Singapore – Solace, a real-time data platform provider, has released a new IDC InfoBrief examining how real-time data practices are shaping the outcomes of enterprise artificial intelligence projects.
The InfoBrief draws on a survey of 623 senior technology decision-makers at enterprises with revenues exceeding $1 billion, spanning eight countries. According to the study, agentic AI has moved from experimental use to operational dependency, reshaping how quickly enterprise data needs to move.
Organisations identified as leaders in real-time data maturity are, according to the findings, more than three times as likely to deliver measurable business results across the majority of their AI projects.
Meanwhile, these leaders reported average annual gains of roughly 23% in areas such as speed to market, risk reduction, and the ability to sense and respond to change.
Commenting on the findings, Paul Fitzpatrick, Chief Marketing and Business Development Officer at Solace, said, “Every enterprise has the same AI models. The winners are the ones whose agents run on trusted, real-time data, and the numbers back it up. When data moves in real-time, AI delivers. When it doesn’t, AI fails to create sustainable impact.”
Turning to the drivers behind this shift, the study found that real-time data and agentic AI are increasingly interlinked. Eighty per cent of respondents said they are investing in or already running AI agents, while 90% reported an increased focus on real-time data specifically to support their agentic AI plans.
When asked what is fuelling the growing need for real-time data, respondents named advances in agentic AI as the single biggest force reshaping their data priorities, ranking it ahead of security, cost, and regulatory pressure.
However, the research also points to a persistent obstacle. Forty per cent of organisations reported that connecting AI agents in real-time to reliable enterprise data ranks among the top barriers to deployment, rather than the challenge of building the agents themselves.
Data quality and consistency emerged as the leading technical challenge overall, cited by 46% of respondents, including organisations considered the most advanced in their real-time data adoption.
Notably, the study found that real-time readiness and AI success do not correlate strongly with company size, and vary only modestly across industries. Instead, what separates leaders from other organisations appears to be operating practice rather than budget or headcount.
Specifically, 76% of leaders said they run real-time data across most or all of their operations, compared with just 7% among organisations classified as emerging. Leaders were also more likely to standardise on a single unified platform, rather than relying on a patchwork of loosely connected tools, and to embed real-time data talent directly within the teams that use it.
The InfoBrief further suggests that returns compound as organisations mature. As enterprises progress through the study’s maturity tiers, both the number of agents in production and overall AI project success rise in tandem.
Among leaders, the share of agents deployed in production climbs to 59%, up from 20% at emerging organisations, while the share of AI projects delivering measurable outcomes rises to 67%, up from 35%.
Despite the challenges identified, satisfaction with real-time data investment remains high. Ninety per cent of organisations surveyed said their real-time data investments have met or exceeded expectations.
Looking ahead, the study indicates that investment momentum is set to continue, with between 94% and 97% of organisations planning to increase spending on both real-time data and agentic AI.
Carlos M. González, Research Manager for Event-Driven Automation and Analytics at IDC, framed the findings as evidence of a broader shift in enterprise priorities. “The survey shows real-time data moving from a technical concern to a strategic one across the board”, he said.
“Organisations are citing agentic AI as the single biggest force driving their need for real-time data. That reordering of priorities tells you real-time data is becoming foundational infrastructure for AI, not an analytics feature”, González concluded.
