Banks Don’t Know Their Customers. AI Can Fix That (If They Let It)
By Justin Estes, Contributor at The Financial Brand
Simple Subscribe
Subscribe Now!
Executive Summary
- In a recent episode of Banking Transformed with Jim Marous, Backbase’s Marc Corbett explains how unified platforms, hyper-personalization, and proactive engagement can unlock true growth in a competitive digital landscape.
- Financial institutions need to move beyond legacy systems and back-office AI to deliver customer-facing solutions that drive growth and enhance loyalty.
- Adopting unified platforms and API-first architectures allows banks to create proactive, personalized experiences across the customer lifecycle.
- Balancing in-house innovation with strategic technology partnerships is essential to avoid technical debt and accelerate AI implementation.
Breaking Through Legacy System Barriers
Q: What are the most persistent challenges you’re seeing when financial institutions try to deploy AI solutions?
Corbett: I think it’s the same song and dance that we’ve seen throughout the last decade, while everybody’s been focusing on, let’s get to the next modernization of our software stack, let’s bring an application to our customers. We’ve seen the evolution of Web 2.0, then into mobile and now AI is a new tool, but it doesn’t mean that the landscape has changed.
So, I still see the same problems I saw a decade ago when I was first starting at Backbase which is legacy systems, fragmented data, organizational silos, business and technology aren’t speaking, aren’t working as one, agility and speed is always going to be probably the largest cog in the wheel.
Q: How are forward-thinking banks addressing legacy technology challenges to enable AI integration?
Corbett: I think you said this best when I was listening to you the other day, that the rope analogy. So, time has passed and institutions have not modernized their technology in the same way. They’re holding onto the rope and it’s getting further away; there’s a lot of catching up that needs to be done.
But at the same time, Rome wasn’t built in a day. We need to gradually replace outdated systems and it just starts with laying everything out, sometimes on a spreadsheet, we do it with our consultants a lot and just saying, “What are you using? Where are the redundancies? How can we look at this from a modularity standpoint and actually use this point solution, this integration pattern more than once across your network and get away from that risky big bang?” So, that progressive modernization, I think, is huge.
API-first architecture is another thing that we always focus on, which is shockingly still a significant barrier to kicking off an implementation, as that API readiness has to be in place.
Unlocking the “Growth Mode” Strategy
Q: Can you explain what Backbase calls the “growth mode” and why it’s critical for banks today?
Corbett: So, AI is a big piece of this; we’re using an AI power platform now to push this, but once again, AI is just a part of that larger strategy. Our banking leadership has taken a look at it and called it growth mode because we need to move away from a slow, disconnected legacy tech staff that hinders our ambitions and strategies and instead adopt a more nimble, all-in-one approach. This is where we come in, with a platform play.
We believe that a single architecture and code base, avoiding the mergers and acquisitions of technology products that often occur in the financial institution space and unifying that data experience are huge. That allows us to be more proactive, enabling us to identify the data layer and we’re also incorporating an AI layer into many of our capabilities, allowing us to tailor experiences.
Q: How does moving to a unified platform approach enable banks to be more proactive rather than reactive?
Corbett: I think you were saying it recently, Jim, as well, that tailoring that value proposition is huge. If you consider different users at various lifecycle events and stages in their lives, I have a mortgage with another financial institution. Why isn’t my primary FI offering me a mortgage solution?
I recall when I first renovated my home, I had to visit a nimble neobank to secure the necessary financial backing, as my bank wouldn’t provide it. Why is that? Why is that relationship so broken? So, tailoring those experiences can only come from disconnecting from legacy experiences and embracing a nimble, all-in-one platform suite.
Dig deeper:
- AI-Powered Growth Strategies for Modern Banking
- Unlocking Banking’s Growth Mode
- How Banks Are Failing Their Most Valuable Clients
From Back Office to Customer-Facing AI
Q: Are most banks still focusing AI efforts on back-office efficiency, or are we seeing customer-facing applications?
Corbett: I think so. So, there’s something we’ve been focusing on and it’s called Customer Lifecycle Orchestration. It’s really about increasing the product holdings per individual, thereby increasing that ratio. You hear about it a lot, but how do you do it better?
