It’s common practice in the banking industry to design for the average customer. In this podcast, Carl Ryden shares why you should design experiences for the edges instead.
Jim Young: Hi and welcome to the Purposeful Banker. The podcast brought to you by Precision Lender, where we discuss the big topics on the minds of today’s best bankers. I’m your host, Jim Young. I’m joined today by Precision Lender, CEO, Carl Ryden. Personally, I’d love to have Carl on more of these podcasts, side note, he’s aware of this, but I usually save my invitations for topics that really, really excite him.
Last month, you hopefully got a chance to listen to our two-part podcast in which Carl and I talked about the exciting potential of artificial intelligence and machine learning in banking. Today, we’re going to talk about designing to the edges. What this concept is, why it’s valuable to banks, and why, quite often, banks do just the opposite.
Carl, let’s start with this TEDx talk, given by Harvard professor, Todd Rose, called “The Myth of Average.” A couple of weeks ago, you ambushed a group of people that just happened to be in the break room and basically said, “Hey, watch this.” You had a captive audience there. Can you explain to us why … Give our listeners a little bit about Todd Rose and that talk and why you felt like you had to share it with anyone you were coming into contact with for about a two-week stretch?
Carl Ryden: Sure. I stumbled upon it over the weekend. One of my personal passions is education. I volunteer a lot in that area. I think somehow through probably some AI or machine learning study, what I watch, what sort of TED Talks I watch, it came to me and they said, “Watch this one.” I watched it.
He’s a doctor in the School of Education at Harvard University. He has written several great books, the last of which was called, “The End of Averages,” I think is the name of it. Maybe we can put links to his books …
Jim Young: Absolutely, they’ll be in there.
Carl Ryden: … into his TED talk. I watched his TED talk. He was talking about how he started out with a wonderful story about fighter pilots and how, I think it was in the 1950s, there was the number of incidents and accidents were going through the roof. Jets had become really fast. They would crash or near crash at a rate never before.
Initially, they blamed the pilots. They said the pilots … “It must be pilot error. They can’t keep up with this.” Those sorts of things. Then, they blamed the training of the pilots. Then, they looked at the aircraft. It turns out there was no mechanical failure on the aircraft. The pilots were just as good as they were before.
Then, what they really started focusing in on was the pilot-to-machine interface, right? Which was, for the pilot, is the seat that they sit in and the cockpit they sit in. They said, “Well, maybe pilots have …” They were still using the dimensions when they would put a plane out for bid. They would say, “Here’s how big the cockpit should be.” They were still using the dimensions of a pilot, of the average pilot, from the 1920s.
They thought, “Well, maybe pilots have gotten bigger.” They had a study. They studied 4,000 pilots and measured ten different dimensions. The arm length, chest size, waist size, leg length, wrist size, all these things, and then calculated what the average pilot looked like.
I think they were heading down the road of, “It turns out the average pilot of the 1950s is bigger than the ’20s and they’re going to make a … We moved the average.” It was this enterprising young researcher, from Harvard, who had done this research before, and had probably a natural skepticism about averages. He went through. He was the guy who was tasked with measuring all these dimensions on the 4,000 pilots.
He had gone out and measured them. He said, “Well, if I take each of these ten different dimensions, chest size, head size, wrist size, waist size, everything, and I say the average is the middle 30% of the range, how many pilots are within that middle 30% for each of those dimensions for all ten?”
Jim Young: Right.
Carl Ryden: For height, the average height, I think, was 5’11”. It was 5’9″ to 6’1″ was, what you would call “average,” the middle range. It turns out of the 4,063 pilots they tested, exactly zero pilots were average in that average range, in the middle range, of all ten dimensions, is that every single one of them were unique and different.
Jim Young: There is no average pilot, in other words.
Carl Ryden: Yes. Then, he went even further and said, “If you narrow it to any three of those dimensions, there was .05% of pilots who had the same on any three of those dimensions.” This came to the realization that if you design for the average, you’re really designing for no one. He does a great job in his TED talk, highly recommend it, particularly if you care about education.
