For the past two years, the standard answer to “why are we investing in AI?” was some version of “it makes our team more productive.” Saves hours. Automates repetitive tasks. Speeds up drafting and research. Businesses told that story to their boards, and it worked.

That story is changing.

Futurum Group’s 1H 2026 Enterprise Software Decision Maker Survey — polling 830 global IT decision-makers — found a significant shift in how enterprises now measure AI return on investment. Direct financial impact, meaning actual revenue growth and profitability improvement, has nearly doubled as the primary ROI metric, reaching 21.7% of responses. Meanwhile, productivity gains dropped from 23.8% to 18.0% as the top success measure.

This is not a subtle shift. Executives are no longer satisfied with time-savings narratives. They want to see the P&L move.

What the Data Shows

The survey found specific metrics that paint a clear picture of where enterprise AI stands in 2026.

The ROI story is maturing. Futurum split their previous “overall financial performance” metric into two distinct measures: top-line revenue growth (10.6%) and bottom-line profitability (11.1%). Together, these hard financial outcomes now dominate the value conversation in a way they did not a year ago.

Agentic AI is the fastest-growing priority. Autonomous agents surged 31.5% year-over-year as a top technology priority among enterprise decision-makers. When combining first and second-place rankings, agentic AI reached 39.3%, up from 32.0% in the second half of 2025. This is not a niche interest any more — it is mainstream enterprise strategy.

Buyers now expect AI to come via agents. 38.8% of enterprise buyers say they expect AI capabilities to be delivered primarily through agents. And 45.7% rank AI capabilities as their top software selection criterion — meaning the era of AI as a differentiator is ending, and it is becoming table stakes.

Deployment focus is on core operations. Where are enterprises planning to deploy agentic AI? Cybersecurity leads at 58.7%, followed by sales, marketing, and service at 51.3%, and supply chain at 47.8%. These are not experimental projects. These are core business functions.

The Productivity Paradox

The data reveals an uncomfortable truth worth naming: there is a large gap between individual productivity gains and enterprise-wide financial outcomes.

Research from McKinsey’s 2026 Global AI Survey found that knowledge workers using production AI agents recover a median of 6.4 hours per week per seat. That is real. But 97% of executives report benefiting from AI while only 29% see significant organisation-wide ROI. The productivity is happening — it is just not translating into the numbers that matter to a CFO.

Bain’s Agentic AI Benchmark 2026 adds context. Median payback periods run 4.1 months for customer service deployments, 6.7 months for marketing operations, and 9.3 months for engineering. These are solid returns — but they require proper measurement, clear baselines, and deployments that are deep enough to actually affect business outcomes.

The companies failing to see financial ROI are usually the ones that deployed AI as a productivity layer and stopped there. They got faster employees. They did not redesign the workflows, reduce headcount in low-value areas, or capture the output as revenue or measurable cost reduction. The tool worked. The strategy did not.

The Governance Gap

One detail from the Futurum data deserves attention: only 1 in 5 companies has a mature model for governing autonomous AI agents.

This matters because the shift to agentic AI is happening faster than governance frameworks are being built. Agents do not just assist humans — they take actions, make decisions, and operate autonomously across systems. Without clear governance, enterprises are exposed to data risks, compliance failures, and the reputational cost of agents that act incorrectly.

The companies moving fastest on agentic deployment are also the ones most at risk of getting governance wrong. This is not an argument for slowing down — it is an argument for treating governance as a first-class priority from day one.

What This Means for Business

If you are evaluating AI investments right now, this data points to a few clear actions.

Stop measuring AI in hours saved. Your board wants to see revenue impact and cost reduction. Build your AI business case around those metrics from day one. What deals will close faster? What overhead will decrease? What new service capacity do you unlock?

Agentic AI is no longer optional. The 31.5% surge in agentic AI as a priority reflects a real shift in what enterprises are deploying. Static AI tools that assist but do not act are being replaced by agents that execute end-to-end workflows. If your AI strategy is still built around copilots and assistants, you are behind the curve.

Governance needs to be built in, not bolted on. The 1-in-5 governance maturity figure is a warning sign for the whole industry. Before deploying agents at scale, organisations need clear policies on what agents can and cannot do autonomously, how decisions are logged and audited, and how errors are caught and corrected.

The gap between leaders and laggards is widening. Organisations that are already measuring AI in P&L terms, deploying agents in core functions, and building governance frameworks are compounding their advantage every quarter. That gap will not close on its own.

The productivity era of enterprise AI served its purpose. It proved the technology works. The financial era is now, and it demands a different kind of strategy — one built around outcomes, not activity.

---

Enterprise DNA helps business leaders move from AI experimentation to AI that drives real financial results. If your organisation is ready to evaluate what agentic deployment could mean for your bottom line, book a discovery call.