Customer experience (CX) doesn't have a technology problem with artificial intelligence; it has a people problem. As we move from tactical AI projects to broader adoption, it's time to call human resources (HR) and think seriously about not just work redesign but work repricing for the humans in the contact center.
CX leaders who were skeptical about AI are a vanishing breed. Valoir's recent study of more than 150 contact center and customer service managers found that more than 95 percent of organizations run at least one AI capability today. Although fewer than 8 percent of organizations have autonomous agents in production, most have applied AI in a combination of areas, including knowledge search, case summarization, and copilots. As AI agents take on more call volume and more complex cases, the question becomes what happens to the humans left standing?
We've found that it's neither the mass SaaSpocalypse some predicted nor business as usual, and it's extremely uneven across industries and sectors. Nearly 60 percent of organizations have no plans to reduce customer service headcount because of AI. We found that many of them are not retaining staff because AI isn't having an impact; they're clearing backlogged queues and easing resource constraints or redeploying humans to more complex work or more proactive roles like customer success.
However, on the other side of the coin, 19 percent have already cut agent jobs because of AI, and another 22 percent plan to do so within the next year. That means that roughly four out of 10 organizations are actively reducing or about to reduce their customer service and contact center workforce now.
The 22 percent of companies that plan to reduce headcount deserves particular attention because of what it represents. These aren't organizations scrambling to respond to disruption; they're making workforce planning decisions as a consequence of AI deployment decisions already being made.
Whether you're reducing headcount or not, the job that remains isn't the same job. We found that 34 pertcent of organizations are revising agent skill requirements and wage bands because of AI, and 30 percent are changing key performance metrics. As AI absorbs routine, low-complexity interactions, humans are being pushed toward higher-skill work, like complex problem resolution, emotional support, and exception handling–the kind of interactions that require judgment that AI can't exercise. A role redefined around higher-complexity work should be compensated and managed differently than one built around high-volume, scripted call handling.
The problem is that in the rush to get AI working, we've often slapped AI on top of human effort without thinking end-to-end about how we redesign contact center work and how we reprice that work to reflect the higher-level skills it will now require to be effective. We found only 20 percent of organizations are changing their compensation structures accordingly. That gap—34 percent of firms redefining customer service job descriptions vs. 20 percent changing how the job is paid—should make contact center leaders and human resources (HR) departments take notice. Although some of the gap is likely a natural lag between redefining a role and formally repricing it, a persistent lag becomes a retention problem, and attrition in a smaller, more skill-dependent workforce is more expensive to fix than it would have been in a larger, less skilled pool.
So, what should CX and HR leaders do?
First, treat job and compensation redesign as inseparable from AI deployment planning, not as a follow-on HR exercise. If the job is being redefined around complexity and judgment, the pay structure should be moving on the same timeline. Keep in mind that onboarding will need to change as well.
Second, we need to build new performance and management frameworks that blend human and machine contribution rather than slapping old contact center metrics onto a hybrid AI/human workforce. It's time to drop the metrics, like adherence and average handle time, that don't make sense in an AI context, and to take advantage of AI to do things like ongoing monitoring of customer and human agent sentiment, effort, and outcomes.
Third, we need to be honest with the workforce about the future we're building. The organization that will be most successful through the AI transition in CX will communicate expectations clearly. If a smaller team doing more high-value work is paid accordingly, we can save on traditional workforce engagement tactics and techniques and, likely, on attrition as well.
As AI adoption broadens, the technology is seldom the constraint. It's the human parts that are the most difficult to manage. The organizational discipline to match pay and performance measurement to the work AI has actually changed will be important for both agent and customer experience.
Rebecca Wettemann is founder and CEO of Valoir.