Domino finds a two-year ROI plateau: production capability rose to 93 percent, yet 57 percent still see returns lag spend - and many scale agents without governance.
Via PR Newswire: AI ROI Fails to Outpace Spend for 57 percent of Enterprises, Unchanged Since 2025, Even as 93 percent Now Report Improved Production
Domino Data Lab fifth annual survey of 639 senior enterprise AI leaders, released via PR Newswire, finds 57 percent still see ROI fail to outpace spend - unchanged since 2025 - even as 93 percent report improved production capability in 2026 (up from 88 percent in 2025).
That is a two-year plateau: models get into production, money still does not clearly win. Domino also flags a last-mile gap between models in production and business users who need to act on what those models found.
For owner-led SMEs, translate the enterprise chart into one operating rule: shipping a model is not delivering value, and agent pilots without governance are how the plateau becomes your problem.
Asked how business users access AI-generated insights, Domino reports fragmentation: 34 percent say access methods vary by business unit - the single most common answer - and 40 percent still rely on at least one fully mediated path such as a scheduled report or an analyst request.
Agentic AI tops 2026 priorities, tied with upskilling business users at 38.5 percent, ahead of other investment types including governance infrastructure. Yet Domino finds 43 percent have agentic AI in governed production while 41 percent are piloting (12 percent) or scaling (29 percent) agentic AI without governance to manage it - with active scalers outnumbering mere pilots by more than two to one.
Governance maturity divides outcomes. Among organizations whose governance fully keeps pace, 67.5 percent have agentic AI in governed production, versus 17.2 percent where governance only partially keeps pace - about 3.9 times as likely. Fully integrated governance correlates with 75 percent reporting significantly improved AI delivery velocity, versus 23 percent where governance falls behind.
The survey covered Director-and-above leaders at firms with USD 100 million-plus revenue across North America, UK, and continental Europe in April 2026. SME owners are smaller, but the pattern travels: production theater without a delivery path and without guardrails is expensive distraction.
Ignore the vanity metric of models deployed. Measure whether a named user can act, under policy, at business speed.
Vendor-reported survey figures are directional. Verify against your own P and L before you buy another platform story.
Domino survey of 639 leaders: 57 percent still see ROI fail to outpace spend, unchanged since 2025, even as 93 percent report improved production (PR Newswire).
This story maps first to a Workflow ROI Audit: separate work that saves money from production theater that only looks busy.
It also maps to Managed AI Operations. If agents are already scaling, someone must monitor, govern, and keep delivery tied to business users - not leave last-mile gaps as someone else problem.
AgentsROI stays vendor-neutral. Domino sells a platform to regulated enterprises. We sell operating judgment for owner-led SMEs: measure ROI, close the last mile, and govern what runs.
If your team can show models in production but cannot show ROI outpacing spend, book a Workflow ROI Audit before the next agent rollout expands the gap.
This article summarizes publicly reported information and is for general informational purposes only. It does not constitute legal, tax, financial, investment, security, or compliance advice. AgentsROI.ai is not a law firm, accounting firm, or registered investment adviser. Facts, pricing, statistics, and product capabilities cited here reflect the sources listed at the time of writing and may change. Readers should verify current information independently and consult qualified professionals regarding obligations specific to their industry, jurisdiction, and circumstances - including applicable New York State and New York City requirements. AgentsROI.ai may have commercial relationships with vendors mentioned; where material, such relationships are disclosed. Nothing in this article is an endorsement of any specific AI product, model, or provider.
