Professor Mohamed Zaki on CX design: Personalisation and digital capabilities

In case you missed part one of the interview, take a moment to read about some of the key challenges related to CX design. 

In a conversation with MyCustomer, Professor Mohamed Zaki, the deputy director of Cambridge Service Alliance and course leader of Data-Driven Design for Customer Experience (CX) at the University of Cambridge, delves into CX design.

This interview explores the concepts of personalisation and leveraging digital capabilities, exploring the considerations of maintaining a human touch while incorporating AI technologies into customer interactions.

AI and human connection

Personalisation is an important topic that can be a concern when it comes to AI. For example, missing out on the human connection when interacting with a brand.

As a human, most of our preferences are human-to-human interactions.

"As a human, most of our preferences are human-to-human interactions because we humans understand each other, sometimes, I would say. You need to have some emotional intelligence, you know, when we fail some of our customers. It's emotionally loaded when services fail, and expectations aren't fulfilled. 

"In customer service centres, experienced employees know how to call a customer, understand their feeling and emotion, and then respond well to some of these failures."

Over the years, we’ve started to see organisations working in an ecosystem. 

"A simple example here is retail, right? So you order from Zara or Next or something. Typically, the one that delivers to you is a separate organisation. That could be, for example, FedEx.

"Now, that delivery service could fail to deliver the product you already ordered through the e-commerce channels or online, making you, sometimes, complain to the brand. You're not complaining to the delivery organisation."

Mohamed explains that when we digitise what we refer to as autonomous services, the interaction of AI with the human customer comes into play. Emotions can be expressed concerning both positive and negative feedback. Despite customers experiencing various emotions, Mohamed explains that there will be little to no reaction when expressing them to an AI-enabled digital service.

"Typically, there are standardised responses that will come from this kind of chatbot. And it doesn't have a personal touch. And we, as humans, more or less pick this up," Mohamed explains. 

Empathic chatbots

Empathy between customers and human agents is one thing, but how does that work with a chatbot?

"As a customer, if a chatbot starts to empathise with me, I don't care. It should get into solving the problem. I'm not interested in the, you know, the human side of it," Mohamed says.

As a customer, if a chatbot starts to empathise with me, I don't care.

"But in general, we found in some of the research we did that to personalise interactions with customers; you need to understand key elements here. One is knowledge about the customer, and the previous experience is really important."

Personality plays an important role in knowing who you are dealing with. As humans, we can quickly gauge if someone is more likely to complain or be more enthusiastic. However, with machines, it's not the same. 

"To define that personality quickly, maybe we're not there yet. But to start using this knowledge can enable you to communicate better with the customers in this case and build the right style of conversation. 

"So if I'm a customer service agent, I'm coaching a little bit based on the information we collected about the customer, the current engagement, the current emotions, and helping the agent to communicate when they should, if you like, empathise with customers."

It’s also important to know when to apply empathy or have these types of conversations. 

"Maybe at the beginning of the conversation, I don't accept it. Leave me to, you know, spit it out first, if you like. Then we can start to talk about it."

In customer service, elements like friendly greetings help build empathy. The question is whether these elements can be automatically integrated into chatbots or text-based interactions to enhance customer experience.

"In the end, we want the customers to feel at these touch points and interactions like they've been heard."

Fulfilling expectations

Mohamed explains it is worth considering how the design of an AI chatbot can address both the need for customer service teams to minimise call volume and customers to avoid long waiting times. Another aspect to consider is how effectively the AI chatbot can engage in casual conversation, and when it should do so.

The customer doesn't want to chit-chat with the machine.

"There is another situation where the customer doesn't want to chit-chat with the machine because they understand it is a machine. They want the functions to be, you know, solved because they expect an autonomous solution, not a human-to-human interaction."

When designing the approach, the aim is creating responses for human agents that cater to the needs of individual customers. These responses should be personalised based on the customers themselves, enabling a more tailored solution, whilst leveraging AI to automate and optimise this process.

The goal, explains Mohames, is to fulfil customer expectations and deliver the desired outcome through effective interaction and conversation.

"This needs to be designed. It needs to build that kind of humaneness and the right components that enable this."

Designing a holistic experience 

While many organisations aim to empower agents using AI, there are challenges to overcome. For those looking to successfully design customer experiences in this context, Moahamed has some practical tips. 

"The organisation needs to think about the full holistic picture right now; customer service is one touch point. It's not the only touch point where customers visit and interact with us."

Customer service is one touch point. It's not the only touch point where customers visit and interact with us.

Mohamed draws up an example for booking a service for his car through a garage: 

"The traditional way of doing it is; I'm going to call the garage, I'm going to say I want an appointment. And then the garage will give me their availability. You as a customer have to bring the car to the garage and stay there for two or three hours until the service has been delivered. You get a report, pay for it, and go back to your destination, right? 

"The service here is configured based on human-to-human interaction. I'm going to customer service; I'm saying I booked an appointment and want my car serviced. A technician will come from the engineering department, take the keys, look at the car, and start building the diagnostics, do the diagnostics, do the assessments, do those fixes, and come back with the bill, and I pay for it. If I'm satisfied with the experience, I spend three hours there doing my work," Mohamed explains. 

He continues: "There are huge physical, social elements here."

Now, if Mohamed ran a garage, and wanted to gain a competitive advantage by differentiating his services, there are a few things he would do differently:

"One of the key frictions here as a pain point for my customers is waiting on the phone line until someone responds. So I will do a digital app that automatically correctly identifies a suitable slot for the customer," he says.

"Now, more is needed because everybody can do this, right? We could go the extra mile here to satisfy the customers. Another big friction is staying three to four hours away from my daily routine. The only way I can do this is probably on weekends. What if I can offer this for the whole week?

"So companies can say, 'Oh sir, we can do this through an app. You can share, you know, a key to your car, and we can send someone to pick up the car from your home and take it to the garage. We will fix it for you. And when we've done everything, we report this to the app. Payment will be taken automatically once you put in the credit card details, and then we can send it back to your home.'

"I, as a customer, now have a hassle-free service and experience. I didn't have to dedicate two or three hours to have that kind of service done. And typically, more customers would be happy with this kind of service."

Driving value creation

The configuration of organisations has changed significantly with an emphasis on digital capabilities. Human employees are not always necessary for customer interactions, as demonstrated by Mohamed’s example.

"That's kind of what I'm doing at the moment, thinking more about value creation elements, what I'm offering to my customers, what resources I give to them, what activities I do for them."

Mohamed continues: "Know the customer's role and how important it is."