Why Banks Must Shift From Speed to Revenue in the AI Era

Hey everybody, welcome back to another episode of What's Going On Banking. I'm Ron Chevlin, Chief Research Officer at Cornerstone Advisors, here, of course, with my co-host Stacy Bryant. Stacy, what's going on?

How you doing, Ron? Listen, I am living the dream. I am in basketball mode right now. Just want to let you know. I know we touched on that a couple of weeks ago, but my girls are in AAU, and it's all basketball these days, particularly because the New York Knicks are up. We smashed the Chicago Bulls by 30-some-odd points. So, you know what? I know you love when I speak about that, and I am on cloud nine.

Apart from being on that sports cloud nine, Ron, I'm gearing up to head down to, I believe, Marco Island in Florida. Never been there before, but shout out to Claire Cusack. She is the President and CEO over at the New York Bankers Association, and they invited me to speak about AI. This time around, we're going to go beyond AI 101 and speak about experimentation and execution.

What does that look like as folks are thinking about their 2027 initiatives? So, super excited to have that conversation.

Also, I'll be moderating two panels, so a trifecta of sorts. I'll be moderating a panel on tokenized deposits and stablecoins. Our guy Peter Gleason is going to be one of the panelists, so I'm super excited to hear about what's going on with him and everything he's doing. Then another panel discussing the C-suite role, what talent looks like going forward, what the future of talent looks like in banking, and what is actually keeping executives up at night.

I'm a geek and a nerd in what we do, and I'm very, very curious. I'm going to resurrect my Biggie Smalls-from-Brooklyn, New York, mad-question-asking hat. I cannot wait to see if they entertain me or if they kick the moderator off the stage. But nonetheless, that's how I'm doing. How you doing?

I'm doing great. You've got an exciting schedule coming up. I've got really exciting stuff too. Next week, I'm going to be mowing the lawn.

But seriously, week after next, and it'll be pretty close to when people actually get to listen to this. Stacy, I am super psyched, and I want everybody who subscribes to this podcast to know that after this one, the next episode you hear will have been recorded live and in person in Nashville, Tennessee.

We are doing a live broadcast as a keynote presentation at a conference, which is really amazing. It's not like we're just sitting off to the side doing this. But, Stacy, I hope you're getting really nervous for this because that will be the first episode we do where we're actually going to be in the same room at the same time, sitting at the same table doing this.

So, I am super excited for that. Really nervous doing it live, or at least recording it live, I should say. Maybe we'll need some editing after that. But I'm really looking forward to it. I can't wait to see you in Nashville.

You know, I'm not nervous. Is that a bad thing? Should I be nervous?

Yeah, we need to talk. It's good to be nervous.

Anyway, let's get into some topics because there's a couple of things I really wanted to talk to you about. Obviously, some of it's kind of self-serving. We published a bunch of research lately, and we've got to cover some of this stuff.

One in particular I really wanted to highlight for folks who are listening because I think there's some really important ideas in the report. Obviously, the whole topic of AI agents is huge. You can't look at any website that has anything to do with banking or fintech without seeing millions of references to AI agents.

And yes, we did a report on AI agents in marketing. But for me, the importance of the report wasn't about AI agents. It was about how speed becomes such an important revenue driver and strategy, and it's not just about marketing.

The real key message in that report was this: everyone's been focused, especially in the marketing world, on time to market. How long does it take you to get a product into market or a message into market? But the idea of the report was that you really want to be focused on time to revenue. How long does it take to actually generate revenue?

I think one of the key findings in that report, Stacy, was that a lot of marketing departments, especially in the bank and credit union world, don't realize that, okay, maybe it took three weeks, or even six weeks, to get a campaign out the door. Then you start measuring campaign performance based on response rates, conversion rates, and all that kind of stuff.

But what they're not realizing is that by not getting the campaign out the door faster, they've lost revenue. They're looking at the wrong aspect of getting things into market. The real question is: how long does it take you to generate revenue?

AI agents enable that, but there's a business aspect to it. That's why I thought the report was important. It's about speed to revenue, not just speed to market.

What was your takeaway?

I told you earlier before we hit record that this is my favorite research paper of the year.

