Three years into the generative AI era, most enterprises are failing to convert growing investment and adoption into measurable business impact, according to new research from Forrester.

The report, Accelerate Your AI Voyage, draws on a survey of 1,500 AI decision-makers and identifies low AI fluency, uneven adoption, and an overemphasis on marginal productivity gains as key barriers to enterprise-scale impact.

"AI urgency is at an all-time high, but too many businesses are paralysed by a lack of understanding and siloed adoption," said Sharyn Leaver, chief research officer at Forrester. "CEOs have a narrow opportunity to shift the narrative."

A central finding is that many employees score poorly on Forrester's artificial intelligence quotient (AIQ) - its measure of AI aptitude - leaving organisations ill-equipped to deploy AI effectively. The report also identifies siloed adoption within individual business functions and difficulty measuring impact as compounding factors.

Data governance and infrastructure readiness emerge as critical differentiators. Almost half (47%) of high AI adopters engage consulting partners to prepare their data and systems, compared to just 26% of low adopters - a finding directly relevant to those tasked with maintaining data integrity.

Forrester outlines four areas organisations must address to unlock AI's full potential: defining business outcomes and success metrics; identifying use cases aligned to those outcomes; establishing a structured runway to test and time AI deployments; and scaling applications using cloud, frontier models, and embedded agents.

The research finds that high AI adopters are more likely to be led by a CEO-driven AI strategy (25%) anchored in customer experience outcomes. High adopters also prioritise customer experience (52% versus 44% for low adopters) and marketing optimisation (48% versus 30%).

Talent development is another dividing line. High AI adopters are more likely to include AI skill requirements in job descriptions (47% versus 33%) and require applicants to demonstrate those skills (54% versus 29%). Forrester recommends embedding AI skills through hiring, upskilling, and structured learning pathways.

"Businesses that prioritise customer-led AI experiences will ultimately build trust and long-term value. The window to outpace competitors is open, and those who act decisively will be best positioned to succeed," Leaver said.

Forrester warns that a wait-and-see approach to AI is no longer viable.