IDC: out of 2,473 organizations using AI in their operations, only 25% have a broad AI strategy, and most report some failures among their AI projects
Despite enterprises' best efforts, painless AI deployments appear destined to remain elusive for all but the most dogged adopters.
Context & Ripple Effects
IDC's 2019 snapshot lands just months after Gartner counted a 270% four-year surge in enterprise AI deployment — adoption was racing ahead of strategy. The new numbers expose the gap: among 2,473 organizations actually using AI, only a quarter have a broad strategy, and most admit to project failures.
The finding reads as the opening data point in a durability problem the coverage keeps confirming: McKinsey later found adoption plateauing between 50% and 60% of businesses, IBM's 2025 CEO survey found only 25% of AI initiatives delivering expected ROI, and MIT's report on 95% of GenAI pilots having little financial impact echoes IDC's failure finding six years on.
First-order effects
- The roughly 1,850 organizations in IDC's sample without a broad AI strategy are running projects piecemeal — and most are already absorbing failures among them, meaning budget is being spent on initiatives without an enterprise-wide frame to judge or rescue them.
Second-order effects
- Vendors and consultancies selling AI strategy and governance services gain a durable market: the same surveys that show near-universal experimentation (Gartner's deployment surge, later 61% of CIOs experimenting with AI agents) also show most buyers lack the internal discipline to scale, keeping demand for external methodology alive.
- Boards and CFOs, seeing failure rates corroborated by IBM's finding that only a quarter of initiatives deliver expected ROI, push procurement toward proven, narrow use cases over ambitious broad deployments.
Third-order effects
- If the pattern holds — strategy-less adoption, plateaued diffusion, pilot-heavy portfolios — enterprise AI consolidates around a small set of organizations that industrialize the practice, while the majority remain perpetual experimenters, widening a capability gap that survey vendors will keep quantifying.
- Persistent failure rates invite governance pressure: with only 35% of execs prioritizing transparent, accountable AI use in the related coverage, repeated project failures give regulators and customers a concrete basis to demand documented AI strategy and oversight.
The trend: Enterprise AI is settling into a long-term pattern where deployment outpaces strategy, leaving a persistent minority of organizations that convert AI spending into scaled, ROI-positive operations.