The belief that a solitary authority can keep product vision pure is a comforting myth that most scaling firms cling to. In practice, that authority becomes a bottleneck because every cross‑functional team must pause for a sign‑off before moving past the next milestone, and the sign‑off itself is delayed by competing priorities, unclear criteria, and the human tendency to defer difficult trade‑offs. The bottleneck multiplies when the gatekeeper’s inbox fills with requests from dozens of squads, each hoping to keep its velocity, while the gatekeeper’s own metrics reward thoroughness over speed.

Consider a midsize cloud‑services company that organized its engineering into autonomous pods, each with a product manager, designer, and two engineers. The CEO appointed a single “strategy council” to approve any feature that touched revenue streams. When a pod finished its design sprint, it queued a request to the council. The council, meeting only once a week, would often return the request with “needs more data” or “re‑align with roadmap”, forcing the pod to idle while re‑gathering evidence. Over time, the pods’ sprint retrospectives filled with the same lament: “we spent more time waiting than building.” The slowdown cascaded to sales, who could not promise delivery dates, and to customers, who saw longer wait times and began to drift to competitors.

The real problem is not the existence of a decision point but the mismatch between the decision’s frequency and the pace of work. When a gatekeeper’s capacity is lower than the sum of the teams’ output, each request adds a hidden delay that compounds across the organization, eroding the very agility the pods were meant to create.