Chatbots have been increasingly popular with various organizations for the past couple of years – both in and out of the context of the financial industry. Just about anyone realizes how helpful they can be for the client, providing quick and thorough answers for basic questions and saving banks’ some precious time. But there is an aspect to the role of chatbots that few realize and take advantage of. Let’s find out how the smartest algorithms can help human consultants improve their interactions with clients – both at the time the interaction takes place and in the long-term.
Despite the fact that last year, chatbots accounted for merely $20 million in savings for financial organizations, Juniper Research estimates that the figure will reach as much as $8 billion per year by 2022 . This astounding prediction owes to both rapid development of the technology and ever-increasing interest of banks that consider chatbots one of the least expensive and hassle-free ways to jump on the Artificial Intelligence (AI) bandwagon.
Indeed, chatbots offer multiple benefits for both banks and customers. But before we get to that, let’s take a step back.
Chatbot – not exactly AI, yet still of tremendous value
In the broadest sense, chatbot is a computer program created to simulate conversation with a human. In the context of the internet, it is usually associated and employed through text-based chats and messaging apps and serves as an alternative way of using them as opposed to human-to-human interactions.
It’s worth knowing that present-day chatbots shouldn’t be considered Artificial Intelligence in the full sense of the word. As opposed to projects such as Google’s AutoML or machine learning solutions employed by Amazon Web Services , they do not include trainable models that keep on evolving seemingly on their own as they come in contact with more and more data. Instead, chatbots are made up of advanced algorithms capable of recognizing limited number of words, signs or questions and associating them with the most appropriate pre-defined actions or answers. While they are clearly not infallible, a user in need of valuable information, be it a client or employee, can really use their help.
How do chatbots help banking agents?
The inherent flexibility of the chatbot solution makes for a number of possible case scenarios already applied in practice. What we’re describing is a unique and rarely used application of automation, which serves to help primarily agents rather than clients (though helping clients in the process). Let’s take a look at the following cases, facilitated by the Chatbot integration module of the LiveBank virtual bank branch solution:
The customer is searching for information on credit cards. They start a chat conversation through Facebook Messenger to learn about specific card parameters. Based on the context, the question is redirected to proper specialists (using LiveBank’s Routing module). However, as all of the most appropriate experts are busy, the inquiry is further redirected to another consultant, who is less experienced in the subject. And here comes the unconventional use of the chatbot functionality: the question also reached a prompter-enhanced algorithm. Normally, the consultant would find it difficult to answer the inquiry in a way that would satisfy the client. This time, however, they can use the help of the chatbot, which immediately suggests four different ready-made and detailed answers to choose from, which contain data that frequently changes, for example, transaction fees. The consultant picks the most suitable one and alters it slightly to make it match the context of the original question even better.
The answer is subsequently sent to the client. As the conversation progresses, the consultant can pick even more chatbot-generated answers to send to the client, modify them or write answers on their own.
Chatbot – banking agent’s best friend?
Thanks to the chatbot, everyone wins. The bank can tend to its clients’ problems easier as even less experienced consultants can answer detailed questions with the help of chatbot-generated templates. What’s more, every time they get familiar with those ready-made answers, they educate themselves and improve their own knowledge of the bank’s offer, learning and providing service at the same time. As a result, the bank improves the efficiency of client-consultant interactions and constantly educates and improves the knowledge of consultants.
It’s also worth adding that the presence of human consultants, who can verify the answers provided by automation and pick the best one, reduces the risk of incorrect responses, which occurs when the client asks questions that the algorithms of a fully automated chatbot were not prepared for. Such incorrect responses are very frustrating for clients and may discourage them from ever turning to them for help. As such, the human serves as a smart interface between the client and the chatbot, making sure that the interaction has just the right mix of automation and human touch. The chatbot can be very helpful for experienced consultants as well, since it may provide them with the fastest way to check the latest data on fees, commissions, card types etc.
At the same time, the client enjoys shorter waiting time, better availability of consultants (as even less experienced consultants can prove sufficient) and improved service quality.
Artificial Intelligence is here to stay
Considering the popularity of text-based messaging apps (e.g. social media websites) and the relative ease of implementing chatbots for banks, it’s easy to see why banks would choose them as an introductory step towards the world of artificial intelligence. And the unique way the LiveBank application employs chatbots shows that artificial intelligence may be used not just to replace, but also support and enhance the skills of human consultants.