The early days of enterprise AI were all about doing things faster and cheaper. Cut some phone time here, automate a process there, save a few percentage points on costs. But something interesting has happened over the past year – companies have figured out how to actually make money with this technology.
Scott Likens, PWC’s Global Chief AI Engineer, has watched this shift happen across dozens of implementations worldwide. What he’s seeing now isn’t about incremental improvements. It’s about companies opening up entirely new ways to generate revenue.
The Big Shift in Enterprise AI Conversations
The questions executives are asking have changed completely. Instead of “How do we save money?” the conversation has shifted entirely toward AI revenue generation.
Scott explains:
I think [in] 2025, though, we’re seeing this pivot with the focus on return on investment. There’s been more of a pivot around how does this really affect my business? Let’s get past the efficiencies and the fun exciting things I can do with it and reinvent portions of what we do. Can I change my revenue and can I get to new markets? Can I create products faster?
This isn’t just theory. The numbers are real, and they’re substantial.
What the Numbers Actually Look Like
Take a major technology company that added AI to their customer contact center. The results were dramatic, with what Scott describes as: “cutting 25% of phone time, cutting transfers by 60%.”
But here’s the thing – those aren’t just cost savings. They’re capacity gains. Suddenly, the same team can handle more customers, take on more complex problems, and deliver better service. In healthcare, Scott describes oncology practices using AI to process unstructured data and: “driving 50% improved access to insights.”
More insights mean better patient outcomes, faster diagnoses, and the ability to serve more patients with the same resources. This demonstrates real AI business ROI beyond traditional cost-cutting measures.
Scott notes:
We’re seeing opportunities not only [to] be more efficient, but they’re revenue opportunities. You’re going to be able to get to more customers and help them.
Opening Up Previously Inaccessible Markets
The most interesting developments are happening at the edges—companies using AI to reach customers they never could have served profitably before. This represents genuine AI market expansion into previously inaccessible territories.
Scott explains:
I like to see [companies] going to more markets that you didn’t hit before because Gen AI is expanding your ability to reach into different languages and cultures.
Again this isn’t theoretical. Pharmaceutical companies are leveraging AI capabilities that Scott describes as: “the ability to translate to local languages and local regulatory schemes”
This is helping them accelerate drug development and expand into new markets. What used to take years of localization work can now happen in months.
These pattern repeats across industries: companies breaking into geographic markets, serving customer segments that were previously too expensive to reach, and creating products that weren’t economically viable before.
The Measurement Problem
Here’s where most companies get stuck. They’re still measuring AI success the old way; looking for cost reductions and efficiency gains. But those metrics miss the real value when it comes to how AI creates new revenue streams.
Scott recommends a different approach:
I say focus on the things that show growth, focus on customer satisfaction, focus on operational efficiencies, things you’re already measuring. When you insert AI, you’re able to see the difference. Generation of new revenue streams.
The better questions, according to Scott:
- Are you differentiating?
- Are you scaling?
- Are you creating new capabilities for your people?
Why Some Industries Moved Faster
One of the most surprising findings from Scott’s work was that the companies everyone expected to be slow adopters – heavily regulated industries like healthcare and financial services – actually moved faster than everyone else when implementing AI return on investment examples.
Scott observes:
When the wave first hit maybe in the last 18 months or so, one of the interesting things was the regulated industries actually moved faster than the non-regulated. They were out of the gate. I think they saw opportunities.
This creates an interesting competitive dynamic. While some companies are still debating implementation, others have already captured new revenue streams and built advantages that will be hard to catch up to.
The Workforce Multiplier Effect
The job displacement fears that dominated early AI discussions haven’t materialized. Instead, something more interesting is happening when it comes to business growth through AI implementation. PWC’s research shows what Scott describes as: “higher salaries for AI enabled workers”
and: “productivity going through the roof.”
When your workforce becomes more capable, you can take on bigger projects, serve more demanding clients, and charge premium rates. It’s a revenue multiplier that most ROI calculations miss.
AI’s Inherent Limitations
Scott is refreshingly honest about AI’s limitations:
It’s math, not magic. It only solves, especially GenAI, it only solves certain patterns really well.
The companies succeeding with AI aren’t the ones applying it everywhere. They’re the ones who understand exactly where it creates value and where it doesn’t.
Scott notes: “There are definitely constraints [as to] where AI actually adds value.”
The Next Wave
Looking ahead, Scott sees even bigger opportunities emerging, particularly in robotics and what he calls “embodied AI.” While widespread adoption is still on what Scott estimates as “probably more on the 5-year” timeline, early movers will capture outsized advantages.
The companies generating real revenue from AI today aren’t just improving their current operations. They’re building the foundation for business models that didn’t exist a couple of years ago.
