Financial sector customer experience is being overhauled as a new generation of artificial intelligence-powered chatbots are taking over as the frontline interface between banks and customers.
Chatbots are about to change consumer banking as we know it, enhancing customer service, cutting costs and improving efficiency and fraud detection.
Bots are micro-apps that reside within larger applications such as instant messaging platforms and merge human interaction with artificial intelligence (AI) to provide a simple and fast means to communicate with users.
For banks, bots promise to be both a brand manager and accountant’s dream. Not only do bots reduce the need for customers to phone call centres for assistance, they also provide better customer service and ensure greater brand consistency.
No longer will a bank’s reputation be imperiled by a disheartened employee at the other the end of a phone line: bots provide a single voice, tone and message to every customer, and run as smoothly at peak hours as during the dead of night.
“I see the migration from a standard banking, stand-alone app, transactional app, over to something much more conversational, much more social,” said Colin Payne, Principal & Global Domain Lead, Digital Banking, Capgemini Consulting.
In October 2016, Mastercard revealed plans to launch a chatbot for its bank clients that would allow consumers to transact, manage finances and shop via messaging platforms. The company partnered with FinTech specialists Kasisto to run a six-month chatbot pilot. This found the most in-demand service requests were balance queries, transaction history and financial advice. The bot is platform agnostic and can be used via messaging apps, text messages and even within a banking website. Mastercard’s bank clients will personalise the bot.
“The ability to help respond to a customer inquiry live via a two-way communication, not necessarily reacting to a set of FAQs that may be on a bank’s website, moves the conversation away from having a single standard way of communicating with everybody to one that is personalised,” said Kiki Del Valle, senior vice president, Commerce for Every Device, Mastercard.
ENHANCING OPERATIONAL EFFICIENCY
Canada’s Finn.ai has developed a white-label enterprise grade platform purpose-built for financial institutions. The AI-powered assistant acts as a conversational day-to-day banking concierge and won the company $50,000 in the FinTech and Mobility category of Capgemini’s InnovatorsRace50.
Finn.ai’s tool – available on any platform with chat or voice capabilities including Facebook Messenger, Amazon Echo, Google Home and native iOS and Android apps – will provide a better customer experience with shorter waiting times to resolve questions, said Jake Tyler, Finn.ai chief executive.
“End users get customized advice, recommendations and suggestions based on their financial health and goals,” said Tyler, whose company’s clients include ATB Financial and several other tier one and two banks in Europe, Latin America, Asia and Africa.
By answering lower level queries, the assistant frees up customer service staff to handle more difficult questions, increasing the bank’s operational efficiency, while customers can perform day-to-day banking tasks through their preferred channel without having to download an app or go to a website.
“Our banking assistant offers a very significant cost saving opportunity in the way that customer care is done,” said Finn.ai’s Tyler.
“It is trained to answer thousands of the most common banking support questions automatically. These questions typically account for 50-80% of in-bound call volume.”
Should the system not understand or be able to answer a question, it hands over to a bank agent to help. It learns from these interactions so it can answer similar questions in future.
Outsourcing bot and app development to specialist companies appears a wise move.
“Conventional banks are limited by their legacy IT systems and restrictive internal policies and procedures,” said Jonas Huckestein, Chief Technology Officer and Co-founder of UK mobile-only bank Monzo.
“We are seeing the same thing happen to banks that has happened to many other industries: they accidentally became software companies without ever developing a core competency in that area.”
Customers today want little and often services from their banks. For micro services, chatbots are an ideal solution. Even for more complex questions and enquiries that take time, chabots can guide you through that process.
For example, the bot might see that a customer has bought flight tickets and so prompts them to buy insurance or notify the bank of the trip. Or perhaps the assistant notices there isn’t enough money in a consumer’s account to pay their upcoming bills and so suggests the customer get an overdraft.
“The possibilities are virtually endless,” said Tyler.