I'll answer your question regarding speed, because that's the obvious premise of the research paper, Ron. One of the main reasons I really enjoy it is because there's so much good material about what's happening today. How are financial institutions, particularly marketing departments, launching campaigns today? What does that process look like? What's standing in the way?

I really appreciate that we're also providing a vision for what that could potentially look like. At Cornerstone, we always pride ourselves on helping institutions become a smarter bank or smarter credit union through 2030. The report really breaks down the root causes of speed and why speed isn't just another metric.

I told you on the last episode that one of the main themes coming out of American Banker's Most Powerful Women in Banking conference was agentic AI. And I'll be honest, while a lot of the conversation was focused on governance, I didn't really hear use cases. I didn't understand what that looks like today or how workflows are actually changing.

For all our listeners and viewers, this is cream-of-the-crop material, especially as we gear up for 2027 planning. I really want to get to the root of speed. We speak about speed and how the lack of speed is taking a huge bite out of how credit unions and community banks scale, absorb market share, and grow.

The root cause of the speed issue is this $5.5 billion beef.

Again, for our listeners, I'm originally a Brooklyn, New York girl. When you've got beef with somebody in New York City, you've got a problem.

And the problem here is $5.5 billion. I took this out of the research paper, from the section "From Adversaries to Architects." The friction is structural.

We talk about how marketing and compliance operate as adversaries today instead of partners. Marketing creates, compliance reviews. Marketing revises, compliance redlines. Again, the friction is structural.

You laid it out in the report. The friction produces a six-week exchange cycle and a five-hour-per-message dynamic that costs the industry $5.5 billion.

What does that really mean?

There's a financial institution, not named in the report, that believes if they add a compliance team member to the marketing team, involved from the very beginning of campaign development, it will reduce the amount of time it takes them to capture market share and acquire customers.

Let's get down to the nitty-gritty.

Suppose I'm a member of ABC Bank and I'm constantly using my debit card at Home Depot and Lowe's. There are lots of home improvement purchases. Wouldn't that be the ideal time to run a campaign promoting a home equity line of credit or some other lending product?

This is just one of many scenarios.

How beautiful would it be if those purchases happened over the weekend and, shortly after, I received a relevant email from my financial institution?

What does that mean? Stickiness, retention, differentiation.

Again, this paper really lays out a plan for what institutions are trying to achieve in 2027 and beyond.

Thoughts?

Well, I'm glad you liked it. That's always a good sign.

A couple of things struck me while we were writing it.

First of all, if you go back 35 to 40 years ago, and I know nobody wants to hear, "Okay, Boomer, tell us about 40 years ago," the hot management topic of the day was business process reengineering.

That was really driven by the adoption of personal computers and the move from mainframes to microcomputing.

As that happened, smart people started saying, "Hey, we should rethink business processes in light of this new computing paradigm."

They looked at processes that took six weeks and figured out how to get them down to three weeks. That was a huge impact.

But we're now at a point where AI, and AI agents in particular, can take those processes from three weeks down to three minutes.

The technology is here. The technology is capable.

But a few things are stopping institutions from doing it.

First, we don't have access to the right data.

Second, much of the required data isn't transactional or financial. It's qualitative data. It's policies. It's institutional knowledge. The data is in the wrong format.

Third, we're not giving software the authority to make relevant decisions.

We're sitting at a point in time where we talk a lot about the Smarter Bank mantra. That phrase came from a book I published in January 2015. And we're a lot closer to January 2027 than we are to January 2015.

If I were writing a new book today, it wouldn't be called Smarter Bank.

It would be called Faster Bank.

I think we've moved beyond the smarter bank goal. The new goal is the fastest bank.

Faster for sure. Fastest bank.

A few thoughts come to mind. I say this almost every episode, but I am obsessed with what the future workforce looks like in banking.

I'm thinking about our annual What's Going On in Banking report that comes out every January, where we survey hundreds of community bank and credit union leaders about what's keeping them up at night, their priorities, and their appetite for risk.

I'm curious if we can add a question to the survey.

I want to hear from participating financial institutions about not just the future workforce, but the org chart. Not who does what, but who collaborates heavily with whom.

I want to understand those collaboration patterns.

At the root of it, I'd love to identify which financial institutions really work hand-in-hand with their marketing and compliance teams on a regular basis.

Again, you speak about speed and becoming a smarter institution. When we think about the future of banking and the new plumbing of banking, it's also the new plumbing of the workforce.

