Emotion is the most underused signal in customer experience

Contact centres have spent decades optimising for efficiency - shorter queues, faster handling times, better resolution rates and lower cost per interaction. These metrics matter, but in my experience, they can miss the thing customers remember most: how the interaction made them feel.

 

A customer can get through quickly and still leave frustrated, or they can have their issue resolved and still feel unheard. As AI becomes more embedded in customer service, this gap matters more, because poor emotional understanding is now one of the clearest signs of a bad automated experience.

 

And as Jamil Zaki, a professor of psychology at Stanford University, recently wrote in Harvard Business Review, “if the last decade in customer relationships was about mastering digital convenience, the next will be about restoring human connection at scale.” 

 

The next evolution of customer experience (CX) will come from treating emotion as a live operational signal, rather than something to analyse after the customer has already left. To do that, organisations first need to understand why sentiment alone doesn’t go far enough.

 

 

Sentiment is too blunt for modern CX

Many organisations already use sentiment analysis, but sentiment is often too limited to guide action. A positive, neutral, or negative score can help teams understand the general tone of an interaction, but it rarely explains what kind of response the customer needs.

 

A customer arranging care for an elderly relative may speak politely while still being emotionally distressed. Someone calling about a blocked bank card may be angry, anxious or worried, and each reaction requires a different response. A simple “negative” label doesn’t tell an organisation whether the customer needs reassurance, speed, de-escalation or specialist support.

 

This distinction matters because emotion is more specific, and therefore more useful than sentiment alone. Language can provide clues, but tone, pace and context often reveal far more about what is happening in the interaction.

 

Understanding these emotional signals gives organisations a clearer view of what a customer needs, how best to respond and how to adapt processes to suit the situation. 

 

 

Emotion should change the customer journey

When an organisation is able to understand the emotions of a customer, it creates an opportunity to respond more effectively. A customer showing signs of anxiety, frustration or distress, for example, shouldn’t be treated in the same way as someone making a routine query. The interaction might need to be prioritised, routed to a more experienced agent, or directed to a specialist team.

 

Making the distinction benefits the customer, but it also has an operational impact. When customers are routed incorrectly, they have to repeat themselves, wait for transfers or start again with someone new. At scale, those avoidable handoffs create additional cost, extend handle times and increase the risk that a difficult interaction becomes a lost customer.

 

This is where the relationship between automation and human expertise becomes particularly important. Gartner’s 2026 research found that 91% of customer service leaders are under pressure to implement AI this year, while more than 80% of organisations plan to expand human agent responsibilities. Together, these trends suggest that AI is reshaping where and how people add value, rather than reducing their role altogether. The challenge is to ensure automation can recognise where human expertise is needed and route customers accordingly, rather than moving them through the system faster.

 

This is where emotional insight becomes operationally valuable. It enables organisations to make better decisions about prioritisation, escalation and resource allocation, rather than relying solely on static routing rules. 

 

 

The tech involved

For emotional orchestration to work, the technology needs to operate quietly in the background. AI starts by interpreting more than the literal words in the conversation. Word choice matters, but so do tone, pace, hesitation and the wider context of the interaction. Together, those signals help identify whether the customer needs reassurance, urgency, de-escalation or specialist support.

 

This emotional signal then becomes part of the decision-making logic. It sits alongside intent, customer history, channel, agent availability and agent skill, so the next step reflects both what the customer is trying to do and the emotional state they are in. This is what makes the journey feel more responsive - the system adapts before the customer has to push harder.

 

The handover to the agent is just as important. By the time the conversation starts, the agent should know what has already been said, which records matter and what emotional context has been detected. This removes one of the biggest and most understandable sources of frustration in customer service: being asked to repeat the story from the beginning. It also helps the agent choose the right tone from the first sentence.

 

Over time, these interactions will help to improve the system itself. If certain routing decisions consistently resolve anxious, frustrated or distressed customers more effectively, the learnings should feed back into the orchestration layer. The result is a customer journey that becomes more adaptive with every conversation, with fewer static rules waiting for someone to rewrite them.

 

 

The path forward

The next stage of CX will be shaped by organisations that understand the limits of automation as clearly as its benefits. They will know when AI can resolve friction quickly, when it risks making the experience worse and when a human needs to step in.

 

Emotion is the signal that makes this possible. If contact centres continue to treat it primarily as a post-call reporting metric, they will risk overlooking its greater value as a way to inform and improve the customer journey. 

 

AI’s biggest impact will come from using emotional context to make better decisions about what happens next, connecting data, systems and people so the customer’s experience becomes faster, more relevant and, as Professor Zaki argues, helps to restore “human connection at scale.”

 

---

 

James Cadman is CCO at Luware

 

Main image courtesy of iStockPhoto.com and sturti