Is Fintech Softening? With Jason Henrichs

I’m a fan of all things marketing and customer engagement when it comes to financial services. I always look at those things, too. It is interesting. You touched on something that I want to drill in on a little bit. That is data and analytics. The FinTechs do a great job at using that. Correct me if I’m misquoting you. The banks have typically been doing a lot of that stuff, particularly some of the larger institutions. It’s been this secretive private thing where they keep all of the analytics and data in one place and don’t necessarily drill that into the business processes where somebody is actually taking care of a customer that, if they have that insight, could have a different conversation.

We need to free the data, which means you need to be able to pull it out of silos and, more importantly, get it clean into the hands of people and then train the people. This is as much a cultural issue. One of our members at Alloy Labs, the head of analytics and I were having this conversation. Their chief banking officer kept asking for these reports.

He was like, “What are you looking for?” She’s like, “The CEO says we need to be more data-driven.” He goes, “One, you don’t understand what data and analytics really are. This is not an old-school management information system. Let me give you some dot matrix, green and white striped paper. You got to tell me what you’re looking for. What kind of insight? Let me go test something.” Just generating reports doesn’t tell you what to do.

There are two approaches to this. There is the discovery phase, which is probably a little complex for a lot of banks to jump straight to, but into that discovery, let me go through some machine learning against the dataset, see what clusters, and then investigate what falls out of it. I don’t know what I’m searching for or if I’ll even find anything. That’s probably a bridge too far for a lot of the banks reading to start.

There’s also this confirmatory hypothesis testing. You need to get in the customer’s head here. What would make their life better in this data? Then let me go set up a test so that I can go do this. This is one of the things that is so valuable in a platform like Total Expert. You come with a whole bunch of user journeys out of the box, but your ability to manipulate and create new user journeys is to test, continually optimize, and look at what data falls out of the bottom. I know this is how groups of the Alloy Labs members that are Total Expert customers use it. If I do it this way, is that better or worse in performance than what I saw elsewhere?

You need a platform. You can’t afford to build your own. You can’t use something hardwired because the cost of experimentation needs to be dropped. Otherwise, you need to do this big ROI. It’s a big lift, it needs to integrate into my core, and it’s going to take us six months to implement. That’s too expensive. You’re only going to do something that’s safe.

I had a conversation with one of our large bank customers, and they’re using us for part of the business and not another part of the business. It was around that speed to market and speed to be able to iterate in testing a different communication, different journey essentially. I won’t name the vendor, but they have 4 or 5 dedicated developers. Every time they want to do something, they got to put it into a queue and bring on developers. By the time they can iterate and get feedback in the form of what happened, it’s 4, 5, 6 months down the road, and we’re just like, “The pace of progress is important for you guys right now.” At least, that’s how we think about it. You agree, I’m assuming.

A big part of this is there’s got to be a cultural change. That cultural change has to be modifying what we did is not bad that is positive because the likelihood that you hit the hole-in-one right out of the gate is so low. If you don’t go back and continue to modify and optimize, you’re squandering a huge opportunity for a sunk cost. The incremental cost of continuing to test and learn and optimize is low. That’s not an admission of failure. That’s a growth mindset of, “We’re going to do some stuff. We’re going to figure it out. It didn’t work as well as what we were doing, so let’s revert back to what we were doing.”

Do you think there are a lot of banks and credit unions that struggle organizationally to have a true growth mindset when it comes to evolving?

There’s a reason I still get to give that talk both on stage and with bank boards and management teams. I can tell things are lightening up. I’ve done it three times in a month, including at one of the big banking schools. I spent an entire day talking about how we teach bankers to have a growth mindset. The biggest challenge is, if you look at our business around lending, it is not about managing risk so much as we’ve gotten into this world where we try and eliminate risk. Not often do you point to the corner office and see the CEO, who almost always has come up through the lending operation in some form or fashion. Remember when Joe had that spectacular failure? That whole strip mall idea was brilliant.

That thing tanked, and we lost a bunch of money.

We gave everyone a trophy, and that was good. Culturally, we struggle with a growth mindset. Now you’re talking about a business that tries to control losses. It’s natural to have a selection bias for people who have a fixed mindset. If we tease them apart, these are two different things. For our core business, we’re not saying, “Anything goes, let’s throw it against the wall.” No.

Safety and soundness matter, and processes matter, but around the edges, we need to be comfortable taking some risks and understanding that taking a risk is for the purpose of learning. That’s what true value is. It’s not a traditional ROI that we’re going to look at, and we’re like, “We’re going to go do this experiment. Over the next years, it’s going to generate $100,000 in revenue.” It’s probably not an experiment if you can forecast all of that.