And I think that AI, while it’s going to help us clean up the back office massively and repurpose a lot of the workflows so that you can get into those digital insights and those high digital touch points that you want your member representatives to be facilitating versus just the minutia of finding and searching, the same goes for the front end in the front office where I’m a client, I log in as the customer and there’s kind of different algorithms that are running based on my life events that are giving me the right campaigns that are guiding me through first financial management flows and wellness flows.
There needs to be a system in place that dynamically changes my dashboard and prioritizes family-centric banking, as opposed to my old life when I was a single person managing a small portfolio.
Hyper-Personalization and Customer Trust
Q: How can banks overcome risk tolerance issues to deploy customer-facing AI that shows they truly know their customers?
Corbett: So, I think there is that aspect, but ultimately, we think that when we look at the industry, these banks have fundamentally built their relationships, especially in the customer financial institution space, even into the sub-regional space, based entirely on relationships themselves, intimately knowing our customers. Hyper-personalization is not a new concept; it was previously used in a niche setting, offering a white-glove feel.
Obviously, that’s not scalable, but hyper-personalization via AI to analyze vast amounts of data in real-time, understanding the customer’s needs and then preferences, leading to highly tailored product recommendations, I don’t see that as a risk as much as a helpful hint into the future.
That proactive engagement for me personally is where I see us predicting customer needs, triggering their perspectives, enhancing their services, increased customer lifecycle value over time so that it’s even harder to leave things that we’re doing with a completely built in family application where I’m ramping up my child now within the same FI giving them access to a virtual credit card.
Commercial Banking’s AI Revolution
Q: What specific AI applications are showing the most promise in commercial and small business banking?
Corbett: What fascinates me about commercial and commercial lending is the talk about obsolete, archaic legacy ways of working with such a new business model every day as well. It’s such an interesting juxtaposition of two factories where you have revolutionary growth and at the same time, we haven’t changed a lot of how we service those users.
And I think a lot of the automation, a lot of the ways that we offer up products is so credit risk assessments for lines of credit personalization of relationship management, your relationship managers and how they communicate with you, how they recognize your entities and sub entities automation of some of the entitlements within your system, fraud detection is massive.
Q: How can AI transform traditionally manual processes, such as credit risk assessment and fraud detection?
Corbett: So, we have a lot of the hooks already in our software where there’s a step up, there’s a real-time payment and it exceeds a certain amount. You need to use biometrics, you get a push or a nudge to a mobile device, someone’s on the go.
But then, by embedding fraud and AI, we can see how many clicks are being copied and pasted, what the behavioral data of that user is and whether it’s unlike what we’ve seen before from that same user. And automating that process is huge.
Then there’s document processing, cashflow forecasting, all these things that have traditionally been very manual, done on spreadsheets or aggregating external systems, now we can kind of interwind those into a single experience and then give the customer once again what they need when they need it versus throwing a million functional areas at them and expecting them to be a subject matter expert on my commercial platform.
Build vs. Buy: The Partnership Strategy
Q: What balance should financial institutions strike between developing AI capabilities in-house versus partnering with technology providers?
Corbett: So, this is going to be contentious because I’m going to be biased as a provider of a platform, so take it with a grain of salt. However, we’ve examined dozens of institutions that’ve built in-house capabilities and those that’ve acquired them from vendors in a single source of truth approach. In other words, if I’ve a problem, I need a point solution; let’s put it in place.
And time and time again, what we see is that when you take that approach, you may end up with unwanted tech debt. There’s the innovation and knowledge you need internally about AI, but I think you should utilize that in your assessment.
Q: How can smaller institutions compete with larger banks when it comes to AI innovation?
Corbett: But ultimately, what we’re driving towards is you come to us and you say, “Hey, Backbase, what’s in your roadmap?”
And we show you what’s in there and say, “Well, actually, I’d like to develop a feature faster than that.” Okay, well, let’s give you the keys to the castle and open it up. The Agentic AI powers it, with an AI and data layer below. Now, you can use your own sources to build and expand on that roadmap and potentially even partner with it if it’s an initiative that we deem fit.
This is a good, I think, harmonious relationship we have with a lot of our even smaller CFIs in the space who want to take that risk but don’t want to necessarily hire 15 AI developers or people in the market who don’t really want to take on that risk in general about the management and cost of that tool, cost ownership over the next five years.
