## Overview AI agent capability now far exceeds enterprise adoption. Coding agents are powerful, but most workplaces still use simple chatbots because real work involves complex, messy data and entrenched workflows. The gap between what agents can do and what organizations actually use them for is the capability overhang. ## Core Framework **Three barriers to agent adoption:** 1. **Data messiness**: Real enterprise data is fragmented, poorly documented, and access-controlled -- agents need clean inputs that don't exist 2. **Workflow integration**: Connecting agents to the right information at the right step requires deep domain understanding of each organization's processes 3. **Change management**: Access control, trust, and organizational willingness to delegate decisions to agents **The overhang opportunity**: The biggest value creation lies not in building more capable agents, but in building the connective tissue (software, services, consulting) that helps organizations actually deploy agents against real work. ## Cross-Domain Applications 1. **Consulting positioning (S021)**: "Claude Code to Production" consulting directly addresses this overhang -- the demand signal from Upwork confirms organizations have the agents but can't deploy them 2. **Product design**: Products that reduce adoption friction (templates, workflows, access management) capture more value than products that increase raw capability 3. **Career strategy**: Being the person who bridges the capability-adoption gap is more valuable than being the person who builds more capable agents ## References - Aaron Levie (Box CEO) tweet, March 2026 (https://x.com/levie/status/2035913027401503002)