A new, AI-inclusive org design for CX

Key takeaways

• AI agents are joining the team, but jobs aren’t going away

• In customer support, human agents are getting new roles, leaving Tier 1 issues for AI

• Success metrics are shifting from handle time to customer happiness

It’s no surprise that AI agents are joining the workforce. They consume and organize information quickly, they follow clear instructions, and they’re getting better at empathy every day. They’re great to have on the team, and when it comes to customer experience, they’re not cutting into headcount—they’re changing the whole organizational structure.

"AI is not just a technology shift," says Venki Subramanian, SVP of product management at Freshworks. "It's a fundamental change in how customers interact with you and how your customer service agents do their work." 

According to Gartner, by 2027, half of organizations that planned to reduce customer service headcount through AI will abandon those plans. That’s because the work left for humans is more complex and more consequential than what it replaced. You don't cut staff when the remaining work is harder. You prepare them for it.

For CX leaders, this starts with rethinking what happens with your human agents when AI absorbs the volume—every tier shifts upward, and the management layer that once supervised high-volume transactional work is evolving with it, too.

Look before you leap

“You should not be restructuring your teams in the abstract,” says Isabelle Zdatny, head of Thought Leadership at Qualtrics XM Institute. “Before you can figure out how AI changes your team, understand what your team is responsible for today.”

Zdatny says leaders should rethink four key areas to identify the division of labor between AI and humans:

• Insights and analytics: Collecting customer feedback, tracking scores, producing reports, building dashboards

• Experience design: Journey mapping, human-centered design, closed-loop processes delivered through customer service and customer support

• Change management: Systematically understanding how to improve customer experiences and conveying to the entire organization how to act on and deliver those experiences consistently

• Strategy and vision: Defining brand promises, priorities, and customer experience principles

Zdatny’s research has found that most companies overinvest in the first bucket (analytics) and underinvest in the other three. That imbalance made sense when insights were manual and labor-intensive. But AI is going to disproportionately automate the insights and analytics function. “So, the other three—design, change, and strategy—become more important, not less,” she says. Because they’re more nuanced, they’re better in human hands.

Human roles shift

In customer support, a human-AI customer service structure starts with a shift in roles, which have traditionally been categorized in part by query types: Tier 1 handles volume — frequent, low-complexity contacts. Tier 2 and above handle the harder cases requiring judgment, institutional knowledge, or emotional attunement. But AI agents are increasingly good at Tier 1 support. 

According to Freshworks’ Customer Service Benchmark Report 2025, AI-powered conversational support now resolves up to 98% of issues on first contact, and in the retail sector in particular, AI agents now handle up to 53% of customer queries entirely on their own without escalating them to human support. 

As AI agents have taken on more Tier 1 queries, 43% of companies in the report have seen productivity gains. Every human agent takes on a new role — Tier 2 and above, able to handle more complex queries. 

inMusic, a global music technology company, created a “customer champion” role as a result of more automation on its support desk.

"The technology got to the point where we were able to take agents away from the front lines and restructure," says Bill Waller, head of customer service operations. "Our KPIs got to where they needed to be, and then the question became 'what's the next step?'" The customer champion role was that next step. Freed from basic queries, agents gained time for knowledge creation and advanced problem-solving, and a new “customer champion” focused on quality analysis and continuous improvement. 

At Broad River Retail, a former content manager whose role had been focused on building and maintaining the knowledge base used by both human and AI agents was elevated to a managerial role overseeing the performance of the AI agents, checking whether responses aligned with what human agents were telling customers, flagging breakdowns. “It was a mind shift,” says Wes Dudley, Broad River’s VP of customer experience, who created the role for his team.

New metrics for managers

Identifying roles and moving folks around is one thing, managing the new structure—and having digital employees for the first time—is new ground for leaders as well. It calls for new ways of tracking performance, and new skills to manage employees who aren’t human.

Supervising a high-volume transactional team is largely about monitoring: first response time, handle time, resolution rates, and such. But when human agents concentrate in judgment-intensive escalations — the billing dispute with a longtime customer, the complaint where tone matters as much as resolution — handle time is the wrong thing to optimize for, according to one CX leader.

He tracks CSAT instead. It resonates across functions, reflects whether human interactions are actually delivering, and doesn't penalize agents for taking the time a difficult case requires. Response time and handling time become "a little bit redundant" when AI handles the volume, he says.

Managing a hybrid team means having humans accountable for AI performance, not just AI deployment.

The organizations that get this right won't just have efficient AI. They'll have human agents sharper at the work that matters most, managers who know how to develop them, and AI colleagues that keep improving because someone is accountable for making sure they do.