AI returns $3.70 per dollar. Most organizations get less than $1.

McKinsey's 2025 State of AI report is the most important dataset in enterprise AI. It confirms three things that should fundamentally change how you approach your AI investment.

First: the returns are real. Companies that get AI right are seeing $3.70 returned for every dollar invested. Revenue increases from generative AI grew significantly between early and mid-2024. The top performers are generating $10.30 per dollar. This is not a technology experiment. It is a commercial advantage.

Second: most companies are not getting it. The average enterprise return sits below $1 per dollar invested. The gap between the top quartile and the median is not closing. It is widening. The firms that got production AI right in 2023 and 2024 are compounding their advantage. The firms still in pilot are falling further behind.

Third: the gap is not about budget or model selection. The report is unambiguous: companies generating real AI value are redesigning workflows and strengthening governance. They are not deploying better models. They are building better architecture, one that is deterministic, integrated, observable, and protected against model drift.

The practical implication: the question your board should be asking is not "are we using AI?" It is "are we in the top quartile, or are we subsidizing the firms that are?"

Weaver's dual-vector ROI model is built on this benchmark dataset. We use McKinsey's own numbers to show you where your specific workflows sit relative to the organizations generating $3.70 per dollar, and what it would take to get there.