Posted by Erica Marois, Head of Content and Community on 27 July 2026
Everyone's Talking About AI. Here's Where Business Value Actually Comes From.
When we began asking enterprise leaders about the AI initiatives that had delivered the greatest business impact, we assumed the conversations would veer toward models, copilots, or the latest wave of agentic AI. Instead, interview after interview brought up workflow diagrams, team structure, cross-functional collaboration, data readiness, and the difficult work of deciding which business problems deserved AI in the first place.
By the time we'd spoken with leaders across healthcare, financial services, manufacturing, software, telecommunications, insurance, HR, and beyond, a consistent story had emerged. The organizations creating the most business value spent more time understanding how the work actually happened before trying to transform it.
The Shift from Pilots to Progress
In the early days of enterprise AI, organizations raced to launch pilots, employees discovered new ways to use generative AI in their daily work, and vendors competed to embed AI into nearly every product imaginable. Back then, the question was, Can we do this?
Today, executives are asking something much harder to answer:
Did any of it actually move the business forward?
It's easy to find sweeping claims about AI transforming business. It's much harder to find concrete examples with measurable effects. We set out to uncover them, and the examples shared were remarkably diverse. A legal team cut contract turnaround from days to hours. Financial advisors reclaimed time to spend with clients instead of preparing presentations. A healthcare organization transformed a manual adjudication process that once consumed hours. A marketing team quadrupled content production by giving writers back time to think strategically.
On the surface, these organizations had almost nothing in common. Yet the use cases show a consistent approach: they resisted the temptation to begin with the technology.
Instead, they began with work. Specifically, they began by understanding which work created value and which work quietly consumed it.
Slow Down to Speed Up Results
Ben Prescott, Head of AI Solutions at Trace3, has repeatedly watched organizations fall into the same trap. Companies become excited about what's possible, but cut corners during the critical planning process. Leaders move quickly from recognizing an opportunity to evaluating tools without spending enough time defining success, mapping the current workflow, or agreeing on the business problem they're trying to solve.
Prescott doesn't argue that organizations are moving too quickly with AI. He argues they're moving too quickly to AI.
His team takes a more deliberate approach. Rather than pulling use cases from a generic playbook, they begin by mapping how work flows through the business today, identifying where friction exists, where success can be measured, and where AI is most likely to improve a meaningful business outcome. Only then do they design a solution.
That approach recently helped a global manufacturer surface hundreds of potential AI opportunities before narrowing the field to 10 that correspond with strategic business objectives. One of those—contract redlining—reduced turnaround time from three to five business days to roughly twenty-four hours while improving consistency and risk management.
Across industries, the organizations making the most meaningful progress appeared to arrive at the same question:
Where is work breaking down, and how can AI help?
Instead of beginning with a tool and searching for a problem, these leaders begin with an operational challenge that already has executive attention, employee buy-in, and a measurable cost. AI becomes one possible solution rather than the starting point.
“With a technology as exciting as AI, we tend to laser-focus on the technology itself. The immediate reaction to a need or ask is to go find a tool or build an application right away, moving quickly into a pilot,” says Prescott. “But I often come back to a saying from my military days – 'slow is smooth, and smooth is fast.” The same is true for scaling AI across an organization. Most of the challenges I see show up at the first mile (strategy and use case prioritization) or the last mile (user adoption, training, and value measurement). Organizations that skip past either usually end up stuck in pilot mode, no matter how good the underlying technology is.”
Michael Murphy, Principal at Adaptovate, has seen the same mistake play out across organizations. Leaders roll out copilots or enterprise licenses first, assuming productivity gains will naturally follow. But he argues that "tools first instead of workflows first" almost always leads to disappointing results because success was never defined around a business outcome.
Murphy recently worked with a mid-sized global financial services organization that wanted to increase the capacity of its talent acquisition team following a series of acquisitions. The initial instinct was familiar: deploy AI tools as quickly as possible. Instead, his team paused to map the recruiting workflow, identify bottlenecks, and determine which tasks were both repetitive and high effort. Only after redesigning the process did they introduce AI. The result was a 2.5x increase in hiring throughput from job posting to accepted offer.
"The tools will come. Before that, we need to hack your process,” he says.
Amplifying Expertise, Not Replacing It
If understanding work is where successful AI initiatives begin, protecting expertise is where they create their greatest value.
The highest-return AI initiatives shared one defining characteristic: they were designed to remove the repetitive, administrative, and cognitively draining work that prevents those employees from doing what only they can do.
Brandon Metcalf, founder and CEO of Asymbl, captured the idea in a single question:
"Don't ask, 'Where can we cut headcount?' Ask, 'Where do my best people spend time on work that doesn't require their best thinking?'"
At Asymbl, that philosophy shaped the company's use of what it calls digital workers—AI teammates assigned clearly defined responsibilities, human managers, and measurable performance expectations. Rather than replacing recruiters, one digital worker assumed responsibility for high-volume screening, scheduling, and outreach while human recruiters focused on pipeline strategy, culture fit, and the conversations where judgment mattered most. The recruiting team ultimately filled 100 roles in 100 days after processing more than 17,000 applications.
Murphy saw the same principle play out inside a recruiting team he advised. The goal was to eliminate scheduling, resume screening, and other administrative work so recruiters could spend more time doing the parts of the job only people can do: building relationships with candidates, partnering with hiring managers, and making thoughtful hiring decisions.
That balanced approach also surfaced in the financial sector.
A wealth management firm partnered with Incedo to help financial advisors prepare personalized planning recommendations. Advisors had been spending up to two hours translating detailed planning outputs into language clients could easily understand. AI now prepares an editable, client-ready version in roughly 10 to 15 minutes, allowing advisors to spend more time on the work clients actually value: building trust, answering difficult questions, and helping families make important financial decisions.
Financial research offered another compelling example. Rather than asking highly compensated analysts to spend their mornings combing through hundreds of pages of SEC filings and earnings transcripts, one organization built an AI-powered research pipeline that delivers a fully cited morning briefing before the workday begins. Research that once took four hours now takes about fifteen minutes—not because the analysts became unnecessary, but because they can devote their attention to interpretation rather than gathering information.
Asymbl’s CMO Lauren Esposito said it best:
"Don't count the outputs. Count the time returned to your best people—and what they do with it."
The Key Takeaway for Leaders: AI Isn’t the Real Story
One thing is clear: AI isn’t just an experimental initiative but a full-blown business transformation. That may be the biggest disconnect in today's AI conversation.
Outside the boardroom, the focus is on the next breakthrough model. Inside the enterprise, the leaders creating measurable business value are asking different questions. Which workflows deserve to change? Where is expertise being wasted? How do we build an organization that gets better every time AI capabilities improve?
That shift is already changing how some leaders think about AI itself. Nitin Seth, CEO of Incedo, argues that AI-native leaders have shifted their focus from pilots to building reusable capabilities that compound in value with every new use case.
And after dozens of conversations with executives like Seth, one conclusion feels difficult to ignore.
AI's greatest impact may be less about reducing headcount than redefining roles. As repetitive work shifts to AI, the work left behind becomes more human: building relationships, exercising judgment, solving complex problems, and creating strategy.
The technology will continue to evolve. The discipline required to generate lasting business value probably won't.
Editor’s note: This is just the beginning. In our next issue, Rewired reporter Jessica Levco goes inside five enterprise AI initiatives providing measurable ROI at mid-size and large organizations. We'll explore the business problems they set out to solve, how they measured success, and the lessons leaders can apply in their own organizations. Subscribe so you don't miss it.