McKinsey consultants argue that AI-powered development requires fundamental changes to agile workflows and team structures. Survey data from 300 enterprises shows top performers achieve 5-6x faster delivery by adopting AI-native workflows (spec-driven development, continuous planning), smaller pods (3-5 people vs 8-10), and consolidated roles where engineers orchestrate agents rather than write code directly. Key bottlenecks include work allocation challenges, manual code review overhead, and tech debt amplification. Success requires comprehensive change management: hands-on upskilling, new incentive structures, and measurement systems tracking outcomes beyond adoption metrics. Case studies demonstrate 51% increase in code mergers and 60x agent consumption when teams reorganize around AI capabilities.
Sort: