The Most Copied Org Chart in Tech Was a Snapshot Spotify Itself Walked Away From.

Walk into almost any agile transformation in the last decade and you'd hear the same four words: squads, tribes, chapters, guilds. Whiteboards filled with them. Consultants billed against them. A generation of engineering orgs reshuffled themselves to look like Spotify, the cool Swedish music company that had apparently solved the hardest problem in software — how to stay fast while getting big. There was a whitepaper. There were diagrams. There was a name for everything. It was the most copied org chart in tech.

The official story is that Spotify invented an operating system for scaling autonomy, wrote it down, and the world adopted it. Almost every part of that is misleading. Spotify didn't formally publish a framework — two consultants described a moment. The company never fully ran the model the world copied. And when the structure was supposedly proving itself, Spotify was quietly walking away from it. The thing thousands of companies installed was a photograph of a system mid-construction — and they hung it on the wall as if it were a blueprint.

The document everyone cited told them not to copy it

The founding text of the movement is a whitepaper titled 'Scaling Agile @ Spotify with Tribes, Squads, Chapters & Guilds,' published in October 2012 by Henrik Kniberg and Anders Ivarsson.¹ Read it today and the first surprise is how tentative it is. It describes a roughly 250-person tech organization spread across more than 30 teams in three cities — and it openly admits the structure 'was introduced gradually over the past year, so people are still getting used to it.'¹ That is not the language of a proven system. That is the language of an experiment in progress, written down before anyone knew whether it worked.

The second surprise is who wrote it and why. This was not an official Spotify framework commissioned by leadership and stamped for export. It was an informal consultant report describing what two people observed at one company at one moment.² Kniberg put the warning right in the introduction: 'This is not a recipe.'² He has since denied being the inventor of the thing named after his paper — publishing a post in 2015 titled, plainly, 'No, I didn't invent the Spotify model'¹⁰ — and the terminology, by insider accounts, emerged organically inside Spotify's engineering organization: empowered engineers reaching for words to describe how they already worked, not a model handed down from above.⁸ The world took a field note marked 'do not follow this as instructions' and followed it as instructions.

Spotify itself never fully ran the Spotify Model

Here is the part that should have ended the cult before it began. The squad model — autonomous teams owning their slice of the product, loosely coordinated through chapters and guilds — was never actually completed inside Spotify. The clearest testimony comes from someone who was there: Jeremiah Lee, a former Spotify product manager who joined in 2017, by which point he says the company had tripled in size. His verdict is blunt. He learned, he wrote, that the famed squad model 'was only ever aspirational and never fully implemented.'³ The diagrams the rest of the industry was tracing were of a destination Spotify had never fully reached.

And Lee names the deeper problem, which matters more than the implementation gap. The model, he argues, 'optimized for autonomy, which is not what the firm needed.'³ This is the causal heart of the whole story. Autonomy is intoxicating to engineers and easy to evangelize, but it is not a goal — it is a trade. Maximize independence between teams and you minimize the friction that forces alignment, shared roadmaps, and someone owning the seams between products. A growing company often needs more coordination, not less. The model the world adored was tuned for exactly the thing a scaling firm should be most careful about giving away for free.

The bill came due in a December memo

If autonomy culture were the engine of disciplined growth its admirers claimed, you would expect Spotify to have been protected from the most ordinary failure mode of a fast-scaling company: hiring too many people. It wasn't. On December 4, 2023, Daniel Ek published a memo on Spotify's own newsroom announcing he had 'made the difficult decision to reduce our total headcount by approximately 17% across the company,' citing the need to right-size after excess hiring in 2020 and 2021.⁴ The most quoted line landed like a confession from inside the doctrine itself: the company had become, in his words, 'more productive but less efficient.'⁹ The structure built to prevent bloat had presided over it.

That December round was not the whole of it. Across 2023, Spotify made three rounds of cuts totalling roughly 2,300 people — about 27.5% of the 8,359 full-time employees it had reported at the end of 2022.⁷¹¹ More than a quarter of the company, gone in a year. The financial residue showed up in the filings: restructuring charges of roughly €130–145 million, mostly severance and real-estate impairments, flowing through into 2024.⁵ By the third quarter of 2024, headcount had fallen to 7,242, and the leaner cost base helped produce €454 million of operating income on revenue of €3,988 million, up 19% year over year.⁶ The business worked. But it started working better after the autonomy was disciplined, not because of it.

But didn't Spotify win anyway — so doesn't the culture get the credit?

The fair objection is that Spotify is a global winner that grew through the squad era, so picking at its org chart is just sour pedantry. Two things are true at once here, and the honest version keeps both. First, the early culture — empowered engineers reaching for their own language to describe how they wanted to work — was genuinely real and probably genuinely useful in the small, fast years; the insider testimony places its origin in that authentic impulse, not in a slide deck.⁸ No one is claiming Spotify succeeded in spite of having good engineers. The claim is narrower and harder to dodge: the thing that got exported was never the operating system that produced the early speed.

Second, correlation is doing all the work in the credit-the-culture argument. Spotify won on a streaming product, licensing economics, and audio bets — and the model was 'only ever aspirational' even as the company scaled past 3,000 people.³ When the over-hiring crisis arrived, the culture didn't catch it; a layoff memo did.⁴ You cannot credit autonomy for the growth and excuse it from the bloat. If the doctrine were the operating system its disciples installed, it would have to own both ledger entries. It owns neither cleanly — because it was a story about how Spotify worked, told more confidently than the working ever was.

The most durable thing Spotify ever shipped may not be a feature or a playlist algorithm — it's a story about itself that the rest of the industry adopted as its own. The squad model became organizational mythology: powerful, repeatable, and largely unfalsifiable, because the company that authored it had moved on before anyone could check the diagram against the building. Spotify didn't sell the world an operating system. It sold the world a photograph, taken mid-construction, of a house it later remodeled — and an entire industry moved in anyway.