The McKinsey report shows: Real GenAI value is not created in pilots, but at scale. An isolated pilot project may demonstrate that a technology works, but it does not yet generate sustainable business value. A pilot often succeeds under lab-like conditions: with a dedicated expert team, clean data, and without the complexity of a real IT landscape.
- Adoption rising, impact focused: The share of companies using GenAI regularly is growing significantly, shifting impact into clearly identifiable business functions. Marketing & Sales (e.g., personalization, conversion lifts), Product Development (faster iterations, co-creation), Service Operations (self-service, case resolution times), and Software Engineering (code assistance, quality improvements) benefit most. Tangible effects appear as combined revenue and efficiency gains rather than isolated pilot successes.
- Rewiring as the strongest value driver: The study compares 25 organizational attributes and finds that systematically redesigning workflows—end-to-end, cross-functional, and data-driven—has the greatest impact on EBIT improvements. This includes standardizing use-case patterns, productized platform components (LLM-Ops/MLOps), robust data pipelines, and portfolio management that prioritizes use cases by business value. This turns PoCs into reusable, scalable AI operating models.
- Leadership, risk & skills in sync: C-level teams are increasingly using GenAI themselves and embedding accountability for target visions, governance, and risk controls. Success hinges on clear guardrails for security, IP/copyright, transparency, and quality, complemented by upskilling programs for business units and tech teams. Organizations with this “people-plus-governance” approach scale faster because adoption, compliance, and performance management come together.
Source: McKinsey & Company (March 2025): The State of AI: How organizations are rewiring to capture value. PDF.
