<\/span><\/h2>\nThe official biographies in the filings confirm part of the founder narrative in the old article and fail to confirm another part.<\/p>\n
Daniel Ek.<\/strong> According to the 2025 report, Ek is the company’s founder and Executive Chairman of the board; he served as CEO until 31 December 2025 and has been a board member since 21 July 2008. Before founding Spotify in 2006, he founded the online advertising company Advertigo (acquired by Tradedoubler), held senior roles at the Nordic auction company Tradera (acquired by eBay) and worked as Chief Technology Officer (CTO) at Stardoll, a fashion and entertainment community for preteens. The 2018 F-1 contains the same background; at that time his title was co-founder, CEO and chairman of the board.<\/p>\nThe following claims from the old article were not found in any of the primary records consulted and were removed: “founded his first company at 14 in 1997”, “dropped out of KTH in the eighth week”, “became a millionaire at 23”, “became CEO of uTorrent” and “met Lorentzon at Tradedoubler”. The only link the filings support is that Advertigo, founded by Ek, was acquired by Tradedoubler, founded by Lorentzon.<\/p>\n
Martin Lorentzon.<\/strong> According to the 2025 report, Lorentzon is a co-founder and board member; he chaired the board from 2008 to 2016 and sat on the board of Telia Company, Sweden’s main telecom operator, from 2013 to 2018. In 1999 he founded the Stockholm-based internet marketing company Tradedoubler and initially sat on its board; he holds a master’s degree in civil engineering from Chalmers University of Technology. Tradedoubler’s own corporate page also states that the company was founded in Stockholm in 1999 by Felix Hagn\u00f6 and Martin Lorentzon and has been listed on the Stockholm Stock Exchange since 2005. The old article’s statement “studied at Handels” does not appear in the primary records and is therefore marked as unverified.<\/p>\nHow does the complementarity show up in the records?<\/strong> Ek’s background rests on product and technology (ad tech, a marketplace, a CTO role); Lorentzon’s rests on founding an internet company and taking it public (Tradedoubler, 1999 and 2005). According to the major shareholders table in the 2025 report, as of 31 December 2025 Ek held 13.9 percent of ordinary shares and 28.8 percent of total voting power, and Lorentzon held 9.3 percent of ordinary shares and 40.5 percent of voting power; the two founders’ combined voting power is 69.3 percent (CEOtudent calculation). The old article’s claim “Daniel Ek is the largest shareholder” is correct on ordinary share ownership according to this table, but not on voting power: Lorentzon leads on voting power. According to the fiscal year 2025 report, the company is run by two co-CEOs: Alex Norstr\u00f6m and Gustav S\u00f6derstr\u00f6m.<\/p>\n<\/span>Why is Spotify’s business model a “two-sided” negotiation?<\/span><\/h2>\nMost of Spotify’s revenue comes from subscriptions: according to the F-1, Premium revenue made up 90 percent of total revenue in 2016 and 2017, and according to the 2025 report, 89 percent in 2025 and 88 percent in 2024. In 2025 Premium revenue was \u20ac15,350 million and ad-supported revenue \u20ac1,836 million.<\/p>\n
The first side of the model is the listener. The F-1 writes that the ad-supported free tier works as a “funnel” and that, based on data tracked since February 2014, more than 60 percent of total gross new Premium subscribers came from this tier. The 2025 report describes the same funnel in more cautious language, as “a significant portion of gross new Premium subscribers”.<\/p>\n