Which department leaders will need to rely on one another for growth?

That's one comment.

The second goes back to a conversation we had about personal AI agents.

Let's say I'm Stacy. I'm in a particular part of the country. I have $75,000 I want to invest. I don't want to go to NerdWallet. I want GPT or some other large language model, or even one of my own agents, to figure out the best rate for me.

The reason I bring that up is because I'm working with a number of financial institutions that are focused on growth, staying relevant, and attracting younger customers.

They're thinking about how they can show up in searches, whether those searches are being conducted by AI agents or large language models.

Going back to the research paper, we talk about journey orchestration.

I want to highlight one example from the report. After an eight-month implementation project connecting a predictive analytics platform to core transactional data and HubSpot, a $3 billion asset credit union in Michigan now runs three to four always-on campaigns triggered by CD maturities, auto loan maturities, and attrition signals, without manual list pulls or campaign setup for each send.

I'm talking about two things simultaneously.

First, how can institutions make sure they appear as relevant options when consumers direct their agents to find products?

Second, how can institutions use agentic AI for marketing and member acquisition?

We're talking both about existing members and prospective new members.

We keep discussing the importance of data cleanliness, infrastructure, integration, and building a smart foundation.

I want our community bank and credit union listeners to understand that this is a $3 billion institution. This is doable.

So often I hear, "We're not JPMorgan Chase. We're not one of the big banks. How can we possibly do this?"

This is where fintech collaboration matters. There are solutions available to help prevent churn and drive growth.

It's important to understand that this is achievable.

My final comment is probably a little biased because I spend so much time on the road talking about AI.

I still hear from some older leaders, respectfully, that because they're not large institutions with big budgets, they can't do these things.

They're discounting their own institution's ability to reach new heights.

I'm really trying to drive home the point that it's doable, and this is how it can be done.

Okay. A couple of responses.

First, I want to address your comments about the cross-functional aspect of all this, specifically marketing and compliance.

One of the things that came out of our research was that it isn't simply about assigning a compliance person to marketing.

The reality is that compliance departments have resource constraints too.

Let's say marketing gets a dedicated compliance resource. That's great.

But generative AI tools are giving marketing the capability to create much more content than before. Let's just say twice as much for simplicity.

If compliance staffing doesn't grow alongside that increase, you've created a bottleneck.

Marketing is producing twice as much output, but compliance still has the same capacity.

People have limits. We can only process so much work.

If marketing doubles the amount of content and campaigns being produced, you can't reasonably expect compliance to double its resources as well.

That's why AI agents become so important.

For all practical purposes, they don't have the same resource constraints. You can keep throwing more work at them, and they'll continue processing it.

One of the challenges with technology adoption in banks and credit unions is that technology tends to develop within functional silos.

There's a marketing technology stack. There's finance and accounting. There's operations. There are branch systems.

But once you start crossing departmental boundaries, very little is designed to address those interactions.

That's where the opportunity with agents lies.

I also had a comment about older leaders pushing back on all this, but I think we've beaten that topic to death a little bit.

We've got a few more minutes before reaching our self-imposed time limit, but I wanted to bring up another topic.

I recently saw a survey conducted by Morning Consult, a very well-respected research firm, commissioned by the American Bankers Association.

Credit unions, don't get mad. It's not really about the banks specifically.

The survey looked at fraud protection and found that about 44% of Americans trust banks most to protect them from fraud, compared to 8% who selected government, 6% who chose non-bank fintech providers, 5% who picked AI providers, 1% who selected cryptocurrency companies and telecommunications providers, and about 29% who said they didn't know or had no opinion.

A couple thoughts, Stacy.

First, I can't imagine only 44% said banks. The 8% who said government are definitely smoking something we don't smoke.

The 6% who chose fintechs clearly didn't understand the question.

The 1% who picked cryptocurrency and telecommunications providers weren't paying attention and were just clicking random answers.

And the 29% who said they didn't know or had no opinion? God bless them. They told the truth.

I've got more thoughts on it, but before I go there, what was your reaction?

Before I give you my take, let me make sure I understand.

What I'm hearing is that you're saying some survey respondents weren't really answering the question.

No, no, no. I'm making fun of the people who think government, fintechs, or cryptocurrency providers are doing the most to protect them from fraud.

