Organizations pour billions into AI. Yet revenue growth lags. Efficiency surges. But bottom lines barely budge. Fortune spotlights this disconnect in a piece drawing from Deloitte’s latest survey. While 74% of firms chase AI-driven sales increases, just 20% deliver today. Productivity? That’s real—66% report gains. Decision-making? 60% see sharper insights. The catch: these wins hide in plain sight, far from P&L headlines.
Deloitte’s 2026 State of AI in the Enterprise report, based on 3,235 leaders across 24 countries, paints a clear picture. Worker access to sanctioned AI tools jumped 50% in a year, hitting 60%. Pilots proliferate. Production stalls. Zero percent of companies have 40% or more experiments live today. Expectations? Still zero in six months. Pilot fatigue sets in without governance. Infrastructure lags. Talent gaps yawn.
And budgets climb anyway. 84% boost AI spending. Transformative impact doubles to 25% from 12% last year. But reimagination? Only 34% create new products or reinvent processes. Most stick to surface tweaks—37%. Jobs untouched. Zero percent redesigned around AI strengths.
Consider Deloitte’s own Sidekick, an internal GenAI tool. Employees save two hours weekly. Time shifts to skills, creativity, relationships. No direct dollars. Pure capacity unlock. A manufacturer deploys AI agents for product development—cost down, market entry faster. An air carrier routes routine customer tasks to agents. Humans tackle complexity. These aren’t revenue bombs. They’re force multipliers.
Leaders chase the wrong numbers.
Financial returns demand time. Deloitte’s earlier AI ROI report—1,854 execs from Europe and the Middle East—shows why. 85% hiked AI spend last year. 91% plan more. Payback? Two to four years typical, versus seven to 12 months for standard tech. Just 6% recoup in under a year. Top 20%—AI ROI Leaders—treat AI as transformation, not bolt-on. They bet early on generative and agentic AI. Allocate over 10% of tech budgets. Redefine wins around revenue opportunities (49%) and model reinvention (45%).
Gartner echoes this. CFOs misjudge by forcing one ROI lens, says analyst Twisha Sharma. “AI does not follow one cost curve, and it does not produce one uniform type of value.” Build portfolios: routine automation, process tweaks, bold disruptions. Nonfinancial value hits first—agility. Decisions. Capacity. “CFOs need to account for that if they want a complete picture,” Sharma adds.
PwC’s data stings harder. 56% of CEOs see zero revenue or cost wins from AI last year. Winners—12%—embed deeply in decisions, demand gen. They rewire workflows, not just license tools.
So what works? Track the invisible. Reclaimed hours compound. Faster cycles beat rivals. Employees reskill into high-value roles. Agentic AI—23% use today—needs oversight. Just one in five govern it maturely. Physical AI hits 58% adoption, eyeing 80% in two years.
Barriers persist. Skills top the list. 53% educate for fluency. 48% upskill. Few redesign roles. Successful firms pair humans and AI—new jobs like operations managers emerge. Governance binds it. “Oversight is everyone’s role,” Deloitte notes.
Leaders fix this now. Audit pilots for activation gaps. Measure capacity shifts. Link efficiency to P&L via dashboards. Demand vendor P&Ls. Target workflows driving EBIT. The 20% who do? They pull ahead. The rest watch budgets evaporate.
AI isn’t failing. Measurement is. Fix that. Value follows.
WebProNews is an iEntry Publication