By Rick DeLisi, lead research analyst for Digital Customer Service at Glia. Glia were finalists in the ‘Best Use of AI in Customer Service‘ award at The 2025 AI Awards.

AI is quickly becoming a fixture in most contact centers. Yet for many, the full value of AI isn’t being fully realized. The unfiltered truth: this isn’t a failure of the technology, but a maturity problem in the way companies are implementing AI.

New research from Glia and Metric Sherpa shows AI adoption has surged among fintechs, SaaS providers, and financial institutions, but the operating models and strategies that turn AI into value have not kept pace. Three-quarters of organizations already use at least some form of AI in their contact centers, and 83% plan to continue expanding usage. But most still measure success with pre‑AI yardsticks and hand its governance to teams that aren’t directly connected to customers.

Six main friction points holding up full ROI

What’s causing this AI “maturity gap”? According to the research, the reasons are numerous.

1. Critical measurement misses

Leaders want AI to be valuable to their customers and employees. They also want it to drive bottom line results. But there’s a problem: Many leaders aren’t consistently or actively measuring the impact AI is creating.

• 90% say customer value is important, but only 61% measure it consistently.
• 85% say employee value is important, but only 60% measure it.
• 85% say economic value is important, but only 49% measure it.
• 86% say strategic value is important, but only 28% measure it.

This last gap in particular is perhaps the biggest miss of all.

Strategic value—i.e., using customer interactions to generate insights that influence product, marketing, pricing, or positioning—is precisely where AI can shine by surfacing patterns often missed by humans. But because so few teams are measuring AI’s strategic impact, this transformative opportunity remains largely underexplored.

2. Adoption outpaces adaptation

Leaders have moved fast: 75% are using or scaling contact center AI. The trouble is, strategic adaptation hasn’t kept pace. Too many programs deploy tools without closely considering roles, metrics, ownership, and governance. That’s how AI ends up as bolt-on automation rather than an operating model change. The consequences are tangible: shallow wins and suboptimal scaling.

3. Disconnected systems hurt

Since one of the biggest goals of AI implementation is efficiency, it’s a shame that call center operations are inherently inefficient. Disconnected systems, poor knowledge access, and high employee effort are creating barriers for agents. In fact, a third of leaders say agents’ struggles to harness AI effectively are hindering the technology’s full ROI potential. Streamlining workflows, surfacing knowledge, and minimizing handoffs allow AI to do what it does best—and lets people focus on providing a personal touch in situations where a human connection is critical to customer loyalty.

4. The contact center isn’t leading AI projects

AI is changing how work gets done, yet only 17% say their contact center—the function closest to customer value—leads AI initiatives. When IT or operational leaders work independently, programs often interject their own priorities over broader strategic outcomes such as loyalty, conversion, and first-contact resolution.

5. Savings aren’t being reinvested strategically

AI is capable of creating massive efficiencies, but it’s what organizations do with that newfound extra capacity that truly matters. Applying savings straight to the bottom line can drive positive results for the short-term, but the organizations that pull ahead in the long run are those that put their savings to work. The most successful are converting time saved into value created—raising hiring standards, investing in employee development, redeploying staff to proactive activities, and more.

6. Strategic intelligence is overlooked

Contact centers are the beating heard of the most comprehensive product feedback a company will ever get—yet nearly 40% of leaders underestimate customer-based strategic intelligence, and only about a quarter measure it. That’s a missed growth lever. When AI organizes this trove of valuable data, interactions can inform product fixes, marketing clarity, and even fraud or risk decisions.

How to achieve full AI maturity

Luckily, there’s good news. For each of these friction points, there’s a practical fix that can accelerate AI maturity in the contact center—and convert potential into value that is both measurable and measured. Here are six.

1. Ensure business alignment

Make sure AI is actually serving business outcomes. For every use case, specify whether the goal is experience, efficiency, risk, growth or some combination—and design requirements, workflows, and handoffs to serve that end.

2. Break down bottlenecks

AI thrives in clean workflows—but disconnected systems, hard-to-find knowledge, manual wrap-up, and inconsistent authentication hinder its effectiveness. As organizations remove friction, both efficiency and experience rise together—leading to fewer transfers, faster resolution, and less rework.

3. De-silo your systems

Coordinate voice and digital so AI and humans operate as one system—one with shared context, smooth handoffs, and clear ownership. Customers shouldn’t have to start over, switch channels or escalate.

4. Make AI governance inclusive

Create a cross-functional forum—contact center/CX, IT, operations, compliance, data, etc.—to decide use cases, model choices, data handling, and risk thresholds.

5. Center human enablement—not replacement

Automate routine tasks so people can do what they do best: complex problem-solving, providing empathy, onboarding new team members, and building customer relationships. Communicate the talent strategy upfront. When teams know AI is there to extend capacity and elevate roles, adoption rises and outcomes improve.

6. Review your metrics

Retire single-threaded KPIs as your sole north star. Set measurement baselines, review on a fixed cadence, and course-correct regularly. What gets measured—and discussed—gets improved. And the implementation of AI enables companies to have access to every keystroke and every word that every customer has ever shared with your company.

“Good enough” automation will not move the needle

The research tells a clear story. AI is no longer optional in customer interactions, and the bar for quality service is rising. Organizations must focus on developing the maturity necessary to derive full value from AI in their contact centers.

Importantly, maturity isn’t just about technical achievements—it’s first and foremost about management priorities. When companies modernize their metrics, realign governance, and reduce internal friction, they unlock a compounding loop: better experiences, higher-performing teams, clearer strategy signals, and stronger unit economics.

With this loop in place, AI-powered contact centers are no longer cost centers—they become engines of customer loyalty and strategic insight.