The Spotify model
Squads, tribes, chapters, and guilds. Widely copied, and by one insider's account never fully real at Spotify either.
Henrik Kniberg and Anders Ivarsson, Scaling Agile at Spotify with Tribes, Squads, Chapters and Guilds
October 2012
Jeremiah Lee, Spotify's Failed Squad Goals
2020; one former engineer's account
SI Labs case study on a bank abandoning the squad model under BaFin requirements
consultancy-sourced
Henrik Kniberg and Anders Ivarsson published the whiteboard paper Scaling Agile at Spotify in October 2012, describing squads of six to twelve, tribes of squads, chapters for craft, and guilds for interest. It was written as a snapshot of one company at one moment, with a caveat that it was already changing.
In 2020 Jeremiah Lee, a former Spotify engineer, wrote that the model as described never fully worked inside Spotify: squads lacked real autonomy, chapter leads had accountability without authority, and cross-squad coordination went unowned. That is one former engineer's account, not a company statement. A separate case study by SI Labs describes a bank that adopted the squad model and reversed it under BaFin documentation and approval requirements.
The canvas carries it as the canonical case of structure copied without preconditions. The chart travels; the beliefs, tenure, and regulatory room it needed do not.
What is the smallest unit that can own a customer outcome end to end?
What synchronizes the work when there is no room and half the team is not a person?
Which decisions may an agent make alone, and who is able to stop it?