I just don't buy it.

Honestly, I was surprised 100% didn't say banks.

So let me ask you a question.

You've been around the block. You always talk about mainframes and 40 years ago. You've seen banking from the institution's perspective.

As a customer yourself, what does trust mean to you today?

I'm glad you brought that up because that's exactly where I wanted to go.

I thought this issue was important because it goes well beyond the headline.

First, the survey was commissioned by the American Bankers Association, so of course they're going to highlight the fact that banks were selected most often.

But what the survey didn't ask was how well banks are doing. How well are fintechs doing? How well is government doing?

The ABA isn't going to love me saying this, but reality is that few, if any, organizations are doing a truly great job helping consumers avoid fraud.

That wasn't asked.

My second concern is that banks will take these results and start beating their chests.

"Look what a great job we're doing."

Wrong conclusion.

This isn't what builds trust.

Trust comes from outcomes.

Imagine your bank came to you next month and said, "Stacy, here are the 33 fraudulent attempts we prevented this year."

You'd be thrilled. You'd feel a level of trust that nobody else had earned.

But then imagine saying, "Okay, bank, that's great. Now let's talk about the 17 irrelevant offers you sent me. You wanted me to refinance a home I don't own. You offered me a home equity loan even though I live in an apartment. You promoted a credit card that's worse than the six I already have."

You'd say, "You're doing an excellent job on fraud protection, but a terrible job helping me manage my financial life."

That's another dimension of trust.

The problem is that we rely on a simplistic understanding of trust when it's actually multi-dimensional.

Okay, I've got a little beef with you.

What I'm hearing is that you're saying personalization is part of trust.

I believe personalization is important and can be part of the trust stew, but when I asked what trust means to you as a banking customer, here's what comes to mind.

Protect me. Notice when something doesn't look right.

Treat me fairly.

Don't do only what's good for you.

Be dependable. Let me access my money whenever I need it.

Respect my control. My information isn't a blank check for you to do whatever you want.

And finally, take responsibility. Don't make me manage your departments when I have a problem.

What I would have appreciated from the ABA survey is understanding what trust actually means to respondents.

If they trust banks, why?

If they trust government, fintechs, or someone else, why?

I'm pushing back because while personalization is important and contributes to how well you know me, we're talking about two different, important values.

We can't conflate them.

Well, believe it or not, I don't think we're disagreeing.

I think we're agreeing.

You identified some important dimensions of trust, and that's exactly the point.

Different people place different levels of value on those dimensions.

You might care far more about fraud prevention than someone else. I might care more about another aspect.

That's the challenge.

We mix everything into one giant pot, and that's not how the world works.

We're reacting to simplistic surveys without doing the hard work of understanding, first, what the dimensions of trust actually are, and second, how different consumer segments value them differently.

Nothing drives me crazier than that.

In fact, our colleague Brandi Gregory and I are doing a webinar next month on this exact topic.

I didn't mean to turn this into a commercial, but I'm basically working through some of those ideas in real time.

Anyway, I'll stop my rant and throw it back to you. Do we keep going, or are we ready to wrap up?

No, I think it's safe to say that the kind of beef we have isn't chuck steak. It's filet mignon, medium rare.

That's the kind of beef we have.

Going back to the ABA survey, my last comment is that when you really understand surveys, the first questions that come to mind are: Who were the respondents? How many were there? Where were they located?

The socioeconomic factors matter.

For our listeners who use survey data in strategic planning or to support initiatives, I know this sounds like common sense, but I've learned that common sense perhaps isn't that common, Ron.

Make sure you vet the data because it helps tell a story, and that's especially important when you're using members' money and stakeholders' money.

That's it.

Apart from that, I think we're ready to rock and roll, and I cannot wait to do that with you in the honky-tonks of Nashville.

Look at us.

All right. Good luck with your conferences over the next couple of weeks.

Safe travels, and I will see you in Nashville.

And for everybody else, thank you for joining us, and I hope we'll see you at another episode of What's Going On Banking.

If you enjoyed today's episode of What's Going On Banking, make sure you hit that follow button and subscribe wherever you listen, whether it's Spotify, Apple Podcasts, or YouTube. We've got more witty, gritty conversations coming your way, and you won't want to miss them. Stay tuned.