AI Is Making Human Conversations with Customers More Valuable

For decades, the contact center has largely been managed as a cost center, reducing expenses wherever possible.

• How can we reduce the cost to serve?

• How do we handle more interactions with fewer resources?

• How much can we automate or move to a lower-cost labor model?

Those are still valid business questions. But I think artificial intelligence is creating a much more interesting one:

What happens when the conversations that still reach a person are the ones that matter most?

And it’s where I believe customer operations are evolving the most. And if you are sitting in the Chief Revenue Officer role, you should be paying attention.

AI Changes the Economics of the Human Interaction

Artificial intelligence is getting better at handling predictable, repeatable work. Routine questions, basic transactions, and information requests are increasingly good candidates for virtual agents and automation. That creates efficiency. But it also changes the mix of interactions that involve a customer.

Salesforce reports that AI resolved 30% of customer service cases in 2025. It projects that number will reach 50% by 2027.

Think about what that means. If technology handles more of the simple work, the customer who reaches a contact center agent may be calling with a complicated problem. They may be considering leaving. They may need help making an important decision. There may be an opportunity to retain the relationship, expand it, or create an advocate.

The human interaction becomes brand-building. That means we should not look at AI only through the lens of how much labor it can remove. We should also ask what we are going to do with the human capacity it creates.

Net is: the more routine work AI absorbs, the more valuable the remaining human conversations become.

Deflection is Not the Same Thing as Success

There is a temptation with any new technology to measure what is easiest to measure.

• How many contacts did we automate?

• How many calls did we prevent?

• How much did cost per interaction decline?

Those metrics matter. But they do not tell us whether we improved the customer experience (CX), much less the long-term relationship.

Most of us have experienced the other side of automation. You have a problem that does not fit neatly into the system, and you find yourself saying some version of “agent, agent, agent,” trying to reach someone who can actually help.

Customers are not necessarily rejecting AI. They are rejecting its dead ends.

Gartner recently found that half of customers believe generative AI can make customer service easier. At the same time, 87% said companies using generative AI for service must provide an option to reach a human agent. That distinction matters.

We should automate where efficiency creates value. We should preserve human interactions where judgment, trust, or relationship-building promotes empathetic connections.

That is especially important in complex financial decisions, healthcare, member benefits, and other situations where the customer may be dealing with something consequential.

I would not build an AI strategy around how much you can automate. I would frame it around how much friction you can remove. Find a friction point. Improve it. Evaluate the results. Then move to the next one.

The Next-Generation Agents Handle What AI Cannot

This shift also changes the talent equation. As AI increasingly takes on routine tasks, organizations will compete on the quality of the people handling the unpredictable work. Those interactions require a higher caliber of customer-service professional.

You need people with strong emotional intelligence and critical-thinking skills. They need to be patient, composed, and comfortable moving beyond a script. They need the judgment and authority to solve problems and the ability to build a genuine relationship with someone on the other end of the conversation.

At the same time, AI should make those people better.

Agent assist, better knowledge management, faster information retrieval, call summarization, and automation of after-call work can remove internal friction. The goal is not simply to do the same work with fewer people. It is to enable agents to resolve difficult customer needs faster and with more finesse.

That is an important distinction for leaders making technology investments.

AI should improve both sides of the equation: Automate interactions where automation makes sense, and make people more effective where the human interactions matter.

Walk the Virtual Floor

The other change I believe needs to happen is measurement.

Traditional contact center metrics still matter. Customer satisfaction, average speed of answer, abandonment, hold time and first-contact resolution all tell us something important. But they are no longer enough.

I have seen interactions receive excellent quality scores on paper and then listened to the actual call. The customer was transferred multiple times. They waited on hold. Their question was not resolved. Someone promised a callback that never happened.

The dashboard states the interaction was completed. The customer experienced something very different. That is why I tell leaders: You have to walk the virtual floor.

Listen to the interactions. Understand why customers are contacting you. Look at sentiment, conversational intelligence, and friction patterns alongside the operational metrics. Then, start connecting those signals with what customers actually do afterward.

• Do they renew?

• Do they buy again?

• Do they expand the relationship?

• Do they leave?

There is evidence that this connection deserves more attention. A recent peer-reviewed study of more than 19,000 banking customers found that higher customer satisfaction was associated with greater customer-level revenue growth, with the relationship persisting for as long as four years.

The next contact center dashboard should not stop at operational performance. It should help explain customer economics.

Customer Operations Belong in the Revenue Conversation

This is ultimately why I believe Chief Revenue Officers need to get more involved in customer operations.

I am not suggesting every contact center should report to the CRO. Organizational structures are different, and there are plenty of good reasons for customer operations to sit under a Chief Operating Officer or another executive. The reporting line is less important than transparency.

Revenue leaders should understand:

• What customers are saying.

• Where they are experiencing friction.

• Why they are contacting the organization.

• What those interactions tell them about retention, expansion, referrals, and lifetime value.

Historically, the Chief Revenue Officer role was heavily focused on generating pipeline, selling, and closing deals. That definition is expanding.

Revenue is also what happens after the sale. It is whether the customer stays, whether the relationship grows, and whether the experience creates an advocate for the business.

That is why I think we need to change the way we talk about this function. It is not a call center any longer. It should be thought of as a revenue center.

Organizations that figure out how to bring sales, marketing, and customer operations together to promote quality service, customer satisfaction, and business growth will have a distinct advantage.

Used wisely, artificial intelligence will accelerate that shift. Not because it makes people less important. Because it makes the right human interactions more valuable than ever.