[scuttleslops] SPOT, UMG, TXN, SPGI, TRU, ADSK, MTCH, CSGP, DLTR
The very first scuttleslop! Let me know what you think. Again, the text that follows is substantially AI-generated, but try not to be too negatively biased by that fact. You might imagine that I hired a junior analyst and tasked them with summarizing the most recent and relevant newsflow and management commentary related to my coverage in a weekly email. Give it a chance!
Spotify and Universal sign a licensed-AI deal that re-cuts the music value chain
At Spotify’s Investor Day on May 21, Spotify and Universal Music Group announced new recorded-music and publishing licensing agreements that, for the first time, give a major DSP the right to let users generate licensed AI covers and remixes of participating UMG artists’ songs. The tool will launch as a paid add-on for Premium subscribers, with artists and rightsholders opting in; UMG framed it around “consent, credit, and compensation,” with payments flowing to both Masters and publishing rightsholders.
Strategically, this is a coordinated response to Suno and Udio, which have built fast-growing user bases on top of unlicensed catalog and are facing label lawsuits. Spotify and UMG are betting that a sanctioned, opt-in experience inside a Premium subscription beats unlicensed generative tools so long as the labels share the upside. Spotify is trying to make generative AI additive to rightsholders rather than a piracy-like threat. The pitch is that fans will create derivative works inside a paid, licensed system, artists and songwriters will share in the value, and Spotify will capture “fan spending amongst our super users that doesn’t exist today.”
For the labels, this is another kind of negotiated rent extraction: any new format eventually has to draw on the rightsholders’ catalog, and the big three can tax it. For Spotify, it’s a higher-ARPU, paid feature monetized above the base subscription that could improve monetization if the economics are favorable (the royalty structure has not been disclosed).
The risk is that this also legitimizes per-track generative play that, at scale, could fragment listener attention away from the catalog itself, especially if it bleeds into discovery in the main app. WMG was not part of the deal, but its stock rose alongside Spotify’s on the implication that similar terms will follow.
Other notable bits from the Investor Day:
Podcasts were presented as a turnaround story. Roman Wasenmüller said podcasts were previously “a highly negative business and a drag on overall company margin,” but are now in their second year of profitability, with a long-term path to 40% margins. He also said podcast engagement has doubled since the last Investor Day (in June 2022), video podcasts have been streamed by more than 500 million users, and Premium users who listen to podcasts in addition to music spend three more days per month on Spotify.
Audiobooks+ already has more than 1 million users paying on top of their Spotify subscription, and management said those users have lifetime values “multiples of premium-only users.” Owen Smith said Spotify has reached roughly 20% share of the U.S. audiobook market in just a couple of years, scaled its catalog from 150,000 titles at launch to more than 700,000, and is on track to hit $100 million of annualized recurring revenue from Audiobooks+ alone by July.
The advertising story was more self-critical. Katie English admitted that Spotify was “too dependent on direct buying, too concentrated in the U.S. and too centered on audio-only budgets,” and that “it took us too long to make the shift” toward automated, performance-based advertising. The recovery plan is a rebuilt ads platform, more programmatic buying, better measurement, AI-assisted ad creation, and broader video inventory; she said active advertisers grew 68% year over year in the first quarter and that biddable channels are now more than one-third of the ads business.
More broadly, the company no longer wants investors to see it as “just” a music-streaming subscription business; management is trying to reposition Spotify as a broader audio, video, creator, advertising, and fan-monetization platform. Alex Norström framed the starting point as scale: Spotify now has 761 million active users, nearly 300 million subscribers, operates in 184 markets, and has a subscriber base “double the size of any other music service,” while Gustav Söderström added that the platform generates 3.4 trillion daily taste signals across its products. The company’s next monetization push is built around the idea that there is “no such thing as an average user.” Spotify wants to move beyond one subscription price by selling add-ons to heavy users across audiobooks, music creation, podcasts, fan access, and potentially other verticals.
Spotify’s AI argument was basically that it does not need to own a general large language model; it needs to own the taste data layer. Söderström said the company is buying general AI capabilities from the market while building its own “large taste model” on proprietary user behavior, content metadata, creator tools, and cultural data. He argued that “taste is constantly evolving,” so the advantage comes from processing years of listening behavior plus trillions of new signals from an “insanely active user base every single day.”
Financially, Spotify emphasized that it has already become a much better business since the 2022 Investor Day. Revenue reached €17 billion in 2025, a currency-neutral compound annual growth rate of 18%; gross margin improved from 25% in 2022 to 32% in 2025; operating margin moved from roughly negative 6% to almost positive 13%; and free cash flow went from close to zero to nearly €3 billion. The 2030 financial targets are ambitious: mid-teens revenue compound annual growth, gross margin of 35% to 40%, and operating margin of at least 20%.
Sources: Billboard — Spotify and UMG strike licensing deal for AI covers and remixes · MediaNama — Spotify and UMG fans to create licensed AI covers and remixes
Related scuttleblurb post: [UMG, WMG] record labels and the music industry: opportunities and challenges (7/18/22)
TI runs the analog pricing-power experiment for real — second 2026 price hike, 15–85% across power and signal-chain ICs
A Texas Instruments customer notice dated May 7 informed distributors of a second 2026 price increase, with pricing adjustments applied to all orders and shipments from July 1. Reported moves range from 15% to 85% depending on product family, with the largest hikes concentrated in digital isolators and power-management ICs.
Analog Devices already pushed through a 10–30% hike in February, and NXP and Infineon are reportedly preparing similar moves for June/July. This is the live test of a key question for TI and the ADI/analog complex more broadly: does the long-tail, design-in, decades-of-incumbency structure actually let analog vendors take rate after a multi-year inventory correction, or does the customer’s ability to multi-source and TI’s expanded internal capacity limit pricing?
Sources: TrendForce — TI prepares second 2026 price increase for June/July · TrendForce — ADI plans 10–30% price hike following TI’s lead
Related scuttleblurb post: [TXN] Texas Instruments (8/8/22)
S&P Global (SPGI) — Bernstein Conference, May 27, 2026: CEO defends benchmark moat against AI disintermediation thesis
CEO Martina Cheung framed the company as “predominantly a benchmarks business” — 2/3 of revenue and 3/4 of profit sit in Ratings, Index, and Platts energy benchmarks where S&P is the sole or dominant provider. She emphasized the unique nature of the IP: “We are the only providers of the S&P Global Ratings, the only providers of the S&P 500, the only providers of the Platts Brent crude benchmark.”
On the central bear thesis that LLMs disintermediate data providers, Cheung pushed back hard, arguing the IP “is actually not available publicly” and recounting a large investment bank that piloted a frontier model in a sandbox, deployed it, and “had to shut it down very, very quickly because they couldn’t trust what was coming out of it.”
Ratings tailwinds remain strong with an $8 trillion maturity wall through 2028 and hyperscaler issuance running ahead of plan. Private credit ratings have grown to a base “in the hundreds of millions of dollars” off rising LP demand for transparent, consistent methodology.
Market Intelligence is the contested segment — Cap IQ Pro is “less than 6% of our revenue” and growth is dragging on slow end-market demand, but early AI-ready data renewals are getting 35–45% uplifts. Critically for capital allocation: at current valuation, “any deal, even a tuck-in size deal, quite frankly, would have to hit a really high bar” to beat returning capital — a strong signal favoring buybacks over M&A.
Related scuttleblurb post: S&P, Moody’s, and the limits of AI disruption: Part 2 (3/13/26); S&P, Moody’s, and the limits of AI disruption: Part 1 (3/3/26)