The state of the marketing stack, from primary sources only.

McKinsey's 2026 global survey finds 88% of respondents report regular AI use, while nearly two-thirds of organizations have not begun scaling AI across the enterprise. Deloitte's Tech Trends 2026 reads: 11% of organizations have AI agents in production, against 38% piloting; 42% are still developing an agent strategy and 35% have none. IBM's 2025 CEO Study (2,000 CEOs, 33 countries) finds only 25% of AI initiatives delivered their expected return and only 16% scaled enterprise-wide. Kyndryl's 2026 People Readiness Report (1,100 leaders, 8 countries) finds 57% now have AI embedded or deployed broadly, up from 35% a year earlier, yet only 32% have achieved even one of their top two AI objectives.

Two circulating numbers failed the source check: an '89% agent failure rate' attributed to Deloitte is a reframe of its 11%-in-production figure, and a claim that 'only 11% hit their top-two AI objectives' misquotes Kyndryl's 32%. The stat roundups that dominate this topic quote each other; this page cites the institutions and retires what does not survive checking. Gartner's June 2025 warning counts only about 130 of the thousands of agentic-AI vendors as real, predicts over 40% of agentic-AI projects cancelled by end of 2027, and names the pattern 'agent washing' — causes that map onto missing disciplines: no owned number, no working memory, no written standard that can refuse. The seven disciplines are on the practice page; grade your own stack with Stack Score.