Pinnacle Performance Champions

Most customer service training spends a lot of time on two things: being friendly and knowing your stuff. Both matter. Customers certainly don’t want to deal with someone who is rude, indifferent or unable to answer a basic question. But there’s a third piece that may matter even more when it comes to solving the customer’s problem: understanding what the customer needs.

Qualtrics recently took a closer look at that in its 2026 Agent Effectiveness Benchmark, surveying 7,001 consumers across seven countries and seven industries about their most recent customer service experiences. The research measured three specific behaviors: friendliness, knowledge and understanding. What stood out wasn’t that one mattered and the others didn’t. It was what happened when an interaction failed to resolve the customer’s issue.

When an issue went unresolved, customers’ rating of the agent’s understanding fell 37 percent, from 4.44 to 2.78 on a five-point scale. Knowledge dropped 34 percent and friendliness dropped 20 percent. Overall satisfaction with unresolved interactions fell to 2.83. According to Qualtrics, understanding was the dimension most closely tied to whether the customer’s problem actually got resolved. That should make anyone responsible for customer service stop and think, because we’ve gotten pretty good at teaching the other two.

Friendliness is easy to recognize. Smile, be pleasant, use the customer’s name and show some enthusiasm. Don’t sound like the customer interrupted something more important. Knowledge is fairly easy to train and measure too. Employees need to understand the products, services, policies and systems necessary to do their jobs. When they don’t know something, they should know where to find it.

Understanding is different. An employee can be friendly and know everything there is to know about the business yet still miss what the customer is actually trying to accomplish. Customers don’t contact a business hoping to be impressed by how much the employee knows. They usually have something they need help with, and sometimes that need isn’t obvious.

A customer asking about price may really be trying to determine whether the more expensive option is worth it. Someone asking about a return policy may be concerned about making the wrong choice. A person asking when something can be delivered may have a deadline they haven’t mentioned. Simply answering the first question may be technically correct while still being less helpful than it could have been.

Qualtrics’ findings become even more interesting when artificial intelligence enters the picture. AI-handled interactions received one of the highest friendliness scores in the study at 4.41, but their understanding score was 4.08. That 0.33-point difference was the largest gap between friendliness and understanding for any channel measured.

There’s an important distinction there. We can program politeness and give an AI agent enormous amounts of information that can be made available almost instantly. But sounding helpful and actually understanding what someone needs are still two different things, and the same is true for people.

Don’t confuse answering the question with understanding the customer. Sometimes the answer really is all they need. There’s no reason to turn a simple request into an interrogation. If someone asks what time you close, tell them what time you close. But when the situation involves a decision, a problem, a complaint or anything where the right answer depends on the customer’s circumstances, a little discovery can completely change the quality of the interaction.

“What are you hoping to accomplish?”

“Tell me a little more about what happened.”

“What’s most important to you?”

“Is there something specific you’re concerned about?”

These questions are simple examples of getting beyond the initial request and understanding what actually matters to the person.

Complaints are a good example. When a customer is upset, employees often become so focused on explaining what happened or defending what the company did that they never identify what the customer is most upset about. It may be the money, but it could just as easily be the inconvenience, poor communication, a missed commitment or the feeling that nobody took the problem seriously. Until you understand that, you may be trying to solve the wrong problem.

Efficiency still matters. Customers don’t want a fifteen-minute conversation when a two-minute one will do. But efficiency should come from getting to the right answer more effectively, not simply getting through the interaction faster.

That distinction will become even more important as businesses automate more of the customer experience. Qualtrics recommends giving AI agents better context from previous experiences, preferences and purchase history, and making sure that context carries over when an interaction moves from AI to a human employee. Otherwise, the customer winds up explaining the same problem all over again, which is a pretty good way to demonstrate that nobody was listening in the first place.

Technology will continue to improve, and AI will almost certainly become better at understanding context. Human employees should be getting better at it too. Friendliness and knowledge matter, but you can be friendly while giving someone the wrong solution. You can know your products inside and out and recommend something that doesn’t fit the customer. You can provide a technically accurate answer and still leave someone feeling like you never really heard them.

The best customer service people understand that. They don’t rush to the answer before they understand the situation. They listen for what matters, ask when they’re not sure and use what they know to help the person in front of them.

Being nice helps. Knowing your stuff helps even more. But if you don’t understand the customer, there’s a pretty good chance none of it will matter.

Steve