Agentic AI Widens Enterprise Performance Gap

Key insights

1. The gap between frontier and average AI users has widened to 8.3x, driven by the adoption of agentic workflows rather than model access. This indicates that strategic implementation, not just technology, determines AI value.

   Adoption Metrics →

   Impact: Companies ignoring agentic strategies risk a significant and growing productivity deficit compared to competitors.

2. Non-technical roles such as legal, sales, and marketing are adopting AI agents at rates far exceeding engineering. This challenges the notion that AI is primarily a technical tool and highlights its broad operational impact.

   Workforce Dynamics →

   Impact: Business leaders must prioritize AI training for non-technical staff to unlock hidden productivity gains across the organization.

3. Agentic usage now accounts for 64% of enterprise output tokens, shifting the focus from content generation to system execution and maintenance. This represents a fundamental change in how work is structured and performed.

   Workflow Transformation →

   Impact: Organizations must redesign workflows to accommodate autonomous agents, moving from human-in-the-loop to human-on-the-loop models.

4. Frontier firms have significantly higher adoption rates of plugins and skills, with 19% of users employing skills compared to 3% at typical firms. This tooling gap is a key driver of the performance divergence.

   Tooling Strategy →

   Impact: Investing in agent integration layers and reusable skills is critical for maximizing AI ROI and closing the performance gap.

5. Institutional inertia and cognitive habits are the primary barriers to full AI adoption, even among advanced users. Overcoming this requires explicit training and cultural shifts, not just tool access.

   Change Management →

   Impact: Successful AI transformation requires a focus on behavioral change and workflow redesign, not just technology deployment.

Action items

• Audit current AI usage to identify the gap between chat-based and agentic workflows. Prioritize migrating high-value, repetitive tasks from chat to autonomous agents.

  Impact: This shift can unlock significant productivity gains by freeing up human time for higher-value strategic work.

• Implement training programs focused on agentic workflows for non-technical departments, particularly legal, sales, and marketing. Provide specific use cases and templates for agent deployment.

  Impact: Targeted training can accelerate adoption in high-growth areas, capturing the 100x+ usage increases seen in frontier firms.

• Invest in agent integration layers, including plugins and reusable skills, to enable agents to interact with existing systems. Focus on creating standardized workflows that agents can execute autonomously.

  Impact: Advanced tooling is the primary differentiator for frontier firms, and investing here can help close the 8.3x performance gap.

• Redesign organizational workflows to accommodate autonomous agents, moving from human-in-the-loop to human-on-the-loop models. Define clear roles for agents in system maintenance and execution.

  Impact: Workflow redesign ensures that agents are integrated into core operations, maximizing their impact on system-level performance.

• Address institutional inertia by creating cultural incentives for AI adoption. Recognize and reward employees who successfully implement agentic workflows, and provide support for those struggling to break legacy habits.

  Impact: Cultural change is essential for overcoming the cognitive barriers to adoption, ensuring that the organization can fully leverage its AI investments.

Quotes