When Spotify first exported its autonomous squad model, each eight‑person squad owned a slice of the product and could ship daily without waiting for anyone else. The promise was clear: give every team full decision rights over its backlog and let speed explode. What fell apart was the next layer up – the tribe, a collection of dozens of squads sharing a broad domain. Tribe leads, eager to keep every squad aligned, began imposing weekly sync meetings, shared roadmaps, and a de‑facto approval gate for any cross‑squad change. The very mechanism meant to preserve coherence turned into a coordination tax that slowed releases across the whole domain.

The root cause was a mismatch between decision rights and information flow: squads had the authority to act, but tribes hoarded the knowledge of inter‑dependencies and forced squads to seek permission before touching shared APIs. This “information hoarding” created a latent queue where every change waited for a tribe meeting, and the queue grew as more squads were added. Because the bottleneck sat at the tribe level, adding more squads only amplified the delay, turning the original speed advantage into a systemic drag.

Spotify responded by creating chapter leads who owned the shared technical standards and by delegating API versioning decisions back to the squads, while tribe leads shifted to a purely advisory role. The change restored the original contract – squads could act, and the tribe merely surfaced risks after the fact rather than gating work. The result was a resurgence of rapid iteration across the domain, proving that alignment must travel with authority, not against it.

The lesson extends beyond music streaming: any organization that stacks autonomous teams under a larger unit must ensure that the larger unit’s information channels do not become a de‑facto gate. If the higher layer hoards the map while the lower layer holds the compass, speed stalls regardless of how many hands are rowing.