The second side of the model is the rights holders. According to the F-1, content licensed from Universal Music Group, Sony Music Entertainment, Warner Music Group and Merlin, which represents independent labels, accounted for approximately 87 percent of total streams in 2017. In the 2025 report, the same four parties account for approximately 72 percent of streams of audio content provided by record labels. The company explicitly presents this as a risk factor: when supplier concentration is high, one or a few parties alone can adversely affect the business. The rise of Spotify’s gross margin from 11.6 percent in 2015 to 32 percent in 2025 (the 2015 ratio is a CEOtudent calculation, the 2025 ratio is printed in the report) is the outcome of this negotiation over the years; the F-1 also notes that the new licensing agreements signed with the major labels in 2017 had a positive effect on costs and margins.<\/p>\n
The lesson applies to one-person businesses too: if a few suppliers control your product’s input (an API, a content license, a single distribution channel), your gross margin is not your design but the result of the other side’s bargaining power.<\/p>\n
<\/span>What do the numbers say? Spotify’s verified 13-year table<\/span><\/h2>\nThe table below combines figures taken as printed from four SEC filings (the 2018 F-1; the 20-F reports for fiscal years 2020, 2022 and 2025). Values are in \u20ac million; subscriber and user counts are in millions of people at year end. Net result is net result “attributable to owners of the parent”.<\/p>\n
\n\n\n| Year<\/th>\n | Revenue<\/th>\n | Gross profit<\/th>\n | Operating income \/ (loss)<\/th>\n | Net income \/ (loss)<\/th>\n | Premium subscribers (m)<\/th>\n | Monthly active users (m)<\/th>\n<\/tr>\n<\/thead>\n |
\n\n| 2013<\/td>\n | 746<\/td>\n | 128<\/td>\n | (98)<\/td>\n | (63)<\/td>\n | not reported<\/td>\n | not reported<\/td>\n<\/tr>\n |
\n| 2014<\/td>\n | 1,085<\/td>\n | 174<\/td>\n | (191)<\/td>\n | (188)<\/td>\n | not reported<\/td>\n | not reported<\/td>\n<\/tr>\n |
\n| 2015<\/td>\n | 1,940<\/td>\n | 226<\/td>\n | (235)<\/td>\n | (230)<\/td>\n | 28<\/td>\n | 91<\/td>\n<\/tr>\n |
\n| 2016<\/td>\n | 2,952<\/td>\n | 401<\/td>\n | (349)<\/td>\n | (539)<\/td>\n | 48<\/td>\n | 123<\/td>\n<\/tr>\n |
\n| 2017<\/td>\n | 4,090<\/td>\n | 849<\/td>\n | (378)<\/td>\n | (1,235)<\/td>\n | 71<\/td>\n | 159<\/td>\n<\/tr>\n |
\n| 2018<\/td>\n | 5,259<\/td>\n | 1,353<\/td>\n | (43)<\/td>\n | (78)<\/td>\n | 96<\/td>\n | 207<\/td>\n<\/tr>\n |
\n| 2019<\/td>\n | 6,764<\/td>\n | 1,722<\/td>\n | (73)<\/td>\n | (186)<\/td>\n | 124<\/td>\n | 271<\/td>\n<\/tr>\n |
\n| 2020<\/td>\n | 7,880<\/td>\n | 2,015<\/td>\n | (293)<\/td>\n | (581)<\/td>\n | 155<\/td>\n | 345<\/td>\n<\/tr>\n |
\n| 2021<\/td>\n | 9,668<\/td>\n | 2,591<\/td>\n | 94<\/td>\n | (34)<\/td>\n | 180<\/td>\n | 406<\/td>\n<\/tr>\n |
\n| 2022<\/td>\n | 11,727<\/td>\n | 2,926<\/td>\n | (659)<\/td>\n | (430)<\/td>\n | 205<\/td>\n | 489<\/td>\n<\/tr>\n |
\n| 2023<\/td>\n | 13,247<\/td>\n | 3,397<\/td>\n | (446)<\/td>\n | (532)<\/td>\n | 236<\/td>\n | 602<\/td>\n<\/tr>\n |
\n| 2024<\/td>\n | 15,673<\/td>\n | 4,724<\/td>\n | 1,365<\/td>\n | 1,138<\/td>\n | 263<\/td>\n | 675<\/td>\n<\/tr>\n |
\n| 2025<\/td>\n | 17,186<\/td>\n | 5,496<\/td>\n | 2,198<\/td>\n | 2,212<\/td>\n | 290<\/td>\n | 751<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n Source: Spotify Technology S.A., Form F-1 (28 February 2018; 2013-2017 data; per the F-1’s note, the 2013 and 2014 figures are presented without retrospective application of IFRS 15 and may not be fully comparable with later years), Form 20-F fiscal year 2020 (2018-2020), Form 20-F fiscal year 2022 (2021-2022), Form 20-F fiscal year 2025 (2023-2025). Premium subscriber counts for 2015-2017 are given in the F-1 as “approximately”.<\/p>\n The table’s story fits in one sentence: while revenue grew 23-fold in 12 years, the company posted an operating loss in 10 of 13 years. The accumulated deficit disclosed in the F-1 just before the direct listing was \u20ac2,427 million; most of the 2017 net loss of \u20ac1,235 million came from finance costs rather than operations, and the operating loss was \u20ac378 million.