Developers creating artificial intelligence for the financial industry promise that their tools will maximize convenience and boost productivity, but when is the “human touch” still needed to ensure the successful implementation of these fintech solutions?
Listen in as Baker Hill’s lending experts share their thoughts and discuss the role of AI in the coming digital transformation in banking.
Mitch Woods: Welcome everybody to Baker Hill’s Lending Made Easy. Today we’re going to talk about a topic that can be a little provocative when we start talking about digital transformation in banking: artificial intelligence.
We discussed Netflix’s impact on banking in a previous episode and how that’s really impacted customers’ expectations. A lot of that’s driven by their artificial intelligence and personalization. But banking, on the other hand, isn’t a transactional business like Netflix, so how do we really experience artificial intelligence or use artificial intelligence in a way that’s going to strike that balance with the human touch?
Today’s episode is about artificial intelligence versus the human touch in banking. I’ll start by just sharing a quick story.
I recently was on a website and having some difficulties and pulled up their chat feature, thinking I’d be able to get the answers that I wanted, but I had a terrible experience with their chatbot and ended up having to email someone. I still haven’t heard back three days later.
But, on the other hand, I’ve had some other great experiences when artificial intelligence has worked really well to get the answers I need.
Today, Bryan, I’ll kick it off to you first here. When we start talking about striking a balance between artificial intelligence and the human touch when it comes to digital transformation in banking, where do you see financial institutions really playing in that space?
Bryan Peckinpaugh: Yeah. Great question, and welcome. Welcome to the podcast, Mitch. So new voice for everybody that’s out there. You know, talking about a bank’s digital transformation is always interesting. This topic came up in our recent digital lending transformation webinar panel discussion with a few experts in the industry.
To me, it’s always a “better together” concept regarding a digital transformation in banking. It’s also a client experience concept, where we’ve always gotta be cognizant of the experiences we’re delivering across all channels. Right?
Sometimes, having an outstanding experience in one channel and a poor experience in another creates such a wide gap between the two that it seems worse than if I’m just average in all channels, right? If I’m average in all channels, it’s consistent from a human experience perspective. We don’t notice a difference. So it’s perceived a little bit better.
When you start to bring the AI, you have to be careful that you don’t go too far.
Like you, Mitch, I had a similar chatbot problem where I just had to keep hitting representative-representative-representative to get some actual person to engage. I had a complicated question that wasn’t in their question tree, and they didn’t account for that.
The power of AI in bringing insights to the people will always be a people business.
You will always need human insights into the data and the human relation to your customer’s current mental state or emotional state. Being able to look a person in the eyes or hear in their voice to discern what is going on (and then responding to that appropriately when armed with all of the most relevant and timely data) is—I think—where we’re going to see the best experience, right?
Again, the key to success with any digital transformation in banking is to try to create that consistency and not have big dips if I’m engaging with a chatbot or calling an automated number.
You’ve got to find that blend. You’ve got to find the technology that will arm your client-facing people with as much relevant data as possible. You’ve got to make sure it’s prioritized in stack rank. I can’t overwhelm them either, right?
If I get lost in the data, it’s a question of finding the balance. Where we have personal interaction, which there’s a lot of in banking as there are in other industries, you just can’t go too far one way or the other.
Mitch Woods: David, same question: Where do you see artificial intelligence really fitting in in the banking industry?
David Catalano: Yes, I definitely like what Brian had to say. I would encourage banks to use it internally and use it with their employees before they embark on the customer aspect of AI if they could do that. Meaning they need to be answering the phone. If they can’t answer the phone, I’d rather have some bot at least begin to attempt to help me out.
But, at the end of the day, there are process automation tools that you can use as part of a digital transformation in banking that is in the AI category that will help you consume or basically remove repetitive tasks from your process and allow the humans to work on more complex processes that require critical thinking.
So getting people out of the robotic work of something like statement spreading (or some of those activities that you really can’t start using the data until you get it in a format that’s usable). Start there. I can also see applications in marketing, where you’re trying to predict the next best product for a client.
What do the other clients have that look like this one? And, which one should you serve up? Or, how about looking at account behaviors and changes in accounts?
When I see changes in behaviors of an account of a deposit account where deposits aren’t coming in the way they used to, loan payments aren’t coming in the way they used to, or balances aren’t maintained in a certain range, can they trigger that human interaction?
Using an automated process to identify who I should talk to and then have a person make that outreach. Making a digital transformation in banking combines the tool and the team to create the best possible experience for the client.
I think that chatbot is a way to field lots of calls and narrow them down to a funnel. Maybe that works better today than it did ten years ago or 20 years ago, for that matter. I’m not sure how long that’s been around, but if it doesn’t work, it doesn’t create a very good experience for the client and shouldn’t be used, as both of you mentioned.
Bryan Peckinpaugh: It’s certainly better. David, I think you and I think a lot alike in this because as you were going through, you’re mentioning many solutions, and this is an area where we’ve got buzzwords just for buzzwords’ sake. I’ll lump AI with generic terms like analytics, big data, and machine learning. To me, those are so whats, right?
What are we actually trying to solve? And that’s where the banks need to get back to first principles and say, what are we trying to solve for? What type of data are we trying to get into the hands of people? Cause it sometimes, and you were giving some examples there, David, where AI isn't necessary if you look at it like a dictionary definition.
