The 68-Point Gap

Writer surveyed 2,400 enterprise workers this month. Ninety-seven percent of executives said they have AI agents deployed. Only 29 percent reported meaningful ROI. A 68-point spread between having the thing and getting value from it.

The money side is just as stark. Fifty-nine percent of companies now spend over a million a year on AI. The budgets cleared. The returns did not.

What separates the 29 percent is not better models or bigger contracts. It is something less interesting: they redesigned how their teams actually work before they deployed. Chow Tai Fook mapped over 400 agents to specific decision points across 24,000 employees — 70 percent efficiency gain. That did not come from a smarter model. It came from understanding the workflow first.

The other 68 percent plugged agents into existing processes and waited for impact. They are still waiting. And the gap compounds. The companies seeing returns are reinvesting into more refined deployments. They have data on what works. The companies still searching have cost, complexity, and mounting pressure to show results — which pushes them toward more deployment rather than better deployment.

The question that matters is not which AI to use. It is what changed about how your team works since you deployed. If the answer is nothing, you are accumulating cost while your competitors accumulate operational intelligence.