Real-Time Data Leaders Triple Their AI Success Rates As Enterprises Struggle To Connect Agents To Live Data

Enterprises that have reached maturity in their real-time data practices are 3.4 times more likely to deliver measurable outcomes from AI projects than those at the earliest stage of adoption, according to new research from IDC.

The findings come from IDC’s Real-Time Data Study, a survey of 623 senior technology decision-makers at enterprises with more than US$1 billion in revenue across eight countries. The study, commissioned by real-time data platform vendor, Solace, was fielded in June 2026.

IDC scored each respondent on a 0-to-100 real-time data maturity scale – blending real-time deployment depth, agentic AI adoption and measurable business outcomes – then grouped them into four tiers: Emerging (bottom 30 per cent), Developing (30th to 60th percentile), Advanced (60th to 80th) and Leaders (top 20 per cent).

The gap between the top and bottom tiers was consistent across multiple measures. Among Leaders, 89 per cent reported delivering measurable outcomes on half or more of their AI projects. For Emerging organisations, that figure was 26 per cent.

Agentic AI has become an operational dependency

The study found that 80 per cent of respondents are either investing in or have already deployed agentic AI, and 90 per cent have increased their real-time data focus to support it. A further 85 per cent described real-time data as mission-critical and 84 per cent called it strategically important.

Among Leaders, 59 per cent have already deployed several AI agents into production, compared with 20 per cent of Emerging organisations. And while 47 per cent of Emerging organisations are investing significantly in agentic AI but have not yet deployed it, that figure drops to 31 per cent for Leaders, suggesting they have already moved past the investment phase and into execution.

Agentic AI was cited as the number-one force driving demand for real-time data, selected by 43 per cent of respondents. Security and AI trust requirements came second at 40 per cent, followed by cost and scalability improvements at 38 per cent, regulatory or compliance pressures at 35 per cent and industry-specific disruptions at 29 per cent.

The bottleneck is connecting agents to data

When asked about barriers to scaling AI agents, 40 per cent of respondents identified connecting agents to real-time enterprise data as a top obstacle, tied with managing agents across different frameworks.

Monitoring agent performance and accuracy was cited by 37 per cent, getting agents to work together as a team by 35 per cent and human-in-the-loop oversight by 33 per cent. Concerns over large language model costs (28 per cent) and the ease of building agents (26 per cent) ranked lower.

On the technical side, data quality and consistency emerged as the leading challenge for real-time data initiatives over the next 12 to 18 months, selected by 46 per cent of respondents. Latency and performance constraints followed at 40 per cent, governance, security and permissions at 39 per cent, legacy systems and technical debt at 38 per cent and tool sprawl and platform complexity at 37 per cent.

Solace President and CEO, Denis King, argued that the data pipeline rather than the AI model itself is now the binding constraint.

“As enterprises move AI efforts from pilot to production, the constraint isn’t the quality of the agents or the models – it’s the accuracy and timeliness of the context, i.e. the data, they have access to,” King wrote in the report.

“The barrier to success IDC ranks first in the study is something we hear in nearly every conversation: difficulty connecting agents, in real-time, to trustworthy information from across their enterprise.”

Leaders run unified platforms while others fight sprawl

The study identified platform architecture as one of the clearest dividing lines between maturity tiers.

Among Leaders, 71 per cent run a single unified real-time data platform, compared with just 8 per cent of Emerging organisations – an 8.5-times gap. For full technology integration, the figures were 36 per cent for Leaders and 10 per cent for Emerging organisations.

Similarly, 76 per cent of Leaders have implemented real-time data across most or all of their workflows, compared with 7 per cent of Emerging organisations. And 91 per cent of Leaders are automating half or more of their real-time processes, against 23 per cent at the Emerging tier.

IDC concluded that while belief in real-time data is widespread, execution is not. Leaders have operationalised real-time data across the enterprise, while less mature peers remain confined to a handful of workflows.

Company size is not a differentiator

One of the study’s more notable findings is that real-time data maturity does not scale with company size. Maturity scores ranged from 50 to 53 across all company size bands, from organisations with 500 employees through to those with more than 10,000 staff. AI success rates were similarly flat, ranging from 46 to 51 per cent.

Industry differences were also modest. Healthcare and education tied for the highest maturity score at 57, followed by software and information services at 55, manufacturing at 53, financial services at 52 and retail at 48. AI success rates by industry ranged from 47 to 58 per cent.

The research concluded that maturity is determined by operating practices and organisational approach rather than by budget or headcount.

Talent is organised differently at the top

The study found that Leaders are more likely to embed real-time data talent directly into business teams. Among Leaders, 34 per cent have embedded real-time data specialists in business units, compared with 21 per cent of Emerging organisations.

Conversely, 60 per cent of Emerging organisations have talent distributed but not embedded, compared with 37 per cent of Leaders. IDC characterised this as a shift from loosely coordinated technical support toward placing real-time data skills within the teams that depend on them.

Conviction grows with experience

The more an organisation has progressed in its real-time data journey, the more strongly it agrees that AI agents depend on live data. Among Leaders, 89 per cent agreed that agents need up-to-date, accurate real-time data, compared with 75 per cent of Emerging organisations.

The gap was sharper on agent autonomy: 92 per cent of Leaders agreed that agents must make autonomous decisions, compared with 69 per cent of Emerging organisations. And 93 per cent of Leaders agreed that agents must react instantly to events, against 75 per cent at the Emerging tier.

Optimistic and sceptical at the same time

The research also revealed a tension in how enterprises view agentic AI. While 80 per cent agreed that agentic AI is changing the way they work and 79 per cent agreed it is making the world a better place, 71 per cent simultaneously described it as overhyped.

Despite the scepticism, investment continues to climb. The study found that 94 to 97 per cent of respondents plan to increase their spending on real-time data and agentic AI. About nine in 10 reported that their real-time data investments have met or exceeded expectations for both business outcomes and performance metrics.

Among Leaders, that satisfaction rate reached 97 to 98 per cent.

Data quality tops the near-term agenda

Looking ahead 12 to 18 months, organisations ranked improving data and analytics quality as their top strategic objective for real-time data, selected by 33 per cent. Supporting generative AI and agentic AI use cases followed at 27 per cent, alongside improving data analytics efficiencies at 27 per cent.

Improving process automation, improving app-to-app integration and reducing the time from data to decision were each cited by 26 per cent.

Five priorities for closing the gap

IDC identified five priorities that separate leading organisations from the rest: treating real-time event-driven data as core AI infrastructure rather than an analytics afterthought; consolidating fragmented tooling onto a unified platform; reframing governance as an accelerator rather than a constraint; connecting agents to real-time enterprise context; and investing ahead of the curve.

The researchers characterised the gap between Leaders and the rest as closable, noting that the market is already moving.

The IDC InfoBrief was authored by IDC Research Manager of Event-Driven Automation and Analytics, Carlos González, and IDC Research Director, Robert Burbach. Respondents were drawn from industries led by manufacturing (n = 223), financial services (n = 155) and retail (n = 90), with 78 per cent serving as primary decision-makers for real-time data initiatives and 71 per cent owning AI or agentic initiatives.

Last Updated on September 11, 2026 by Nick Ross