AI isn't the problem, change management is
For all the noise around AI breakthroughs, the truth is that most enterprise AI still fails to deliver. A recent MIT Media Lab report found that 95 percent of enterprise AI pilots never produce measurable business value.
That number isn’t a technology problem, it’s a leadership problem.
“The single biggest problem in the AI journey is it’s a 10 percent technology problem, a 20 percent data problem, and a 70 percent people problem,” said Tiger Tyagarajan, Sr. Advisor, BCG, Bain Capital and ex-CEO at Genpact on the Dialed In podcast.
Tyagarajan’s formula distills what many executives overlook: the hardest part of AI transformation isn’t the model or the math. It’s preparing an organization, its people, processes, and expectations, to work differently.
Technology isn’t broken, our systems are
Too often, companies assume that buying AI means buying progress. But even the smartest models fail inside rigid systems built for manual workflows.
As Wayne Butterfield, Partner at ISG explained, the real friction starts when old operating models meet new automation:
“Organizations don’t fail because the tech doesn’t work,” he said on the Dialed In podcast. “They fail because they’re still structured for humans to handle every edge case.”
Change management, not code, is the difference between a clever demo and a live deployment that scales.
Leadership defines AI success
AI doesn’t transform companies; leaders do. That’s why Tyagarajan frames AI adoption as a culture shift, not a technology project.
He described how the most effective organizations start small, test fast, and let the results speak for themselves. Success is not achieved by announcing a sweeping “AI transformation” that no one’s ready for.
“Start with a problem that matters to customers and employees,” he said. “Show impact quickly, then expand.”
This mindset echoes what John Walter, President of the Contact Center AI Association and founder of ProxyLink, shared about his own experience leading transformation efforts:
“If your executive team isn’t aligned on why you’re doing it and what success looks like, no amount of technology will save you.”
When leadership treats AI as a management exercise, not a marketing one, adoption accelerates naturally.
People are the real platform
Tyagarajan’s “70 percent people problem” isn’t a criticism; it’s a reminder that transformation is human work. Employees need to understand how automation will change their day, why it matters, and what it means for their roles.
That clarity turns fear into focus. As Butterfield put it, the organizations that win “make employees part of the design, not just the deployment.”
When people feel ownership, the technology follows.
From pilot to proof
The difference between companies that talk about AI and those that scale it isn’t luck, it’s discipline.
They plan for process change as rigorously as they plan for the technology itself. They budget for communication, training, and iteration. And they measure success in outcomes, not headlines.
At Replicant, we see that same pattern every day: the companies realizing the most value from automation are the ones that treat it as an organizational evolution, not a product launch.
The path forward
AI’s potential isn’t in question. What’s in question is whether enterprises will build the muscle to use it well.
“When you make change management the core of your AI strategy,” Tyagarajan said, “the technology finally starts delivering on its promise.”
At Replicant, we’ve seen firsthand that sustainable automation isn’t about having the flashiest AI, it’s about preparing your organization to use it wisely.
For leaders looking to chart their next step, our AI roadmap for contact centers outlines how to move from experimentation to enterprise impact, without the hype.