Take risks for the purpose of learning. Click To Tweet

If we look at it and say, “For a small cost of dollars and the potential downside risk associated with it, let’s assume it all goes bad, and we lose all those loans,” what’s the minimum amount we could test to say this is worth it? Once I have a platform and want to optimize my system, what number of leads in the mortgage business am I willing to risk to say, “What if we changed our onboarding and flipped it around that you could do an automated approval or a workflow?” We’re willing to do twenty of those mortgages and look at what falls out the other side. You’re recognizing the whole thing could blow up. That’s probably worth learning, and then you build your business case once you have data.

When you’re working with banks, credit unions, and some of the FinTechs out there that you guys work with, and you’re talking about transformation and projects, are you advising them on doing more proof of concepts, more tests, and more 90-day, 180-day type scenarios? Is that part of your advice package on a typical basis?

Alloy Labs operates a reverse accelerator. It’s reverse in the sense that what we’re accelerating is the development of new types of partnerships and what the use case is. Not about, “Let’s go teach a startup to be a better startup,” or even to go meet a bank. Let me give you a great example that came out of it. One of our four thematic areas is around healthcare for both small businesses and for consumers. Specific to consumers, we’ve been focused on aging tech.

One of our VC partners hasn’t been partnering with banks yet. They’re a direct-to-consumer platform called Carefull out of New York. What they do is help adult children or caregivers manage the financials of aging parents. They had started down this path of, “Can you actually map cognitive decline based on cognitive data?” It turns out you can. You start missing payments and doing double payments. It’s all of these other places that they can help highlight. Having lived with my parents when we first moved back to Minnesota, I can tell you that elder abuse is real. Their phone rings off the hooks.

What Carefull solves is that this is not just a lever you throw where Jason is on the outside to Jason is now with the full power of attorney. That’s what ended up happening, and it was not a smooth transition. There’s this gradual change where you’re moving from, “You have read access,” to now, “You have limited read-write access,” to, “You have full control,” over the span of ten-plus years. That’s the problem they solve. They came to the concept lab, and our banks were intrigued.

I’m intrigued. That sounds fascinating.

One of our smaller banks was all over it. They looked at this and were like, “We have an aging population base. It’s a huge issue for us.” Their approach was like, “Let’s play out what would a paid pilot look like for you. It’s not full integration. It’s going to get us the data we need.” For Carefull, they can prove that banks can use this and do something with it. They figured out what the minimum is for both sides to say this is a good fit. It worked stunningly well. Now they’re on the fifth iteration of it. Frankly, even after you’ve signed the long-term contract, you should still have a proof of concept mindset.

You’re learning. You’re constantly evolving.

This is not a core conversion where the project is done the day it goes live. It should be a perpetual proof of concept until you’re like, “This thing is just running and humming. There’s no more to be done. Let’s innovate on something else.” If you sign a three-year contract, you should probably think of that as twelve proofs of concept over that span of time.

It’s so mind-blowing when you think about it. It’s been driven in to a lot of the people in these organizations because maybe the cores and legacy technology that, “We’re going to buy technology, and that’s the destination. We turn it on, it works, and we don’t need to do anything.” That is not how your technology partners should be approaching.

I know you guys see that a lot. Here are a couple more questions here. You guys obviously see front and center, ground zero, a lot of digital transformation projects. If you had to pick a couple of key points of failure, and we’ve talked about some of this, where would you point most often are the failure points in some of these organizations when it comes to big digital transformation projects?

Two. The first one is always looking at cost savings. We love those cost savings because they’re the easiest to point to, “In our efficiency ratio, what’s the impact going to be in hard dollars saved?” Except not everything can be cost savings. You’re not going to cut your way to greatness, as I already mentioned. The other is there are just some foundational things that need to happen that you’re not going to attribute the ROI to. The two big ones are that you need to figure out data and how you free the data. Second, you need to figure out APIs. Those are two things that you cannot point to today and say, “Here’s what that direct impact is.” You need to go do it.

You have to do it.

The longer you wait, the worse it’s going to be. If your board says no, you need new board members.

It’s really that profound. If they’re not investing in freeing the data, having APIs set up, and those types of things, you’re sealing your own fate. It’s a question of time.

Exactly. One of the other places that banks overthink, “I need a full digital transformation strategy.” They’re going to spend so much time planning, and the world’s going to move so much further before they catch up. This was one of the big insights of our Robotic Process Automation Center of Excellence. This would be true of anywhere that you’re looking at, “Where are the places I can automate?” Don’t focus on the biggest impact things first. Start doing some of them. This was a big a-ha for several of the members. We did one in the labor savings that we’re able to redeploy and work on other things. We just kept finding the next easy things that quick hits added up a lot faster.