<\/p>\n Spotify’s About page (accessed 6 November 2026, an undated company statement) lists 788 million users and 306 million subscribers in 184 markets; it was not included in the table because it is not audited year-end data.<\/p>\n <\/span>How many years did the move from loss to profit take? CEOtudent analysis<\/span><\/h2>\nThe table below is CEOtudent analysis: it was recalculated with calc.py from the filing figures above. No row is printed in this form in the filings.<\/p>\n \n\n\n| Indicator<\/th>\n | Value<\/th>\n | Calculation basis<\/th>\n<\/tr>\n<\/thead>\n | \n\n| Founding to direct listing<\/td>\n | 12 years<\/td>\n | 2006 -> 3 April 2018<\/td>\n<\/tr>\n | \n| Founding to first annual operating profit<\/td>\n | 15 years<\/td>\n | 2006 -> 2021 (\u20ac94 million)<\/td>\n<\/tr>\n | \n| Service launch to first annual operating profit<\/td>\n | 13 years<\/td>\n | 2008 -> 2021<\/td>\n<\/tr>\n | \n| Founding to first annual net profit<\/td>\n | 18 years<\/td>\n | 2006 -> 2024 (\u20ac1,138 million)<\/td>\n<\/tr>\n | \n| Loss years in the filings<\/td>\n | 10 of 13 years<\/td>\n | 2013-2025 operating result<\/td>\n<\/tr>\n | \n| Cumulative operating result 2013-2023<\/td>\n | \u20ac(2,671) million<\/td>\n | 10 loss years + 2021 profit<\/td>\n<\/tr>\n | \n| Cumulative operating result 2024-2025<\/td>\n | \u20ac3,563 million<\/td>\n | 1,365 + 2,198<\/td>\n<\/tr>\n | \n| Cumulative operating result 2013-2025<\/td>\n | \u20ac892 million<\/td>\n | sum of 13 years<\/td>\n<\/tr>\n | \n| Revenue growth 2013-2025<\/td>\n | 23.0x; CAGR 29.9%<\/td>\n | 746 -> 17,186<\/td>\n<\/tr>\n | \n| Premium subscriber growth 2015-2025<\/td>\n | 10.4x; CAGR 26.3%<\/td>\n | 28 -> 290 million<\/td>\n<\/tr>\n | \n| Monthly active user growth 2015-2025<\/td>\n | 8.3x; CAGR 23.5%<\/td>\n | 91 -> 751 million<\/td>\n<\/tr>\n | \n| Gross margin<\/td>\n | 11.6% (2015) -> 20.8% (2017) -> 25.6% (2023) -> 32.0% (2025)<\/td>\n | gross profit \/ revenue<\/td>\n<\/tr>\n | \n| Premium subscribers as a share of active users<\/td>\n | 30.8% (2015) -> 46.4% (2018) -> 38.6% (2025)<\/td>\n | subscribers \/ MAU, year end<\/td>\n<\/tr>\n | \n| Revenue per average full-time employee<\/td>\n | \u20ac1.38 million (2017) -> 1.45 (2023) -> 2.36 (2025)<\/td>\n | revenue \/ average employees<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n Source: CEOtudent analysis; input figures are taken from the verified table above and from the average employee counts in the F-1 and the 2025 20-F (2016: 2,084; 2017: 2,960; 2023: 9,123; 2024: 7,691; 2025: 7,287). The 2025 gross margin is also printed in the report as 32 percent; the other years are calculated.<\/p>\n Three findings stand out. First, Premium subscribers as a share of active users peaked at 46.4 percent in 2018 and fell to 38.6 percent in 2025; this is the math of geographic expansion, because according to the 2025 report the “rest of world” region accounts for 37 percent of users and grew 21 percent in one year. Second, revenue per employee rose from \u20ac1.45 million in 2023 to \u20ac2.36 million in 2025; the average headcount fell by 20.1 percent. The 2025 report records two workforce reductions, of approximately 6 percent in January 2023 and approximately 17 percent in December 2023, and \u20ac212 million in severance costs; the first large operating profit (\u20ac1,365 million) came the very next year. Third, Premium ARPU in 2025 was \u20ac4.63, below \u20ac4.69 in 2024; the report attributes this to currency and product and market mix, and puts the effect of price increases at \u20ac0.25. In other words, the 2025 margin improvement came more from cost and scale than from price.<\/p>\n <\/span>CEOtudent editorial framework: 7 lessons from the Spotify story for the 2026 professional<\/span><\/h2>\nThe framework below is a CEOtudent editorial framework; each lesson is tied to a single verified fact sourced above and translated for a professional into “the CEO move” and “the student move”.