Exciting articles several times a month
7 February 2018
Chatbots are exploding in the finance industry, and it is no surprise. With the ability to automate operations, reach more customers, and provide a more friction-free banking experience, chatbots are streamlining and optimizing many banks’ digital services. So, who is doing it best? Let’s take a look at some of the big names in chatbots for banking.
Bank of America’s Erica
Erica, the new bot from Bank of America, can take inputs via text or voice, and is intended to “get smarter” the more data “she” collects. Customers will be able to also make better financial decisions, apparently, as they have an advisor who will crunch more numbers than any given bank teller.
Announced at Money2020, it is not being used yet, but it will use “cognitive messaging and predictive analytics to assist customers with payments and checking balances.”
Erica will help customers make payments, check balances, save money and pay down debt. She will also direct people to look up their FICO score and check out educational videos and other content. The bank hopes Erica will provide the personal service of a top-tier customer, to the masses.
The Royal Bank of Scotland (RBS) launched Luvo, their chatbot that answers customer questions online and helps direct them to the right place.
Luvo is able to understand questions and then filter through huge amounts of information in a split second before responding with the answer. If Luvo is unable to find the answer, it passes the query on to a member of staff who can solve problems that are more complex. It will support staff to help them answer customer queries more quickly and easily. It has to be trained when dealing with new subject matter, but crucially, it learns from its mistakes and its answers become more accurate over time.
One problem with Luvo, however, is that it can interact only with bank staff and not directly with customers. Also, there is still the need for human experts despite having AI deployed.
Swedbank was the first Swedish bank to introduce a virtual assistant, Nina, in 2014. Nina is a cognitive robot, using Natural Language Processing (interpreting, understanding and responding in written human language) to address incoming queries from customers to contact centres, so that the agents can focus on value-adding businesses. Nina does not listen or speak, but communicates online via chat.
Swedbank has received some two million customer queries annually, and Nina is now handling 40,000 per month, 81% of which she successfully resolves. This means that Nina deals with almost a quarter of Swedbank’s incoming customer queries. Swedbank says that so far the number of staff at its contact centres has not been reduced; instead, employees are spending time on other types of calls, typically more complex queries, which increases value for the bank.
Nina has no avatar, or virtual face; she only exists as a search field where customers type their questions. But that could be changed in the future. The bank plans to include her inside the internet bank (after customer have logged in). Then she would be able to help the bank’s clients with individual transactional business. Eventually, she will also be available as part of Swedbank’s mobile app, with a voice-driven functionality, similar to Apple’s Siri.
In 2016, Skandinaviska Enskilda Banken (SEB) introduced a customer service chatbot called Aida. It is based on an AI platform called Amelia from technology group IPsoft, whose Natural Language Processing was first applied in Swedish in SEB’s Aida (name chosen by SEB staff voting – not a coincidence that the first two letters are “AI”).
Built with semantic understanding, it interacts with users through natural language and senses emotions. The first trials of Aida were among some 700 staff in SEB’s IT support. Users type in a question in the chat conversation, and Aida answers through both voice and text. According to SEB, more than 4,000 conversations were held with the 700 employees during the first three weeks after implementation and Aida solved the majority of issues without delay. Aida currently works on the bank’s internal IT Desk, where she speaks her original language of English, and in Swedish customer service, where she speaks Swedish.
DNB on Facebook Messenger
In collaboration with Convertelligence, a supplier of natural language processing AI, DNB has developed a chat bot that responds to customer queries on Facebook Messenger. The robot is expected to reduce the waiting time significantly, even for those customers who do not receive an answer from the bot.
Currently the robot is under development and all inquiries that it is unable to handle are then handled by a human being. Through the chat history, the robot will learn to successfully answer more and more queries itself. By 2020, the robot is expected by DNB to handle 80% of the bank’s customer queries. If the expectation is met, it will free up a lot of time for customer service employees, who can then take on more complex problems.
In small numbers, on messaging apps and digital assistants, bots are already bankers. Their number is growing and whilst their success is not assured Moore’s Law and the commodification of Artificial Intelligence technology make it more likely than not.
Machine learning will allow banks and financial services companies to cut costs and scale up to offer customers a personalized and always-on service wherever they are. Chat bots will never be a standalone solution to business challenges; rather a part of a company’s larger portfolio of digital touchpoints.
Just as banks adapted to technological shifts by making themselves available over the phone, on the internet and on our smartphones, they must now add chat and messaging to their omni-channel strategies.
The bots are ready, are you?
Exciting articles several times a month