Beyond the realm of revamped underwriting solutions and better backend operations, artificial intelligence (AI) is making quite the wave on the customer-facing front. However, while it promises to revolutionise the way all industries work, many have voiced discontent toward the notions that while it may prove to be cheaper and more efficient, it would be an astronomically unethical replacement for the human workforce.
In fact, just like your typical Second World War-era flick, many fear that it’s the boots on the ground – agents and other intermediaries – who will be the first to feel the tip of a theoretical AI warhead. In conversation with Insurance Business Asia, Verint North Asia and Korea vice president Matty Kaffeman (pictured above) explained that while there is something to this idea, the current iteration of this technology and the way our world works today will still need the human touch – or, more appropriately, expertise.
This service is where the “boots on the ground” comes in. Not one to be left out of the AI revolution, customer engagement specialist firm Verint has touted its offerings to a variety of big-name carriers like Allianz, IAG, RSA, Aegon, BGL, and more. In this field, Kaffeman said that AI technologies have proven to be incredibly useful, especially in making an agent’s work-life much easier.
“The customer’s history and all relevant data, the agent could access it all very easily on the desktop,” he said. “Maybe you’re a person that sends us an email every week. Maybe you’ve complained before, maybe never. I can see the history of what we talked about in the past. Maybe we even made some mistakes. It lets the agent easily know the context of how to deal with a customer.”
With this differentiating factor in mind, Kaffeman believes that people will not be disappearing from contact centres soon, a sentiment echoed by most in the region.
“In our view, it’s all about how we empower them. And you know, if I look at the tedious routines that they must do, how do I take them away? If you just look at a call summary, that’s a great example. That’s tedious work. That’s something that really takes time. If I can make my agent happier with the work they’re doing, the service will be better,” he said.
All of that said, Kaffeman does think about a future where everything is automated.
“Think about yourself going to the bank 15 to 20 years ago,” he said. “You may ask the banker how much money you have in your account. Now, just how ridiculous does this question sound today? I take my phone, I open it, I click on HSBC, and I’m looking… 10 seconds later, I know how much money I’ve got.”
“I will never wait in line for half an hour to ask how much money I have. That doesn’t make sense. We see how it has evolved. However, while people now want to use something more digital and they don’t want to always talk, on the other hand, they are still looking for a more sophisticated solution,” he said.
Citing an instance where he raised a very complex question regarding his house policy to an insurer, Kaffeman said that a bot will not always have the answer, and this is where the expertise of an agent will come in. However, it does come with an understanding that the value that agents are expected to provide needs to be above what an AI can offer.
“If it’s more complex, you still need the people involved behind it,” Kaffeman said. “However, I want to make sure when agents are involved, it’s because it’s critical, and not something that’s less important. If I’m talking to you, it’s probably because I’m complaining – I’m calling you because the bot could not handle it, and so you better know what’s going on with me.
“I think there is something to the idea where you can automate everything and use AI to do everything. I don’t think we’re there yet, very far from it, even. Another thing we see is that there’s no one AI that can solve everything. ChatGPT cannot answer everything. It can try, but then you need to see what will happen if it answers incorrectly – for the insurance industry, that’s a very real risk. Can you let AI automatically answer customers? That’s a very big risk you will have to take.”
Looking to trends in the space, Kaffeman said that it comes down to the aspect of empowering agents and streamlining the customer experience. With the latter, it becomes a question of passing customers to different channels without the process feeling like a chore.
“How do I make sure I can streamline the customer experience when they move between a bot to human and when I move across a different omnichannel?” Kaffeman asked. “Maybe I started with a text on the cell services and move to a person that forwards me to another webform and followed up with WhatsApp – I want the entire experience to be related.
“How do I make sure all of it creates a one-way flow of experience that is consistent? And how do I make sure that the customer repeats themselves every time? Sometimes you talk to an agent, and you give them your requirements, and a few days later, somebody calls you, and while they really want to help you, they don’t know all the details.”
Throughout this journey, Kaffeman also stressed the need for the relevant information to be there when it is needed, something that is essential to making the customer experience as hassle-free as possible. Going back in time, this was very hard to do, as he reminisced about calls which can take hours and result in nothing useful for the customer.
“In the past, all we could have was a synchronic discussion,” he said. “If someone back then did not know the answer, you can be put on hold for a few hours. Today, when I go into WhatsApp and interact with an organisation, it makes sense if they don’t answer on the spot. I’m not sitting and just waiting. I’m doing other stuff as well.”
Part of enhancing the customer experience also means keeping a close eye on their interactions with agents. This is where AI-based compliance checks come in, Kaffeman said, something that insurers are very well-acquainted with as an industry that works with risks – both internal and external – on a daily basis.
“For insurance, there is a growing concern around compliance, making sure that everyone’s following regulations,” he said. “Being able to automate the process and use AI that can help me manage and make sure that we can look at all the interactions, that I have 100% of them, is great.”
“If you look at automated quality management, there’s a list of questions. It could be a simple one: did the agent present himself properly? Did he say, ‘how are you doing sir?’ Was he polite? Did he use some bad words? Did he interrupt the customer all the time? You can identify all these things from the analytics we’re doing.”
Risk mitigation also means proper upkeep of data, and Kaffeman understands the consequences of improper upkeep, especially in the insurance industry.
“Another thing is data validation,” he said. “It essentially means that I can take AI to make sure that the interaction is reflected in our record in the right way and highlight anything which is not. I think this is a critical thing for risk mitigation. In this industry, if I write the wrong information, that can be devastating for some people. Customers that rely on insurance would be affected, and it could turn into a big problem for the insurance company to explain it.”
Part two of Matty Kaffeman’s conversation with Insurance Business Asia will be published in the coming weeks. Stay tuned.
What are your thoughts on this story? Please feel free to share your comments below.