<\/p>\n \n\n\n| #<\/th>\n | Lesson<\/th>\n | Verified fact it rests on<\/th>\n | The CEO move (direction and decision)<\/th>\n | The student move (learning)<\/th>\n<\/tr>\n<\/thead>\n | \n\n| 1<\/td>\n | Timing a broken model<\/td>\n | Recorded music revenue fell from $23.8 billion to $16.9 billion between 1999 and 2008 (F-1)<\/td>\n | Look for opportunity not in a growing market but where demand persists while the payment model has collapsed<\/td>\n | Read the breaks in your target industry’s revenue model over the last 10 years from primary reports<\/td>\n<\/tr>\n | \n| 2<\/td>\n | Supplier concentration sets the margin<\/td>\n | Four rights holders accounted for 87% of streams in 2017 and 72% in 2025; gross margin rose from 11.6% to 32%<\/td>\n | Write the number of parties controlling your input and their contract terms into the business plan; treat margin as “negotiated”, not “designed”<\/td>\n | List your license, API and platform dependencies and learn an alternative for each<\/td>\n<\/tr>\n | \n| 3<\/td>\n | The free tier is a funnel, not a gift<\/td>\n | More than 60% of gross new Premium subscribers came from the ad-supported tier (F-1, data tracked since February 2014)<\/td>\n | Build the conversion path and its measurement for everything you offer for free from the start<\/td>\n | Measure the free-to-paid conversion rate in your own work for three months; there is no funnel you do not measure<\/td>\n<\/tr>\n | \n| 4<\/td>\n | Patience with losses, but patience with criteria<\/td>\n | First annual operating profit 15 years after founding; 10 of 13 years loss-making; two workforce reductions in 2023, \u20ac1,365 million operating profit in 2024<\/td>\n | Write down in advance how long you will accept losses and your exit criterion; Spotify’s 2023 restructuring was a decision, not an accident<\/td>\n | Learn to read unit economics (gross margin, revenue per subscriber); revenue growth alone misleads<\/td>\n<\/tr>\n | \n| 5<\/td>\n | Founder complementarity<\/td>\n | Ek: Advertigo, Tradera, Stardoll CTO; Lorentzon: founded Tradedoubler in 1999, took it public in 2005; combined voting power 69.3%<\/td>\n | When choosing a partner or first hire, look for a track record that covers the area your own background leaves empty<\/td>\n | Map your own skills; decide whether to close the gap with a partner or through learning<\/td>\n<\/tr>\n | \n| 6<\/td>\n | Build the structure early, then hand over<\/td>\n | Luxembourg legal entity in 2006; direct listing in 2018; the founder stepped down as CEO on 31 December 2025, and two co-CEOs were appointed<\/td>\n | Make your goal not to become indispensable but to build the structure that can fill your place<\/td>\n | Document processes so you can hand them over; unwritten knowledge cannot be handed over<\/td>\n<\/tr>\n | \n| 7<\/td>\n | Go deep first, then go wide<\/td>\n | 61 countries and territories (2017) -> 184 (2025); 100 million tracks, 7 million podcasts, 500,000 audiobooks (2025 report)<\/td>\n | Decide to enter a new market or launch a new product only after the unit economics of the existing core are settled<\/td>\n | Set a deliberate ratio between exploration and depth; do not try to maximize both at once<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n Source: CEOtudent editorial framework; facts from the Spotify F-1 (2018) and 20-F (fiscal year 2025) filings, calculated ratios from calc.py.<\/p>\n The common thread of the framework is this: in the Spotify story, structural decisions matter more than a “brilliant idea” moment: the contract structure with rights holders, the free-to-paid conversion mechanism and 15 years of loss tolerance. For a professional, the equivalent is less the question “what to do” and more “which dependencies, which conversion path and which patience period to accept”.<\/p>\n | | |