An MIT Technology Review Insights survey of 400 executives across eight industries reports that an average 48% of insights led to action, versus a recalled 37% a year earlier. Respondents cited missing context at 55%, fragmented information at 45% and information overload at 52%. This is self-reporting by executives at larger companies, not a direct measure of productivity across all businesses or proof of AI’s causal impact. It is more useful for identifying information and decision bottlenecks before deploying agents.
Back to Tech
Executive survey reports a 48% insight-to-action rate
Self-reported improvement is not causal evidence that agents increased productivity.