I said this to folks I know, chancellors of school and education, said, “Watch this.”, because he says, “Our U.S. educational system is actually exactly that. We design for the average student, age-appropriateness. It turns out when we design for the average, we actually design for no one.” I think there are now technologies and tools, AI and machine learning being one of them that allows us to customize that experience and design to the edges.
The reason I brought in everybody from here to look at it is what they came up was a solution for this, was adjustable seats, is allow the plane to be adjusted to fit the pilot. When they did that, their incidents and accidents went way down.
It also opened them up because they used to have to turn away talented folks who didn’t fit the average profile. There were no women pilots at that point. Now, they could bring in women. Now, they could bring in folks who were a little bit taller than before, folks who were a little bit shorter than before, but who still had the skills to be excellent pilots.
Overall, the efficiency, if you’re thinking in banking terms, the effectiveness, all these dimensions went up. Definitely the rate of incidents and accidents went way down, and what allowed them to do that was the adjustable seat.
I came to folks here and told them, I said, “Really what we build with Precision Lender is that adjustable seat for that relationship manager, customer interaction. How do we actually help put together the exact solution that worked for that customer and meets their needs, which really gives us dimensions to become more valuable, just like the pilots became more effective and more efficient and had less incidents and accidents.”
In the banking world, that would mean lenders close more deals. Right? They build stronger relationships. They build a better portfolio for the bank. They do so in a way that fits, right? Giving that freedom, those degrees of freedom into the process really helps.
Jim Young: That sounds awesome. I think you can listen to it and go, “Okay, sure, awesome.”, but I’m sure you’ve heard some very reasoned arguments from bankers on why they would want to include average prices of things or average deal sizes or these sorts of numbers that would, at least, on the surface seem like they would be very useful information for the bank to provide their relationship managers when they’re sitting around to make a deal.
Carl Ryden: There’s a couple flavors of this. One, I’ll talk about. We get folks all the time because we have a lot of data around pricing. Folks say, “Can you just give me the average spread of deals on this sort of deal in my area?” The other piece, that’s feeding averages back into the acquisition process, and we’ll talk about the problems with that.
The other one is, using averages from the back of the bank, another flavor of this, portfolio averages, from the back of the bank and again feeding those into the acquisition process. We’ll talk about a little bit about each of those, I think.
Folks ask us all the time, “Can’t you give us …” There’s services out there that a give-to-get service, where you give your data and in return, you get everybody’s data. What you get is a summary of it, the average.
Jim Young: For some benchmarks out there.
Carl Ryden: Benchmarks. For a commercial real estate deal in south Louisiana of that size, here’s the spread, that’s the average spread. One of the problems you run into with that is, one, and I was talking to a banker about this, who was actually a user of that service, I said, “How do you price, say, a line of credit? How do you price $100,000 line of credit that’s part of a deal with a $15,000,000 commercial real estate deal alongside of it is different than how you price $100,000 line of credit stand alone?”
They said, “Absolutely.” “Is that reflected in this data you’re coming back?” “No.” Then, there’s just data cleanliness stuff of the fee goes straight to the GL. It’s not tied to the record. The data’s just messy.
One of the things I hear from a banker on that subject is, “Well, we know our data’s a mess. We’re hoping everybody else’s is better.” I’ve heard this from multiple of the folks for fee-for-service. It’s also kind of one of these … It’s almost like a scheme. Everybody thinks theirs is crap, what they’re feeding in is garbage, but they’re hoping everybody else’s is better. I don’t know that that works out.
Even assuming it’s all right, and assuming the spread was right, and you get the average spread for a deal like that, what does that do? As a human, as the lender, the relationship manager on the front of that, very seldom do they start at the average and move up.
Jim Young: Right.
Carl Ryden: The average becomes the ceiling. The average becomes the places they start because I can justify the average. That’s the average. Then, they go down. What happens is, your pricing of your deals actually moves below the average. Everybody else’s average becomes your ceiling.
Jim Young: Mm-hmm (affirmative).
Carl Ryden: Then, you feed all that back in, dirty as it may be, and then it comes back, and now the ceiling moves lower. That’s not a good place to be. It’s a dynamic system that feeds back on itself in a bad way. That’s kind of a complicated way of explaining it.
