You Can Replicate Amazon’s Powerful New Tool
Banking Transformed with Jim Marous
October 14, 2026
Hear the full conversation with Jorge Camargo of Bank of America on Banking Transformed: Banks and credit unions can let customers see and correct what the institution assumes about them, the way Amazon's About You page does for shoppers.
Speakers Jim Marous
TopicsBusinessTechnology
Jim Marous (0:00)
Amazon asks its customers what matters to them, and then lets them correct what it thinks it knows. Most banks or credit unions haven't done this. On Amazon's app and website, there's a page called About You. A shopper can view the preferences behind the recommendations, change them or remove anything they don't want used. They can even ask Amazon Shopping Assistant, what do you know about me? And get an answer. So the company we spent years worrying about, which knows too much about our customers, is letting those same customers tell it when it's wrong. Any banker or creditor can start asking customers three questions, plus a follow up that checks whether you understand the answer. This could be the difference between relationship growth and silent attrition. Think about how we actually handle this in banking. Imagine a customer tells her bank that her biggest worry is paying down her credit card debt. Over the next few weeks, let's say she receives three promotions encouraging her to open another card and spend enough to earn a bonus. Nobody meant any harm.
Somebody heard her, but nobody recorded where the next conversation or the next offer would see it. My own primary financial institutions know me very well, but they only prove it very infrequently. And that's a customer who has asked. Most of the time, we don't ask at all.
We determine what customers need from their transactions, and customers never see how we base our decisions. Transaction data is very good at telling us what a customer did, and not very good at explaining why. So when we guess wrong, the customer receives an offer that doesn't fit, deletes it, and quietly concludes that their bank or creditor doesn't really know them. In a recent Banking Insight video, you don't really know me. I made the case for having real conversations with customers. Amazon's move raises the next question, which is what do we do with the answers, and whether the customer ever gets to see them? The pushback I usually hear is that the customers don't want their personal financial institution asking personal questions. Bank of America proves otherwise, and I'm using them only because they've measured it at scale. Asking a customer a good question and answering it well doesn't require Erica or a big technology budget. I sat down with Jorge Camargo, who runs digital platforms at Bank of America at the Financial Brand Forum. Erica, their virtual assistant, has had more than three billion conversation with customers. Jorge said those conversations helped the team identify about 700 topics across consumer banking. He also described an overarching concern customers raise in the relationship. In his words, am I good? Is everything in my world okay? That's a financial wellness question and it's the kind of reassurance a bank or credit union should be able to provide. The bank also asked customers about their goals through a tool called LifePlan. And by 2023, those customers had scheduled more than 2.3 million follow-up appointments with the bank's financial specialists. The top goals were budgeting, saving and improving their credit. I also asked Jorge whether Erica would start asking questions back to understand the customer better. He said yes, and that large language models will allow much richer conversations.
That's the opportunity for every bank and credit union. Ask and then let the customer check whether you understood. I haven't seen many financial institutions do that second part yet. And a community bank or credit union may have the edge here because the person reading the answer might already know the customer by name. Let me give you an example. Picture a customer named Beth who lives in an apartment. For two years, she's been moving $400 into savings every month. Her bank places her in a group. It targets for mortgages.
She starts receiving home loan offers and an invitation to a first-time homebuyer seminar. Best money is actually for her mother who's moving closer to her and will need some help with expenses.
None of those offers are a disaster, but each one tells Beth that her bank has her figured out wrong and she has no way to tell them. Now, give Beth what Amazon gives its shoppers. A page in the app that says, here's what we think you're saving for. Did we get it right? She taps no and types, helping my mom. The app asks whether she'd like help planning for those expenses, and she says yes. The mortgage offers stop. When she sits down with the banker or calls the contact center, they already know why she's saving and start asking her which costs she expects. So here are the questions. The first question is, what's the biggest financial challenge facing you today? Is Jorge, am I good? Asked out loud and it puts the customer's needs first. What you owe the customer return is a next step, a tool or a person and a time they can expect it. If she says it's an emergency fund she can't seem to build, that might be an automatic savings tool or a call from someone who can set up one for her. The second question is, how can we do better? When a customer answers that, they deserve a reply that explains what you can change, what you can't and when. If a customer says it's hard to find their routing number in the app, fix it and then tell them you fixed it. The third question is, where else do you do business? Now, I'd ask this last, after you've helped the customer the first ones, so it doesn't sound like it's the beginning of a sales pitch. Your data already shows money leaving for a credit card in another bank, to Venmo or mortgage at a credit union. It's the flow of funds analysis I've discussed in an earlier Banking Insight video and now you're asking the customer to explain it. They're the only ones who can tell you why the money goes, where it's going. And after every answer comes the fourth part of it. Here's what I understood.
4 more minutes of transcript below
Thousands of transcripts fetched by people building searchable podcast archives
Fetch the whole transcript
The demo key returns a sample episode in full, no card needed:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000796288380