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.