Customers, via a bot, can use social indicators to validate transactions and help reduce fraud. Similarly, geolocation can tell the bank a customer is in another country and so they will no longer need to go through laborious process of informing the bank they’re abroad.
In terms of fraud detection, users can use the finn.ai system to alert their bank to transactions they don’t recognize, while in future it may alert users to unusual spending habits or purchases, said Tyler.
Analytics have also advanced to the stage where AI technology can build up a behaviourial profile of an individual – AI can measure stress in a person’s voice and other mood indicators such as the way they text, the language used and the rapidity of response.
THE PSD2 PHENOMENON
From next year, new banking regulations – PSD2, or the second Payment Services Directive – will end banks’ proprietary hold on their customers’ data. Customers will be able to give third-party companies permission to retrieve their account data, manage their finances and provide financial services on top of their bank’s infrastructure.
“Either banks take the view that PSD2 is also an opportunity, because it means they can then get access to the data of other banks’ customers, opening up the possibility of creating new business models,” said Capgemini’s Payne, “or some banks may be content to simply maintain the existing relationship they have with their customer. It’s possible to do that, but there’s always a risk that a nimbler FinTech could use the data in a better way – it could be an aggregator, it could be one of several new types of mobile banks.”
Perhaps the biggest threat to banks from PSD2 will come from the likes of Apple, Samsung and Facebook, which have detailed social data on their hundreds of millions of customers.
“Say you’ve had an account at your bank for 10-20 years and you then give this data to Facebook, it’ll have the sophistication to mine that data to more carefully predict your spending patterns, your interests,” said Payne.
“Link that with the social data they already have on you and it’s going to be a very interesting proposition. For Apple or Facebook, it wouldn’t be practical to utilize a traditional service with purely human intervention, so they need a great solution and a chatbot is one element of that.”
END OF STANDALONE BANK APPS?
Consumers are spending more time on apps, but access a smaller variety of apps – a 2016 comScore report showed 45% of users’ time on their phone was devoted to a single app and 73% on their three favourite apps combined.
Such statistics show how difficult it is for new applications to grab users’ attention, so making a bot that works within a “super-app” such as Facebook Messenger makes commercial sense, especially as bots involve significantly lower development costs than a standalone app.
“We know consumers are spending a lot more time on messaging applications and we’ve also seen a decline on downloads of apps. Consumers are engaging less in individual applications,” said Mastercard’s Del Valle.
“We believe that for a time bots will play a complementary role to a banking application, mainly because today you still can’t provide every service that you’re providing on a mobile banking app in a chatbot, for example.”
That perhaps explains why mobile-only bank Monzo has no plans to launch a bot.
“We believe that our users should always be able to speak to a human, especially given that we don’t have any retail branches,” said Monzo’s Huckestein. “That said, a human that is assisted by AI will probably provide a lot better support than they otherwise would.”
Monzo’s approach seems to be succeeding – having launched a pre-paid debit card in November 2015, it had 240,000 accounts as of February 28, 2017, according to the company’s annual report. These accounts spent £250 million during the reporting period.
Promising instant payment notifications, built-in budgeting and fee-free travel, Monzo expects to have up to 800,000 accounts by 2017-end, while it was granted a full UK banking licence in April this year.
“We use technology to allow our customers to communicate with us via social media,” said Huckestein. “We also use technology to run our community forum, which is key to the Monzo brand. Most of the third-party software we use is open source. As a result, our technology costs are incredibly low.”
London start-up BUD, whose clients include three of the UK’s six biggest banks, is another firm to bet on the durability of standalone apps. It has created a fully customizable, white label app for financial institutions that aggregates financial service products from multiple providers.
“The customer can use any product through one app,” said Jamie Campbell, Head of Awareness at BUD. “Banks are often criticized for selling products that are unfavourable to clients. By implementing our services, banks for the first time show their customers the whole market of options, and the power will be in the customers’ hands.”