If I’m looking to identify profitable clients and untapped market segmentation, I don’t need a digital transformation in banking or artificial intelligence to do that. I need different types of loan origination tools.
You must think about your entire toolkit and then pick the right tools for the right jobs. What you really need is the human capital that understands what those jobs are.
- What are you trying to solve when you think about the customer experience?
- What are you trying to solve when you think about the next best product you brought up?
- What do you want that to be?
- What tools can you use to optimize that?
I may want to use analytics, not artificial intelligence, machine learning, or something else. I may also want to take baby steps and start to figure out my own data before I start trying to turn loose “artificial intelligence” and machine learning because they’re only as good as where we point them.
And, you know what we’re asking it to get? Even though some of the Google engineers said their AI is sentient as of a week or two ago, it isn’t there yet.
David Catalano: Yeah. I’m a huge fan of everything related to digital transformation in banking and creating ways in which the employees can do their work faster, more efficiently, and more accurately to create wow experiences for customers through that process. Because they are efficient, the customer feels really good about their interaction with a bank.
Bryan Peckinpaugh: Yes, and I’ve been fortunate to talk to some of the biggest banks in the United States and the world and some of the smartest people in banking that are out there, and even they struggle with some of these concepts.
I’ve got smart people, tools, and data. What do I go look for? That’s the crux of all digital transformation in banking solutions. These are not just the higher-level concepts we have at our fingertips, but what are the basics? What building blocks do I need to use in the lending processes? It’s great to talk about AI-driven pricing engines or AI-driven machine learning-driven decision trees.
If I don’t know what I’m comfortable with, I shouldn’t be pointing tools at it and relying on them from a risk perspective. That’s how you can eventually get yourself in trouble.
David Catalano: Yeah, I agree. Just thinking about it in isolation doesn’t make sense. You’re really trying to solve a base problem. What’s the best way to go about doing that? Is AI one of those things that we should consider? And, if not, that’s okay, too.
We’re just trying to solve problems. People get wrapped around the idea that they have this interesting technology, so they want to see how they could deploy it, even if it may or may not make a lot of sense for them or their customers, for that matter.
Mitch Woods: Yeah. I’m hearing that just because the technology’s available doesn’t mean it’s necessary, right? It’s taking a step back and understanding what problem I am trying to solve. Really how can moving forward with this position, my bank, or my credit union be more competitive in the long run, whether retaining my talent or my client base and making my customers happier?
With that in mind, when we start thinking about that strategy—as we’re wrapping up here—where do you start? Where do you decide this is what I’m going to use artificial intelligence for, or is this the use case for artificial intelligence?
Bryan Peckinpaugh: A great question, Mitch.
I’d say how a digital transformation in banking can help will be different based on each institution. It’s going to be how far into their digital journey they are. We talk about a digital lending transformation journey here at Baker Hill. I’m still leveraging highly-manual tools—whether paper and pencil or a spreading tool to capture my financial analysis data.
If that’s where I am, it’s going to be really hard to jump to a fully automated, single-platform loan origination system, where I’m going beyond just a score for automated decisioning and looking at cash flow and automating my portfolio analysis. That’s a big, big jump. I’m much better off having that as an aspiration and figuring out how I work my way there by adopting tools on a chunked basis.
It’s the same thing with these ideas, right? It’s where am I? When I look at my overall processes, I go back to the first principles and reevaluate everything I’m doing. Where do I feel defective? Where do I think the biggest lift is? Then, go target those.
In some instances, yes, it’ll be from companies that have AI or machine learning or their company names or their product descriptions or mission and vision statements.
Maybe it won’t, right? Perhaps it’ll be those doing the basic blocking and tackling because I need that first. It’ll be wildly different.
I always come back to look at your institution, your processes, and those areas where a digital transformation in banking can improve. Where do I want to differentiate? Where do I have the opportunity to get better? What solutions align with where I want to go as an organization? Bring those in, implement them, ingrain them in the DNA of your organization, and then just continue that over and over. Look for those continuous improvement cycles.
Mitch Woods: Awesome. And, David, any last thoughts from you?
David Catalano: Yes, I wouldn’t look for a problem that AI can solve.
I would evaluate what I’m trying to solve and then find a tool to solve that. I wouldn’t go to a hardware store, buy a hammer, and look around for projects on which I can use my hammer, right? That’s backward.
I want to look at what I am trying to fix and then figure out the best digital transformation in banking tools to go about fixing that. People within your organization likely have processes they use to identify which tools are best for the solution they’re trying to solve or the problem they’re trying to solve.
Just follow that. That’s the process I would follow. If AI lands in one of those solution sets, that’s great. If it does, that’s okay, too.
Bryan Peckinpaugh: Although that is why I have a lot of tools in my garage, David
David Catalano: We like tools, don’t we?
Mitch Woods: Absolutely. Brian and David, thanks for today's discussion about AI and digital transformation in banking. I think there are a lot of great points here, and the thing I take away is always looking for a way to put your employees on value-added tasks, but then also consider how you add value to the customer experience and—ultimately—make their lives easier as well.
So with that, thanks, everyone, for tuning in for today’s episode.