AI has won the adoption race. The return-on-investment race, however, is still wide open. Nearly nine in 10 organizations now report regular AI use, according to McKinsey’s 2025 global survey, yet nearly two-thirds have not begun scaling AI across the enterprise. Only 39 percent report an EBIT impact at the organizational level.

The spending is accelerating anyway. Deloitte’s 2025 research found that 74 percent of surveyed organizations had invested in AI or generative AI during the previous year, making it the most heavily funded technology capability in its study. Yet another Deloitte study confirmed that only six percent of respondents saw payback from a typical AI use within 12 months.

The gap exposes a problem that is becoming harder for executives to ignore. Having AI somewhere inside the business is increasingly easy, while making it materially improve the business is considerably harder. McKinsey’s findings point toward workflow redesign as a defining characteristic of organizations capturing more value, while its 2026 analysis argues that productivity improvements alone are unlikely to provide a lasting competitive advantage.

The temptation is understandable. Competitors are announcing AI initiatives, vendors are selling increasingly capable models, and executives are being asked to demonstrate progress expansively.

Wells Stringham recognizes the danger in treating that pressure as a technology-shopping exercise. “People are treating it like it’s a plug-and-play tool,” he says. The problem, in his view, begins with assuming AI can be purchased like another piece of enterprise software and ends with a license sitting on employees’ desktops without a meaningful change to how work gets done.

Stringham, a partner of experience at better&co, approaches the question from years spent working inside large-scale technology environments. At better&co, the team positions its work around moving important projects from strategy into execution, including AI-native product development and operational evolution. Its current focus includes agentic workflows, internal tooling, and data capabilities designed to alter how teams operate. Stringham believes that the starting point remains rooted in understanding the business problem first, then determining where AI can actually change the economics of solving it.

In his view, this difference matters because what used to be an effectively structured team may be counter to the needs of an AI-focused organization. A company can hand employees an AI assistant and see isolated productivity gains while leaving the larger workflow untouched. Stringham believes the more meaningful opportunity comes further down the maturity curve, where AI becomes embedded in the workflow itself, behind the interface, and changes the team’s time spent on the tasks.

He points to an accounting firm that better&co worked with as an example. During tax season, Stringham recalls how documents arrived simultaneously, requiring interns to manually sort files and identify the relevant tax year. The team helped build a system that automated the initial processing, flagged uncertain cases, and produced organized client files with confidence ratings. “Work that could consume almost a business day for some clients was reduced to 30 minutes. This gave accountants more time on the higher-value work of developing tax strategies,” he says.

There is no chatbot at the center of that story. The AI sits underneath the workflow. Employees interact with the system they already need to use, while the technology handles the repetitive processing in the background. The team sees this as a more important direction for enterprise AI than simply giving every employee another application to learn.

“Its not just about learning an AI chatbot,” he says. “It’s about it becoming a part of a daily toolset.”

The impact extends into how companies manage change. Stringham notes that technology projects have historically failed when teams were given a technically sound system that did not fit how they actually worked. AI raises the stakes because relatively small deployments can now produce major changes in how employees spend their time.

He cites IKEA’s use of AI as an example of how organizations can redirect human expertise once repetitive work is reduced. He explains that customer-support employees with deep product knowledge were able to contribute to new digital tools instead of remaining tied to routine service work. The same logic is shaping how Stringham sees better&co’s role in the current AI market.

Mid-market companies, he says, often have enough complexity to benefit substantially from AI but lack the internal capacity to investigate every emerging technology or rebuild a system from scratch. better&co’s experience working on large-scale technology projects is intended to help those organizations move from an attractive idea to something capable of operating in the real world.

The rise of “vibe coding” offers another warning. AI can now take an idea remarkably far in a short period, Stringham explains, but the initial development of a product is different from the final 20 percent required to make it secure and capable of handling enterprise demands. better&co has increasingly encountered technology leaders who have built prototypes with AI and then reached the point where the product needs deeper engineering expertise. Stringham calls the transition “vibe to viable,” highlighting that it’s the ultimate blocker to success.

He says, “To get from vibe to viable, you’re going from the quick stage of building without constraints to real-world requirements. That can be disheartening for a team that hasn’t navigated this roadblock before.”

McKinsey reports that most organizations remain in experimentation or pilot phases, even as AI becomes nearly ubiquitous. Stringham argues that the competitive advantage will accrue to companies willing to examine where work is actually getting stuck, to redesign the process with the constraint and the technology in mind.

“Wouldn’t you want to actually go after the larger opportunity rather than settling for the ability to say your company uses an AI chatbot?” he says. Whether technology changes what a business can do, rather than simply giving it something new to talk about, may be the greater consequential measure of purpose in the era of AI dominance.