Enterprises report AI maturity but struggle to show ROI

Komprise has published survey findings that show a gap between enterprises' stated AI maturity and the returns they have measured from AI. Most respondents said their organisations were maturing or mature in AI.

The data management company surveyed 500 technical decision-makers, including IT managers, data engineers, directors of AI, chief information officers and chief technology officers. The results suggest unstructured data remains a major obstacle, with 93% to 96% of organisations reporting that most of their data is not ready for AI systems or lakehouses.

While 68% of respondents said their organisations were either maturing or mature in AI, only 28% on average said they had quantified return on investment from it. The figures suggest a mismatch between how businesses assess the progress of their AI programmes and the data foundations needed to support them.

Readiness was low across several measures. Only 7% of organisations said 70% or more of their data had been cleaned, prepared and classified for AI, while 57% said half or less was ready.

Preparation for lakehouse analysis appeared weaker still. Just 4% said more than 70% of their unstructured data was ready with schema for analysis in a lakehouse, and only 24% said more than half was prepared and available.

Data bottlenecks

Respondents identified finding the right data for a project and cleaning up redundant, obsolete and trivial data as the main readiness gaps, with nearly half citing those issues. Metadata extraction and enrichment followed at 42%.

More than 70% of organisations do not yet have advanced classification methods to identify stale data and help users find data suited to particular AI use cases. On priorities, 46% said making unstructured data available for queries in AI and lakehouses without moving it was more important than putting AI-ready infrastructure in place, selected by 34%.

Use of lakehouse and analytics platforms was spread across several vendors. Microsoft Fabric was the most widely used at 49%, followed by Google BigQuery at 32%, Amazon Redshift at 28%, Databricks at 27% and Snowflake at 22%.

ROI measures

Measured returns were strongest in customer service, where 38% said they had quantified ROI. Employee productivity and decision speed each stood at 28%, followed by revenue growth at 26% and operational savings at 21%.

Those figures are below the headline assessment of AI maturity reported elsewhere in the survey. Komprise defined "maturing" as beginning to establish standards and use AI more strategically, while "mature" referred to deployment across multiple interconnected departments with agentic workflows and a unified governance and security strategy.

Krishna Subramanian, Chief Operating Officer and Cofounder of Komprise, said the results reflected a disconnect between AI adoption and the data work needed to support it. "Our survey showed that although most enterprises are maturing concerning their adoption of AI, only 21%-26% of enterprises have quantified ROI on the critical areas of operations and revenue growth," Subramanian said.

He added that data quality remains difficult to address across large organisations with fragmented storage estates. "Enterprise IT and AI leaders recognize that data intelligence is lacking but the pathway to ensuring high data quality is not straightforward. It begins with visibility and analytics across storage silos, unstructured data classification and industry-specific metadata extraction," Subramanian said.