Signal Efficiency metrics dominate AI agent measurement while revenue impact stays unquantified

Summary

Enterprises that measure AI agents today mostly count efficiency gains rather than business growth. In a CB Insights survey of 59 director-level or more senior respondents at organisations with 200 or more staff, productivity gains, time saved and cost savings were the most cited success indicators. Each of these leading measures was selected by roughly 58 to 63 percent of respondents. Revenue impact, by contrast, sits much lower on the list and remains hard for firms to put a number on. The pattern suggests that early agent programmes are justified by doing existing work faster and cheaper. Whether agents actually lift sales is still largely unmeasured.