Early AI gains prove fleeting without governance and clear leadership
Nearly half of UK business leaders report early productivity improvements from AI, but structural barriers are preventing most organizations from scaling beyond pilot projects, according to new research from Snowflake.
In a poll of 500 executives, 45% said AI is delivering tangible benefits, with 23% specifically citing productivity gains. The results suggest that after years of questions about AI's return on investment, businesses are finally seeing measurable results.
Yet optimism alone won't sustain progress. The research reveals that organizations face a critical gap between early wins and enterprise-wide adoption-one rooted in people, processes, and accountability rather than technology.
The real obstacles aren't technical
Only 19% of respondents identified technology as a primary barrier to AI adoption. The actual blockers tell a different story.
The top obstacles include:
- Lack of workforce skills
- Poor data quality
- Organizational silos
- Unclear leadership or strategic direction
This pattern mirrors broader industry trends. Google Cloud research in 2024 documented significant numbers of firms stuck in "AI pilot purgatory"-unable to move from proof of concept to production. The problem has worsened with agentic AI, where roughly half of projects remain in early testing phases, according to analysis from Dynatrace.
Governance fragmentation slows decisions
Many organizations distribute AI responsibility across multiple departments without designating a clear owner. Executive leadership typically controls investment decisions, but governance often remains undefined.
This fragmentation creates bottlenecks. Decision-making slows. Accountability dissolves. Teams operate in isolation rather than toward shared objectives.
Jennifer Belissent, principal data strategist at Snowflake, said the findings show that "belief alone is not enough" for successful AI adoption. She stressed three requirements: clear ownership, strong data foundations, and alignment between AI initiatives and measurable business goals.
What executives need to do now
Organizations planning to increase AI funding over the next 12 to 24 months should address fundamentals first. Fabian Stephany, economist at the Oxford Internet Institute, noted that technological breakthroughs typically require time for organizations to adapt workflows and governance structures-a process that cannot be rushed.
The research suggests that productivity gains depend on getting three things right: governance structures that assign clear responsibility, data quality sufficient for reliable outputs, and strategic alignment so teams understand how AI serves business objectives.
For executives, the message is straightforward. Early wins are real. But scaling those wins requires moving beyond enthusiasm to build the operational foundation that transforms pilots into business-critical systems.
Learn more about AI for Executives & Strategy and AI for Management to strengthen governance and leadership approaches in your organization.
Your membership also unlocks:
