# Half of AI agent insights fail to drive business value

> A new report finds that fewer than half of insights generated by analytics agents lead to business improvement, with poor context and data fragmentation as key culprits.
- Title: Half of AI agent insights fail to drive business value
- Summary: A new report finds that fewer than half of insights generated by analytics agents lead to business improvement, with poor context and data fragmentation as…
- Keywords: agentic analytics, generative ai, enterprise analytics, ai agents, data quality, business insights, ai governance, technology, science, Half, agent, insights
- Source: MIT Technology Review — https://technologyreview.com/2026/11/02/1147660/raising-the-quality-of-insights-in-agentic-analytics
- Read time: 1 min
- Topics: agentic analytics, generative ai, enterprise analytics, ai agents, data quality, business insights, ai governance, technology, science
## AI agents promise to transform analytics
Generative AI has made analytics accessible across organizations. Agentic AI promises to automate entire workflows and take actions based on agent insights, but quality remains a serious problem.
## Less than half of agent insights drive improvement
On average, forty-eight percent of insights generated by analytics agents at surveyed companies lead to actions that improve the business. That is up from thirty-seven percent one year ago.
## Top performers achieve much better results
At a small cohort of organizations termed insight leaders, more than seventy percent of agent insights lead to actions that improve the business. These leaders give their agents' insights higher marks overall.
## Lack of context is the main barrier
When analytics agents deliver substandard insights, the main reason is lack of business context, cited by fifty-five percent of surveyed executives. Fragmented or inaccessible data is the other major factor, according to forty-five percent.
## Too many insights overwhelm human reviewers
At a majority, fifty-two percent, of surveyed organizations, analytics agents generate too many insights to be properly evaluated and acted upon. Less than one-quarter, twenty-four percent, can evaluate all agent insights.
## Companies focus on governance and semantics
Surveyed companies are prioritizing initiatives to strengthen AI and data governance, cited by sixty percent. Forty-five percent are adding a semantic layer to their data architecture to improve agent performance.
## Key takeaway

Fewer than half of agent-generated insights at surveyed organizations lead to business improvement, with insufficient context and fragmented data identified as main barriers.