My simple way is to take pricing and everything out of the equation and say, “Suppose you measure the average height of everybody …” I call this the door height problem. “Suppose you measure the average height of every person who walks into your bank.” You say, “On average, everybody’s 5’11”. You decide, for whatever reason, we can save you some money if we move the door heights to 6′.”
Now, you measure it a month later and the average will be 5’6″ because you fed that back into the acquisition process that affects the average going forward. You’ve introduced a feed back loop into the system.
Jim Young: Yeah, you’ve tainted the data.
Carl Ryden: You’ve tainted the data. What happens is now, it’s 5’6″. The worst part about this you folks doing that average thing was really cool and blindly do it again, now let’s move the door heights down to 5’7″ on all the … It keeps moving down. You see that sort of thing happen again-and-again. It’s a flavor of a death spiral.
Jim Young: I was about to say, it sounds a little bit like the version of a death spiral.
Carl Ryden: It’s the different flavor of the version of death spiral where understanding, “What are we doing? What are we trying to achieve?” What you do is instead of feeding the averages, you understand what’s important to the bank. Understand that fully, then build systems that allow you to understand what’s important to each customer. For each customer, handcraft the solution that works best for them, that allows you to elevate each conversation, and earn a better return, just like we talk about in the book.
We see that as a huge shift in how you think about things. We’ve seen so many folks go down. They feed in the averages and then all of a sudden, one of the common things I’ll talk to them is folks who have had it for any period of time, they say, “We have this system. We feed in the averages. Can you do that too?” I said, “Yes, I can do that, but we won’t do that for this reason.”
One of the things I’ll often ask them, I’ll say, “How’s that working for you? Has it gone up since you started doing that or has it gone down?” They’ll say, “Oh, the market’s really competitive. It’s gone down.”, invariably, because …
Jim Young: Right.
Carl Ryden: … you create that feeling that it is because you continue to tell them the averages are going down.
Jim Young: Yeah. Let me do a little devil’s advocate here because the story that Todd Rose tells is it’s the U.S. government, the Department of Defense that is telling the defense contractors, “Build an adjustable seat.” They get pushed back originally from the contractor saying, “Hey, that’s hard. That’s going to be expensive.”, all that sort of stuff.
If I’m a banker looking at that, what would keep me from saying, “That sounds awesome. Giving a good personal experience and really customizing things for each customer, but that sounds really expensive. I need to be efficient. I realize I’m not going to get everybody perfect by doing averages and that sort of thing, but it’s the best case scenario that I can do within the means of what I have to spend.” How would you counter that argument?
Carl Ryden: Well, one it’s a false economy. There’s two pieces I would counter that with. One is a false economy because if you look at even the Air Force data, they, in the worst month, they crashed 17 planes. Right? They were lucky in that when an plane crashes in the Air Force, they see a plane burning on the ground. They lost a pilot or they lost a lot of money. It’s there.
Jim Young: Right, I got you.
Carl Ryden: There’s visible, tangible …
Jim Young: Clear evidence.
Carl Ryden: … clear evidence that something went wrong. The problem you have in the bank is it’s not always … You don’t see what you don’t lose. You don’t see what you lose. Right?
Jim Young: Like you said, they can blame it on tough market.
Carl Ryden: Tough market. Whatever. There’s several deals that you lost there because of this or several deals that you drove to the bottom. Then, maybe even lost them, but you don’t see it in a big ball of fire on the ground. They were lucky enough to at least have that bit of feedback they could act on, but the other piece …
For them, it was a false economy like sure, the vendors are going to say, “It’s going to cost more to have a plane with an adjustable seat.” “Well, how many planes crashing does that equal?”
The math, I think, was easy for them once they look at the larger picture, not just the unit cost of a plane, which again within banks, you see that happen too. They focus on the narrow thing as opposed to the wider thing, especially when the cost incurs in one silo and the benefits inure to another silo.
The cost hits my budget, but the benefit hits his budget. It’s very hard to do that, which is often why this needs to be at the top of every CEO’s list because this is a blind spot that exists within the bank.
That was the first piece about, I would say, is a false economy. It was easy for them to see, maybe harder for a bank to see, so you really have to go look for it and make sure you look at the full picture.
