Hosted by top 5 banking and fintech influencer, Jim Marous, Banking Transformed highlights the challenges facing the banking industry. Featuring some of the top minds in business, this podcast explores how financial institutions can prepare for the future of banking.
Why Banks With More Data Still Don't Understand the Customer
Banks are collecting more customer data than ever before. Yet most institutions still do not understand the customer any better.
In this Banking Insights episode, I explain why that gap is widening, what Bank of America has built with Erica over 3.2 billion conversations, and why the real advantage in banking is shifting from collecting information to learning from it in real time.
Drawing from my conversation with Jorge Camargo, Head of Digital Platforms at Bank of America, I explore why most personalization strategies still fail, why transaction histories are not customer insight, and why the institutions that win will be the ones that ask better questions and act on what they learn.
Bank of America has had a private conversation with 1 in 5 Americans. Their Erica platform has answered more than 3.2 billion questions from over 50 million people. And every one of those interactions taught the customer something about how their bank worked.
In addition, every one of those interactions also taught the bank something about the customer that it could not have learned in any other way. That is where the advantage is building — not in the model or even with the technology used to build Erica. The advantage is in the accumulation of learning over time.
And while most financial institutions may be behind on AI, what is scarier for me is that they're falling further and further behind in knowing their customers.
To understand why that gap is hard to close, you have to look at what most institutions are actually working with. Most banks and credit unions have built the equivalent of an old fashioned map. The map shows transactions, account holdings, and even shows balances. It shows where the customer has been and where they are today, but does not know where the customer is going. It does not adjust when conditions change, and it certainly doesn't respond in real time. But the bigger issue is this — it never asks a question.
PwC has shown that most banks only see a small portion of the customer's entire financial life. The customer accounts in many different places. The rest may sit somewhere else — maybe it's at another bank or nontraditional financial institution, maybe the insight is on a favorite payments app, or maybe it's outside the financial ecosystem altogether.
But even within the portion they do see, most financial institutions treat that data as the entire picture instead of only the starting point. This is where the argument usually breaks down, because data is mistaken for insight, and these are definitely not the same thing. Data tells you what has happened. It does not tell you what it means, and it certainly doesn't tell you what to do next.
I usually call this the rearview mirror perspective, and no driver has ever reached a destination by looking into the rearview mirror. As I mentioned, most banks and credit unions have a transaction history, but they call it a customer view. What they really need is the GPS of the entire customers' financial relationship — not a record of past or even current activity with no understanding of intent. Because if you don't understand the intent, you cannot move from reacting to actually helping.
I recently interviewed Jorge Camargo, the head of digital platforms at Bank of America, when we were at the Financial Brand Forum, and he framed the problem differently — not around the data, but around the question that is underneath this data. Let's take a look at that conversation.
"The way we look at Erica is, first, it's all about answers, then it's about actions. And then finally it's going to be autonomy. So we've gotten really good over the past eight years at giving you the right answer and really being able to route you where you need to go to get that answer you need. We're getting better at providing those actions. So this is what agentic is all about and why we're so excited about the possibilities there. All of a sudden, things that were uneconomically feasible in the past to automate — all of a sudden agentic technology gives you the ability to say, hey, we can do that for you. You don't have to do it yourself."
So did you hear what the customer really wants? Jorge said the customer wants to know: am I good? Is everything in my world okay? That is the question that sits under every transaction, every relationship in your organization. That's the destination question. That's the address you're typing into the GPS — and it's the only question your data alone can't answer.
Actually, if you want to see the full video of me interviewing Jorge at the Financial Brand Forum, it's in the link below. You know what? It's really worth the time.
In my most recent insight video, I made a case for looking at outbound transfers as a sign of real attrition, and to get an idea of how much of that relationship you actually have. Let me make this practical.
I use Alexa at home to order items from Amazon — vitamins, dog supplies, paper towels, just like everybody does. And I also put in things that I forgot when I went to the store. Amazon does something that most banks still do not do. It builds context. Alexa and Amazon start to understand patterns — when I buy something, when something might be running low — and then it acts on that understanding. It suggests a missing item before I even know it's missing.
Now compare that to your bank or credit union. Your institution probably sees almost every paycheck, every bill, and every transfer to another institution — even if it's only part of the customer's financial life, it is still meaningful. And almost nothing is done with it. That is the gap of understanding that the customer is expecting us to know.
Camargo described Erica as a progression. It started in the beginning with just basic information — giving answers to basic questions like balances or did a transfer actually happen — helping customers get information quicker and easier. Then it moves to actions, using what it knows to service what matters in the moment. And now it's moving toward autonomy, where the system can anticipate needs and take steps on behalf of the customer across the entire organization. And it does most of this instantly — not the next business day or after a call back, but right then.
This progression matters because most institutions are still focused on improving the answers, while the advantage is shifting to acting on what you actually know and doing it in real time to help the customer get to their desired destination.
And here's what most executives miss. The eight-year head start that Erica has is real. The blueprint is not a secret. The discipline of asking, listening, and acting is available to any bank and credit union that decides to build it, and most institutions can build this far quicker than it took Bank of America.
So here are the questions that really matter.
Number one: Are you looking at your customer's entire financial portfolio, or simply what you have under your roof?
Number two: Are you asking questions to help understand both how big that relationship is and what that movement means?
Number three, and most importantly: Are you doing anything with what you learn? And can your customer actually feel it? Because if the customer cannot tell that your bank is getting better at serving them, then it's not. Instead, you're collecting more data to create better reports instead of building loyal and trusted relationships.
If this insight video resonates in any way, share it with others within your organization. If you want to hear about how Bank of America is building Erica's capabilities across their entire organization, watch or listen to my interview with Jorge Camargo that I did at the Financial Brand Forum. The link is below, and if you have a question or comment, please leave it in the episode notes. Thank you again for your support.