The finding
In March 2026, KPMG asked roughly 2,100 senior leaders whether AI was delivering meaningful business value. Nearly two-thirds said yes. In its Q2 wave, fielded April to May and published in June, KPMG asked a separate sample of 2,145 leaders a narrower question: had the organization established ROI? Seven percent had. That rose to 14% where accountability was clearly defined, particularly at CEO level, and sat at 4% where accountability had not been assigned.
BCG puts a number on the same point. Surveying 11,749 workers across fourteen markets, it found that a clear strategy lifts measurable business impact by 25 percentage points, while better tools without that strategy move it by roughly 5.
The context the headlines usually miss
Nearly every AI study in circulation surveyed large-enterprise executives. The research on smaller organizations reads differently. Ipsos surveyed 750 owners and operators of businesses with 500 or fewer employees for the US Chamber of Commerce Foundation. Among those already using AI, 54% report a positive impact on how long work takes, 50% on employees taking on more challenging work, and 47% on the quality of the work itself. Three to four percent report a negative impact on any measure, within a credibility interval of ±4.4 points.
The US Census Bureau puts AI use at 32% among firms with 100 to 249 employees, and finds that 57% of organizations using AI apply it in three or fewer business functions. Very few companies at this size have gone deep. The distance between a first win and a real advantage is shorter than the headlines suggest.
The four opportunities
- Most organizations have a mandate to use AI. Few have a target. Among operations leaders, 22% report an AI strategy that is both developed and implemented. (Grant Thornton)
- Your next step: one paragraph in writing naming the outcome expected this quarter, the workflow it applies to, and who owns the result. (Strategic Alignment dimension)
- Deloitte found that 48% of organizations introduced AI without redesigning the workflows or roles it sits within. The work is already changing. The question is whether anyone is designing that change.
- Your next step: pick one high-impact process still running on manual effort. Give it a named owner, a documented before state, and a date ninety days out when it gets measured. If the redesign does not give those specific people hours back inside the ninety days, the work competes with their day job and loses to the first real deadline. (Workflow Clarity dimension)
- Seventy-eight percent of organizations lack strong confidence they could pass an independent AI governance audit within 90 days. Three in four boards have approved major AI investments, and just over half have set governance expectations alongside them. (Grant Thornton)
- Your next step: answer three questions in writing before AI touches customer data or anything customer-facing. What is the AI doing? What is it deciding without human review? And what happens when it is wrong? (Technical Readiness dimension, Operational Readiness dimension)
- Only 37% of employees have received any organizational guidance on AI use, and 14% have been offered training. (Google/Ipsos) Smaller companies are nearly three times less likely to offer it than large ones. People are not unready. They have mostly been handed a login.
- Your next step: name the two or three people already experimenting, give them one shared workflow problem and a few weeks, and make the result visible to everyone else. (Talent & Ownership dimension)
What separates the two groups
None of the four are technology failures. They are questions of direction, structure, accountability, and ownership. Choosing a tool and changing how work gets done are two different efforts, and the second is organizational work: deciding what should change, agreeing on what better looks like, and carrying people through it. Treating it as a continuation of the first is the most common reason progress stalls.
The organizations pulling ahead are not spending more. They are working in a different order: understanding what AI can realistically do in their own operation, identifying where it is most likely to produce a measurable result, and adopting it in one workflow before expanding.
Prepared by Forjic Technology. Sources: Grant Thornton 2026 AI Impact Survey (n=950) · Deloitte AI Institute, Enterprise AI trends in 2026 · BCG AI at Work, fourth edition · KPMG Global AI Pulse, Q1 and Q2 2026 · US Chamber of Commerce Foundation and Ipsos, What Small Business Owners Say AI Is Doing at Work (±4.4) · US Chamber of Commerce Foundation, Google and Hiring Our Heroes, AI in Action · Ipsos and Google, AI Works for America Poll · US Census Bureau.