Customer Experience in the Age of AI


Photo by rawpixel.com from Pexels

The experience of bank customers in recent years has been largely defined by a push to self-service, first through ATMs and online banking and later to mobile devices. As transactions migrate online, the number of branches (and banks) continues to decline. Customers that visit branches to do their banking similarly find that to be a transactional, sales-driven experience. Where have the relationships gone, and can AI bring them back? 

Banks now have the capacity, through analytics performed on customer data, to determine and in some cases predict events of significance in customers’ lives. Banks can offer advice for financial events as they happen – if a new direct deposit is larger, the customer’s status may change, and new products may be appropriate. Yet many customers, logging into their mobile apps several times a week, are confronted with the same screen, showing account balances. Checking account balances is the most frequent reason for logging in (though a recent IBM study noted the most common question customers ask chatbots is “What’s my password?”) but much more information can be provided.  

We see the beginning of these possibilities in neobanks such as Chime, now worth $1.5 billion since the injection of new funding, and the parallel banks launched by Wells Fargo, called Greenhouse, and Finn, from JPMorgan Chase. Finn and Greenhouse take their cues from other consumer-facing apps bank customers use every day, and similarly, banks are advised by pundits to take cues from outside the industry, particularly big tech players such as Facebook, Google, and Amazon. 

However, 2018 saw significant stumbles from these businesses, whose business is largely monetizing customer data by it selling to other companies. Google’s connected home devices showed themselves to be leaky and insecure, and Facebook was shown on multiple occasions to be less than transparent and less than honest about how it handles customer data. So while the user interface of these apps and services may be a model for banks to emulate, their businesses certainly are not. Consumers may no longer be interested in engaging with a bank chatbot on Facebook Messenger because there is no telling what Facebook will do with the data it collects.

As transactions steadily move to digital, banks are looking at new ways to manage the relationship. Many young customers have no idea what their parents’ or grandparents’ banking experience was like. Their knowledge of interacting with companies is limited to social media – the idea of speaking to a banker on the phone is outlandish. Banks are therefore following other industries in introducing in-app chatbots to engage with customers. The most notable example is Bank of America’s Erica, which claims more than 4 million users, against its mobile user base of 25 million. 

A simple set of stock responses can handle the vast majority of queries, but beyond that banks will need an intelligent system to handle variations in language and ultimately, provide responses suited to the customer based on circumstances and demographics. A variety of fintech firms specialize in just this area, but chatbots cannot simply be bolted on and expected to work. Recall the frustration customers experience with IVR systems – chatbots need to handle the conversation better and hand interactions off to humans less frequently. This means connecting the chatbot to internal systems such as the CRM, and developing a model for managing these interactions internally, perhaps a dedicated team or shared services model. 

The customer is changing – to meet his needs, the bank must change too. And without a strong employee experience strategy, it will be difficult to create a positive customer experience model.