Walk into almost any company that declared itself agile in the last ten years and you will find the same wall chart: squads, tribes, chapters, guilds. The vocabulary comes from a single 2012 white paper in which Henrik Kniberg and Anders Ivarsson described how Spotify then organized roughly 30 teams and 250 people into autonomous squads, each owning a slice of the product (Kniberg and Ivarsson, 2012). The paper was careful to call itself a snapshot of one moment, not a blueprint. Almost no one read that line.
The model described a company that did not exist
The most useful correction came from inside. Jeremiah Lee, who joined Spotify in 2017 as a product manager, reported that the celebrated model was aspirational and never fully implemented, and he quoted an agile coach who had helped write it: even at the time, they were not doing what the paper described (Lee, 2020). Ivarsson, a co-author, acknowledged the discomfort of watching firms copy something that did not really exist.
The failure was specific and worth naming. Spotify documented team autonomy in exhaustive detail and never finished the promised material on alignment and accountability. The org chart that spread around the world was half a system. Copying half of a self-management design is worse than copying none of it, because the missing half is the part that keeps autonomy from turning into drift. A dozen squads pointed in a dozen directions do not add up to a product. They add up to a coordination bill that someone pays later.
The research says the ritual is not the mechanism
Academic work on scaling agile points at the same gap. In a systematic review of 52 publications covering 42 industrial cases, Dikert, Paasivaara and Lassenius (2016) found that large-scale agile efforts rarely fail for want of enthusiasm. They fail on coordination between teams, on inconsistent management support, and on organizational mindset. The authors add a caveat that should humble anyone selling a framework: almost 90 percent of the studies they could find were practitioner experience reports rather than rigorous research, so the evidence base for scaling agile is thin.
When the question is turned around, the same answer appears. Surveying 298 practitioners, Palopak and Huang (2024) tested whether project success tracked the ceremonies of agile or the underlying principles of the Agile Manifesto, and found the association ran through the principles, not the rituals. The symptom of getting this backwards now has a name. Havstorm, Karlsson and Gao (2025) built an analytical tool from an ethnographic study of daily stand-ups and catalogued twelve cargo-cult behaviors across seven categories: teams that perform the ceremonies of agile while missing the principles underneath. The chart on the wall looks right. The behavior does not change. A separate review of how agile teams actually reach effectiveness reaches the constructive version of the same point: what drives results is autonomy, shared leadership, and coordination as emergent team states, not the calendar of ceremonies (Steegh, Van De Voorde and Paauwe, 2025).
Alignment is engineered, not exhorted
This is where Team Topologies supplies the half that Spotify's chart left blank. Matthew Skelton and Manuel Pais (2019) treat a team's cognitive load as a first-order design constraint. A stream-aligned team can only stay fast and autonomous if it is not forced to understand and operate everything beneath it. That is the entire point of a platform team: its job is to lower the cognitive load of the teams it serves, offering capabilities as a self-service interface. Read that way, a platform is not another layer of control. It is the infrastructure that makes autonomy survivable as an organization grows.
Aligned autonomy, the phrase Kniberg and Ivarsson (2012) actually used, was always two words doing equal work. Teams are free to decide the how, provided they are aligned on the why. The companies that pinned up the squads chart imported the freedom and skipped the alignment, then wondered why coordination collapsed.
The practical read
The question to ask before adopting any agile structure is not whether you have squads. It is whether you have built the alignment that makes squads safe: clear purpose, open information, platforms that reduce cognitive load, and coordination mechanisms between teams. Dikert, Paasivaara and Lassenius (2016) put management support and model customization at the top of their success factors for a reason. Borrow another company's org chart and you inherit a cargo cult. Borrow the principle, aligned autonomy backed by platform and purpose, and you inherit the only part that ever worked.
Sources
- Kniberg and Ivarsson (2012). Scaling Agile @ Spotify with Tribes, Squads, Chapters and Guilds. The source white paper that coined the model and explicitly framed it as a momentary snapshot built on aligned autonomy, not a template. https://blog.crisp.se/wp-content/uploads/2012/11/SpotifyScaling.pdf
- Lee (2020). Spotify's Failed #SquadGoals. An insider account, with named colleagues, documenting that the model was aspirational and that autonomy was detailed while alignment was not. https://www.jeremiahlee.com/posts/failed-squad-goals/
- Dikert, Paasivaara and Lassenius (2016). Challenges and success factors for large-scale agile transformations. A systematic review of 42 cases showing failures come from coordination and management gaps, on a notably thin evidence base. https://doi.org/10.1016/j.jss.2016.06.013
- Palopak and Huang (2024). Perceived impact of agile principles on project success. A survey of 298 practitioners linking success to Agile Manifesto principles rather than ceremonies. https://doi.org/10.1016/j.infsof.2024.107552
- Havstorm, Karlsson and Gao (2025). Agile software development method cargo cult. Names twelve cargo-cult behaviors, teams doing the ceremonies without the principles. https://doi.org/10.1016/j.infsof.2025.107851
- Steegh, Van De Voorde and Paauwe (2025). Understanding how agile teams reach effectiveness. A review locating effectiveness in autonomy, shared leadership and coordination rather than rituals. https://doi.org/10.1016/j.hrmr.2024.101056
- Skelton and Pais (2019). Team Topologies. Frames cognitive load as a design constraint and platform teams as the infrastructure that makes team autonomy sustainable. https://teamtopologies.com/book