“Chatbot” was once a buzzword used across various industries – from financial institutions to online retailers. Today, artificial intelligence (AI), which interacts with customers through online platforms, is the norm. From a customer’s point of view, chatbots can be a quick way to get answers to straight-forward enquiries; for corporates, they are a cost-saving tool. A study from Juniper Research found that the operational cost savings from using chatbots in banking will reach US$7.3 billion globally by 2023, up from an estimated US$209 million in 2019. Apart from financial cost savings, such tools also free up customer service agents to concentrate on more complex enquiries.
However, the proliferation of chatbots also comes with frustration and skepticism around customers’ experiences with them. The stakes are high for companies that decide to use chatbots for customer service, as the experience delivered can have an outsized impact on a customer’s perception, sentiment and loyalty towards a brand. As such, there is much to be done before chatbots become a tool that consistently elevates customer satisfaction, rather than creating friction. In this context, financial institutions should consider a few key points when launching customer-facing chatbots.
Most chatbots today are able to handle the basic day-to-day enquiries efficiently and accurately, but there are limitations in their ability to understand context. When a customer request is complex, chatbots can get ‘stuck’ when a question is worded less directly and produce inaccurate identification of an action item.
This is where advanced AI in the form of Natural Language Processing (NLP) comes in. Using NLP, new chatbot technology can understand the context and intent of requests, making it able to process requests that are more complicated and written in a more conversational way.
The new generation of advanced conversational user interfaces will not just save time for both for banks and for customers, they also add value by making proactive recommendations. They do this by accessing information from many places at once using open APIs, and bring it all together in myriad useful ways. For example, a person might ask an app if they can afford to buy a particular car. The app might then gather information such as credit score and income to make a recommendation on that car’s affordability, and even recommend alternatives. This kind of computer-driven proactivity will not just improve the customer experience, it will also boost revenue.
While the prospect of more capable chatbots will be attractive for organizations, most people would agree that chatbots should not replace human assistance entirely, but rather supplement it. As such, another key consideration during a chatbot conversation is to allow the customer to be directed to a human assistant if they need it – before the limits of the chatbot become a frustration.
An example of a seamless chatbot experience is AliMe, a chatbot developed by e-commerce giant Alibaba. The chatbot supports all end users of its e-commerce platform, which includes retail buyers and sellers. Its intensive use of machine learning and natural language processing allows it to analyze customer enquiries and emotions, and in turn prioritize or introduce intervention from customer service staff when such a need is detected.
It also predicts customer behaviors on the platform and delivers timely prompts and preempts (such as discount voucher reminders), hence ensuring the chatbot adds value to the customer’s experience. By doing so, the chatbot succeeds in becoming part of their experience unintrusively and seamlessly, without losing touch of the human elements of customer interaction.
With the capabilities afforded by NLP, the augmentative benefits of chatbots do not have to stop there. Imagine a scenario where a small business owner has called his banker to discuss options for financing growth. A chatbot listening to the conversation could ascertain the context and pull data from various sources to seamlessly ensure sure the banker has all the necessary information to hand, tell the banker how much the bank would be willing to lend, and even start the loan application process. All this saves valuable time for both the banker and the small business owner while enhancing the customer experience.
Powerful AI is making the chatbot experience more successful in Asian markets, where the sheer diversity in language and culture in the region highlights the importance of delivering a relevant and localized experience. In Hong Kong, the rolling out of a Cantonese-literate chatbot by local bank Hang Seng caters to the fact that a typical dialogue in Hong Kong may involve a mixture of English and Chinese words, and the chatbot is designed to be able to interpret bilingual requests. This means a user does not need to worry about translating English terms they use in their daily Cantonese conversations, and thus reduces communication friction.
Conversational user interface channels will be increasingly important and subject to comparative scrutiny by customers, as banking moves further into the digital space. The rise of virtual banks also means customers are increasingly likely to rely on chatbots or digital forms of communicating with the bank for their enquiries.
While the technology is promising, there is still much to be done. A recent research index from Forrester has shown that banks in Singapore, for example, are still unable to provide satisfactory customer experience on the three key metrics of emotion, ease and effectiveness. With chatbots the first line of customer contact in many cases, it has never been more important for banks to invest in R&D and partner with third parties with NLP expertise in order to take their chatbot capabilities to the next level.
– Contact us at [email protected]