Top Highlights
-
Agentic analytics is improving, but has a long way to go: 48% of insights now lead to business-improving actions, up from 37% last year—meaning most still don’t.
-
High-quality insights create an edge: “Insight leaders” turn more than 70% of agent insights into useful actions, especially benefiting from stronger prescriptive insights.
-
Context gaps and insight overload hold agents back: Poor business context and fragmented data hurt quality, while 52% of organizations say they receive too many insights to evaluate effectively.
-
Better foundations can boost results: Companies are focusing on stronger AI and data governance (60%) and adding semantic layers to make data more understandable to agents (45%).
Progress, with room to improve
Analytics agents can turn company data into useful guidance. Yet their value depends on whether people can act on that guidance—and whether it improves results.
At surveyed companies, 48% of agent-generated insights led to actions that improved the business, up from 37% a year earlier. That is meaningful progress. Still, more than half did not lead to a business-improving action.
Context separates stronger results
A small group of “insight leaders” reported better outcomes: More than 70% of their agents’ insights led to actions that improved the business. These organizations also rated insight quality higher, especially when agents recommended specific steps.
Context helps explain the gap. Some 55% of executives cited a lack of business context as a reason for poor-quality insights, while 45% pointed to fragmented or inaccessible data. Without the right background, an agent may spot a pattern but miss what it means for the business.
Quality requires focus and oversight
More insights do not always mean better decisions. At 52% of surveyed organizations, agents produce too many insights for teams to review and act on. Only 24% can evaluate every one; most review them case by case, a process that takes substantial human effort.
Organizations are responding by strengthening AI and data governance, cited by 60%, and adding a shared layer that gives data consistent meaning, cited by 45%. These steps can help agents produce more relevant guidance. However, companies still need people to check important recommendations and weigh potential risks before acting.
Discover More Technology Insights
Explore the future of technology with our detailed insights on Artificial Intelligence.
Explore past and present digital transformations on the Internet Archive.
AITechV1
