Ignore what vendors say about ‘unlocking new levels of loyalty.’ There are three requirements driving the need for chatbots in banking.
The pandemic of 2020 may have been bad for a lot of businesses, but it was certainly good to technology vendors providing chatbots—i.e., conversational AI (artificial intelligence)—to financial institutions (FIs).
According to Cornerstone’s What’s Going On In Bankingstudy, heading into 2020, just 4% of mid-size FIs had deployed a chatbot. By the end of the year, that percentage had more than tripled to 13%.
Another 16% planned to invest in chatbots in 2021. But even if they were all to do so—that never happens—it wouldn’t even get chatbots to the 30% adoption level among FIs.
The question that financial institutions must address: Has the time come for chatbots to become as ubiquitous in banks as mobile banking apps?
The answer is yes—although the reasons aren’t obvious. In fact, a lot of the evidence points in the other direction.
Are Chatbots a Competitive Necessity for Banks?
An article in Finextra argues that chatbots have become a competitive necessity:
“The magic behind voice banking isn’t just the convenience of the device; it’s the artificial intelligence that powers intelligent customer self-service. New enhancements including AI combined with voice biometrics will unlock a new level of loyalty among financial institution customers.”
I wouldn’t recommend telling that to an FI’s executive team or board of directors.
Proponents of every banking technology innovation of the past 25 years—from online banking to online bill pay to eStatements to PFM to mobile banking—have promised deeper relationships and higher levels of loyalty.
What Do Consumers Think?
Not a single technology has lived up to the promise. Why would chatbots be any different?
You’ve probably never heard of Shevlin’s Law, but I’m doing my best to make it known. The law states:
For every statistic that proves a point, there are two that refute it.
Nowhere is this truer than with chatbots. There are stats that prove that consumers hate chatbots, and stats that assert that consumers love chatbots. For example, one survey found that 86% of consumers prefer to interact with a human agent, and 71% said they would be less likely to use a brand if it didn’t have human customer service representatives available.
There’s another, study, however, that found that 74% of users prefer chatbots while looking for answers to simple questions.
It doesn’t really matter what consumers say about chatbots.
As with the technologies that preceded it, banks will find that there are two types of chatbot users: 1) those that are eager to use them, and 2) those that have to be dragged kicking and screaming to get them to use chatbots.
Sooner or later, the latter group will acquiesce—they always do.
Why the Time is Right for Banks to Deploy Conversational AI
Ignore what the vendors say about “unlocking new levels of loyalty.” There are three requirements driving the need for chatbots in banking:
1) The need for speed
Abandonment rates for digital product applications in banking are horrendously high. According to a recent study from Cornerstone Advisors, roughly half of the banks surveyed said that in 2020 half of their checking account applications on digital channels were abandoned. The abandonment rates for unsecured and secured loan applications were even higher.
Even more troublesome is the finding that just a minority of institutions follow up with would-be applicants within a business day. That’s unacceptable. Banks need chatbots integrated into digital account opening systems to close that gap.
Banks need to make chatbots components of critical business processes (like account opening)—not just generic sales and service tools.
2) The need for data
Chatbot vendors like to use “providing advice” as a use case for deploying chatbots. It’s an over-sold justification of chatbots. Personal financial management (PFM) tools have been trying to provide advice to bank customers for years with little success.
The problem isn’t the user interface. That is, providing advice through a chatbot versus an email or a pop-up in a PFM tab or tool isn’t the magic bullet.
The problem is lack of data. Some people will tell you that “banks have a lot of data about their customers.” Maybe. But they don’t have the “right” data—that is, the data they need to truly figure out what advice is the right advice to provide.
Banks need chatbots in order to collect data, not display data.
Attempts to codify and store “data” collected through human interactions—and even from clickstream data—is incomplete, generally inaccessible to other applications that could benefit from the data, and hard to analyze.
Data gleaned from chatbot interactions can overcome these shortcomings.
Banks need to make chatbots part of their data management strategies—not just their sales and service strategies.
3) The need for personalization
Many banks recognize the importance of personalization in customer interactions. Some, unfortunately, think of it too narrowly, in terms of personalized messages.
The smart banks understand that good personalization requires personalized conversations. They still wrestle, however, with two things: 1) getting the data to deliver good personalization, and 2) creating opportunities to have personalized conversations.
Banks need to make chatbots part of their marketing strategies—not just as standalone, ad hoc marketing messaging and offer delivery tools.
Banks’Digital Transformation Delusions
Prior to 2021, 56% of financial institutions had launched a digital transformation strategy or initiative. Among those that believe they’re at least half-way through completing their strategy, just a little more than a quarter have deployed chatbots to date.
Just 12% of the banks half-way or more through their digital transformation efforts plan to invest in a chatbot in 2021, and a quarter of them said chatbots aren’t even on their radar.
My take: That makes no sense at all. How can a bank be half-way or more through completing its digital transformation strategy and not have deployed—or even thought about deploying—a foundational digital technology like conversational AI?