KPMG Report Finds Enterprise Disconnect Between AI and Its ROI

Why It Matters

The report highlights that traditional ROI metrics are failing to capture AI’s true impact, prompting firms to rethink investment criteria and risk‑adjusted budgeting in a rapidly maturing market.

Key Takeaways

• KPMG finds only 18% of AI leaders report clear ROI
• 75% of global executives will keep AI spending despite economic uncertainty
• UK firms say AI investment continues even without measurable returns
• Traditional ROI models struggle to capture AI’s time‑saving and decision‑speed benefits
• Only one in ten firms have talent and governance for AI returns

Pulse Analysis

KPMG’s latest Global AI Pulse Survey paints a nuanced picture of enterprise AI adoption. While the hype around generative and agentic AI continues to drive headline‑grabbing budgets, the data shows a stark performance divide: a small cohort of AI leaders are already quantifying value, whereas the majority remain in exploratory mode. This divergence is amplified by macro‑economic uncertainty, yet three‑quarters of senior executives pledge to maintain or increase AI spend, signaling confidence that AI is becoming a core strategic pillar rather than a discretionary expense.

The crux of the ROI dilemma lies in measurement. Traditional financial models demand clean input‑output relationships, a framework ill‑suited for AI’s intangible benefits such as faster decision cycles, talent augmentation, and risk mitigation. Analysts like Ben Grant argue that time reclaimed and decision speed are the real currencies, yet finance teams struggle to translate these into spreadsheet‑ready metrics. Consequently, firms are adopting hybrid evaluation approaches—combining pilot‑level productivity gains with longer‑term strategic positioning—to justify continued investment despite the absence of conventional payback periods.

For investors and corporate boards, the implications are twofold. First, capital allocation decisions must accommodate a broader set of performance indicators, recognizing that AI’s competitive advantage often manifests as a defensive moat against industry peers. Second, the talent and governance gap—only about one in ten enterprises possess the requisite expertise to harvest compounding AI returns—creates a market where early adopters can secure outsized gains. As AI matures from experimental labs to enterprise‑wide infrastructure, the pressure will intensify on organizations to develop robust measurement frameworks and talent pipelines, lest they fall behind in the next wave of digital transformation.