids paying to run the same reasoning twice, and it tracks inference spend down to the individual developer.
The strategic point is that Spotify is treating vendor-agnosticism as a structural cost advantage rather than a hedge. Söderström noted that headcount has been flat for three years while revenue per employee is on track to double, meaning operating expense growth is now coming entirely from compute and marketing, both variable costs the company controls directly. Notably, management said Chirp itself may be offered to other companies, a detail that hints at a potential enterprise tooling business investors haven't previously modeled.
Spotify Is Deliberately Trading MAU Growth for Monetization in Emerging Markets
Management confirmed it is intentionally increasing friction in the free tier across select emerging markets, including tweaks to sign-up flows, deprecation of lower-end Android device support, and increased ad load. Co-CEO Alex Norström was direct about the trade-off: "This strategy carefully increases friction in our free service with a goal of driving higher user conversion and revenue growth down the line." The company's Q3 guidance of 788 million MAU, up only 11 million sequentially, reflects this deliberate slowdown, a sharp deceleration from the growth pace investors have grown accustomed to.
Norström framed this as a lagged version of the playbook that worked in Latin America, where low initial conversion eventually gave way to strong subscriber growth once the funnel matured. CFO Christian Luiga was explicit that this will not come at the expense of subscriber growth in the near term, and the Q3 subscriber forecast of 305 million, implying 5 million net adds, was presented as consistent with that claim. Still, investors should note that this marks a real strategic pivot from growth-maximization to monetization-maximization in Spotify's largest-population, lowest-ARPU markets, a segment that has driven a disproportionate share of MAU beats over the past several quarters.
The Large Taste Model Is Moving Metrics Spotify Calls the Hardest to Shift
Söderström offered unusually specific performance data on the Large Taste Model, the proprietary recommendation system trained on 3.4 trillion daily user events. Since deploying the new Autoplay system powered by the model, active days have increased even at Spotify's already-high engagement baseline, alongside growth in Autoplay minutes, track saves, and a decline in Autoplay drop-off. In chat experiences, minutes are up double digits with more active days and saves. Söderström argued this validates the scaling-law bet the company made when it shifted from older machine-learning approaches: "The old type of machine learning had capped out. More data and larger models did not produce better results. The sequence-based LLM follow different laws, the laws that are called scaling laws... Now we're seeing those." He added the company is "watching it play out with our Taste LLM on a model that no competitor can buy," a competitive-moat framing investors should weigh against the broader AI capex debate.
Reserved Ticketing Is Outperforming Internal Expectations
Spotify's Live Nation-powered ticketing feature, Reserved, has supported multiple U.S. tours since its June launch, with nearly 100,000 tickets reserved through the platform. For some tours, Spotify sold through its entire allocation, prompting Live Nation to upsize allotments mid-run. Söderström called it "probably the feature that the most people ever have said, 'this is the best thing you ever did at Spotify.'" The product is still U.S.-only and tied to premium, but management is framing it as one of the most valuable premium differentiators in the company's history, with clear optionality to expand into new markets and tiers over time. Investors should note this remains an early-stage, narrow-scope initiative; the read-through to financials is not yet material, but the sell-through and mid-run upsizing data points suggest real fan demand rather than a marketing gimmick.
Advertising Business Nears an Inflection Point After a Multiyear Rebuild
Automated ad channels now represent nearly 40% of ad-supported revenue, up from just over 30% in the first quarter, while active advertisers grew 60% year-over-year to 33,000. Ad-supported revenue growth remained muted at 3%, consistent with the prior quarter, as strength in automated channels was offset by declines in legacy direct sales. Management reiterated its expectation that the ad business will inflect to double-digit growth in the second half of 2026, supported by the completed migration to Spotify's in-house ad server and the rollout of a biddable exchange. Luiga added a specific long-term margin target: advertising margins should move from the 20% range today toward 40% over time as the business scales through automation. A smaller but telling data point: of the 33,000 active advertisers, 7,000 are already using Spotify's AI-powered audio asset creation tool, deployed via plug-ins to Claude, ChatGPT, and Gemini.
Music Remix Product Gains a Second Major Rights Holder, But Timeline Remains Vague
Following its May agreement with Universal Music Group, Spotify announced a new licensing deal with Merlin, the digital rights body representing over 30,000 independent labels, for its AI-powered remix and covers product. Management described a "3 Cs" framework, consent, credit, and compensation, as the model differentiating its approach from generic AI music generation. Söderström said the next step is a research preview allowing fans to generate and rank remixes, generating reinforcement-learning preference data he called a unique advantage given Spotify's 777 million users. Notably, management clarified that a full catalog of major-label participation is not required to launch, drawing a parallel to Spotify's own early years operating without complete catalog coverage. Investors should treat this as a multi-year monetization story rather than a near-term revenue driver; management was explicit that "there is more work to do before launch."
Margins and Cash Flow Continue to Compound Even as Investment Spend Rises
Gross margin hit a record 33.4% in the quarter, 30 basis points ahead of guidance, aided partly by timing shifts and a one-time reversal of a Canadian Digital Service Tax accrual. Operating income of EUR 655 million beat guidance by EUR 25 million, and free cash flow reached EUR 797 million, up 14% year-over-year. The company reiterated its full-year expectation of roughly EUR 200 million in incremental operating expense tied to AI and marketing investment, with Luiga noting that expense growth in the third quarter should be roughly consistent with the second quarter before moderating in the fourth. Buybacks totaled $662 million year-to-date through August 3, a 30% increase over 2025 levels, with EUR 9.4 billion in cash and no debt outside lease liabilities. Management's repeated framing, that margin is "a managed outcome, not a byproduct," underscores confidence that current spending is discretionary and reversible rather than structural.