While agentic AI holds significant potential to fully automate enterprise analytics workflows, current implementations are falling short of corporate expectations. According to a recently published survey, only 48% of insights generated by analytics agents result in actions that improve the business. Although this figure represents an increase from the 37% average reported the previous year, it indicates that more than half of the recommendations produced still fail to benefit businesses. While executives consider descriptive analysis successful, the outcomes of action-oriented prescriptive analysis receive decidedly mixed reviews.
The primary reasons for this inadequacy in insight quality prominently include a lack of business context. 55% of surveyed executives attribute poor results to this missing context, while 45% point to fragmented or inaccessible data. Furthermore, 52% of organizations state that agents generate an overwhelming number of insights, causing information overload that cannot be properly evaluated. Less than 24% of companies have the capacity to review all the recommendations produced, meaning the evaluation process remains highly human-intensive.
Despite these challenges, the report highlights a small cohort of companies, termed "insight leaders," that perform well above the average. Within this group, more than 70% of agent-generated insights translate into positive actions in business operations. Overall, surveyed organizations are taking steps to enhance the effectiveness of their agents; 60% of respondents prioritize strengthening AI and data governance, and 45% are focusing on adding a semantic layer to their data architectures.