Redefining the contact center agent for the AI era

Agentic AI is expected to autonomously resolve 80% of common customer service issues — like password resets, order-status questions, straightforward troubleshooting, basic routing — without human intervention by 2029. This also represents significant financial savings for customer service teams, with Gartner projecting a 30% reduction in operational costs. 

This also means that the calls that reach a human are disproportionately the ones that are ambiguous, emotionally heightened, or otherwise high-stakes. And since agents won’t be handling rote exchanges anymore, they’ll spend more time on other tasks, like wrangling the AI bots who are handling those exchanges.

Accordingly, Gartner also says more than 80% of customer service organizations plan to expand human agent responsibilities, with 84% adding new skills to agent profiles. That means longer training — and hiring for the type of agent who can handle more complex situations and varied responsibilities, not just the one who’s comfortable doing repetitive work and following a script. 

The contact center agent role is metamorphosizing, and contact center leaders need to plan for it now. 

But before we tackle how exactly to do that, let’s start with the key metrics contact centers care about in 2026.

What’s important to contact centers?

In a 2026 report from ContactBabel, contact service leaders rated "drivers for strategic contact center change" according to importance on a scale from 1 to 100 points. 

“Customer satisfaction score” received the highest average score across responses, with 71 points — followed by “new technologies” with 63 points, “agent attrition” with 59, and “reducing service costs” with 58.

Implementing AI seems like an apt response to “new technologies” and “reducing service costs” as drivers, since AI applications can save contact centers millions of dollars in labor expenses. ContactBabel estimates that the cost per automated support session (accomplished through self-service) is between 40 and 100 times cheaper than a live call to an agent.

As for customer satisfaction, contact centers often attempt to improve their CSAT (customer satisfaction) scores by improving first-contact resolution (FCR), ensuring customers can get their issues resolved without repeated calls, using CRM systems to provide agents with real-time information and context to reduce customer effort, and cutting down on wait times with self-service systems and intelligent routing.

As contact centers leverage AI to reduce headcount and labor costs while hiring agents capable of handling more complex, high-impact work, they can turn their attention to the third most important driver for change: attrition.

Agent turnover happens to be directly tied, in an inverse relationship, to customer satisfaction — the contact centers with the highest CSAT scores often have annual turnover rates of 10% or less, according to data from SQM Group reported in CX Dive. Experienced agents have developed both the knowledge and experience to handle customer calls both efficiently and with empathy.

Smaller workforces and more intentional hiring allow contact center leaders to focus on building both more sought-after contact center jobs and career ladders to support long-term hiring. They can begin this process by redefining the ideal profile for a contact center hire.

Empathy & critical thinking become more valuable

In 2025, 55% of decision-makers named empathy and listening skills as the single most valued agent characteristic, up from 26% in 2014. Meanwhile, the importance placed on handling repetitive tasks and multitasking has fallen (a possible sign that new technology systems are doing more of that work for agents).

This sounds counterintuitive because AI itself can produce polite, empathetic-sounding language. But once simple interactions are filtered out, human agents get a higher concentration of angry, confused, or anxious customers. 

About a quarter of people calling a contact center have already tried to answer their question through web self-service. A complex issue requiring a live agent is one of the two most important reasons customers escalate from web self-service, and 93% of surveyed contact center leaders felt customers wanted the reassurance a live agent brings.

Future agents won’t only need increased empathy — they’ll also need to have high critical thinking ability. eSkill consulted with a veteran contact center leader to get her thoughts on how AI will transform the skill profiles for contact center agents. A big theme in the discussion was going from butts in seats to brains in seats — i.e., hiring not for the ability to handle rote tasks and follow a script, but for greater problem solving ability. “Before, we would hire for customer centricity and empathy. Now you need to be smart and a critical thinker, because you're not going to be handling ‘When is my bill payment due,’ or 'I can't reboot my TV,’ anymore,” she said.

Upskilling takes center stage

While only about a third of contact center leaders say AI will replace agents, 100% agreed that AI would be used to assist agents in their work. Almost 80% of organizations also plan to transition some of their agents into new roles due to automation. Both training agents to use an AI copilot and transitioning them into new roles requires additional training and upskilling efforts from contact leaders.

While some contact center leaders are recruiting more experienced or more educated agents, most are trying to upgrade the existing workforce, according to research from ContactBabel. The majority are increasing training, particularly soft skills and, to a somewhat lesser degree, technical and product knowledge, and the “great majority” are increasing both hard- and soft-skill training/coaching to keep up with the changing nature of the job.

Where AI is used in contact centers

Hiring better people isn't enough for contact centers; agents also need more autonomy and better tools. When contact center leaders were asked for the single most important factor that would improve agent performance, 27% chose improved knowledge management, 24% a unified omnichannel desktop (providing agents with all the information they need on a single screen), and 17% agent empowerment. More training came in at 9%. 

Currently, technology systems as most contact centers are highly fragmented. 77% of respondents report problems with new agents familiarizing themselves with systems and 70% report problems with post-call work. Agents use an average of 3.5 applications during a call and 2.3 afterward. Right now, 88% of respondents say agents spend more than 10% of a call navigating applications and screens; ContactBabel estimates the industry-wide average at 24% of call time. Another 80% say agents have to duplicate or cut-and-paste information between systems. 

We know contact centers are deploying AI on the customer side for self service. But what about on the employee side? AI can not only assist customers, but it can help out agents as well: LLM tools can provide next-best-action recommendations, relevant knowledge, customer-history/context information and real-time coaching, which is particularly helpful for inexperienced agents or unfamiliar subjects.

The next-generation agent will increasingly be handling harder conversations alongside an AI copilot. Currently, 21% of contact centers use real-time in-call AI agent assistance, but another 72% plan to use it within two years. Call-summary generation is already used by 48%, transcription by 59%, and process initiation/form filling by 26%. 

AI is taking the “robot” out of the human, as one contact center agent quoted by AP puts it.

The new hiring scorecard

In aggregated historical data, “wrong type of person for the job” ranks as the top reported cause of attrition, even above low pay and lack of development. That seems to point to a longstanding problem with how contact centers are handling hiring. 

With today's contact center requiring a different type of hire than the traditional operation did — the winning agent profile is shifting away from traits like tolerance for repetitive tasks and rote multitasking and toward empathy, complex comprehension, initiative, judgment and the ability to use AI/knowledge tools effectively — this issue may become even more pronounced if contact center leaders don’t take a hard look at how they hire. 

Finding the right agent for the job is even more important as contact centers plan to reduce their workforce while investing more in their employees, with longer training and upskilling measures.

One way to address wrong-fit hiring is with competency and behavioral assessment during recruitment. For example, recruiters can look at empirically keyed biodata — one of the top predictors of job performance — by identifying the behaviors and characteristics of people already succeeding in each role, and pre-screening applicants against those criteria. It could also mean general interest surveys. It could mean simulations, where agents are asked to handle a hypothetical call with a “difficult” LLM rather than rote follow-the-script exchanges, in order to evaluate empathy and problem solving. And it could mean choosing learners and committing to their development.

The contact center of the future needs a different kind of agent. It's time for hiring strategies to catch up.