The other piece I would say, I would flip it around and say, “If the U.S. government can find a way to be nimble and flexible, then by God, your bank ought to be able to.” I mean, on the scale of “nimble and flexible,” banks aren’t … You’re saying “bolt.”
Jim Young: Right.
Carl Ryden: The government, you saying “bolt scale” being the right, I don’t know who the far left would be, Rosie O’Donnell or something we’ll pick.
Jim Young: They might be turning around and cruise for a while, the U.S. Government would be turning around an aircraft carrier.
Carl Ryden: Correct. You’re probably more nimble than the U.S. Government in total. I think what it does, it takes someone with a unique perspective, like the lieutenant here. I think it was Gilbert?
Jim Young: Gilbert. Mm-hmm (affirmative).
Carl Ryden: Lieutenant Gilbert, who said, “Wait a minute, guys.” The good way he presented this was, he did the data, looked at it and said, “How many folks would you expect to be within the average, the middle third, on all ten dimensions?” Most folks thought, “Well, most of them, almost all of them.”
He says, “Factually, it was none of them.” That’s the sort of thing that really helps wake them up, shake them out of their groove and go, “Wait a minute, we might be missing something here.” I think you see that again-and-again in banks as well.
Jim Young: Then, let’s say you’ve convinced me that, “Okay, this is the way to go.” How do I get there? How do I get from where I am at a bank with whatever systems I’ve got. To be clear here, we’re not pitching this as by Precision Lender, what is it that’s going to make this sort of thing possible to be able to customize these experiences, but do it again, in a way that’s going to work within what my resources?
Carl Ryden: The concept here, I think, is true to Precision Lender, but is beyond Precision Lender. It’s true across a lot of things, even on the consumer side of things, that you do at the bank. I think you’re seeing some of these, I’m doing air quotes, “FinTech”-type companies finding ways to better satisfy customers. There’s a whole list of those, but with the tools, I do think AI and machine learning are going to be a big part of the tool set that helps you do this efficiently at scale.
I think, but early on, it’s simply having a commitment to it. Right? What you might do, is you can’t have one person handcrafting an experience for every individual, but you might can have that for a little while just to train and to learn which experiences are the best and then cloudify those and put them in at the edges.
I was telling you before we went on about a story I was reading and maybe we’ll put a link to this in the podcast notes. A great article, I think the lady’s in the U.K., wrote. In the article, she was talking about how banks have a unique set of data and insight around what their customers are doing. They treat it as really, really valuable, which it is. They put firewalls around it. They build the wall around it. They protect it from the outside world because if it got out, but they don’t do anything to act on that value and deliver that value back to their customers.
One example she gave, which, I think was an interesting story, Amazon, who monitors all their customers’ activities and what they look at, what they don’t look at, what they buy, what they don’t buy, which things they’re browsing. They have machine learning algorithms working all the time.
They detected someone at a house browsing for infant supplies and newborn stuff and that sort of stuff. They sent a package to the house, which is, “Congratulations on …”
Jim Young: On expecting or … Right.
Carl Ryden: … expecting your new child or the addition to your family, all this other stuff. Here’s the ways we can help you. Here’s some coupons to get you started, which was great. The problem is, the young girl who was at the house who was browsing for this stuff hadn’t yet told her parents …
Jim Young: Oh, no.
Carl Ryden: … that she was with child. What’s interesting about that, I think, is that Amazon figured out that the girl was pregnant before her parents figured out the girl was pregnant and acted on that. That’s not what we should strive for, but it also is you can see what’s happening in the wider world.
Take a bank though, and the lady in the article covers wonderful pieces of this. Suppose I look at your transaction history and I see what sort of things you buy. I see that you buy, you spend “X” amount of dollars on laundry service or internet service or whatever and you shop here. Well, there’s someone just like you who’s similar profile, who’s able to get what you get, but at a lower price.
Your bank could reach out and say, “Hey, you could actually save some money if you did this.” I think you’re going to see those sorts of experiences, those delighters. The bank is not just about the transaction processing. It’s about helping me manage my money, manage my finances, and get to a better point.
You’re going to see, there’s in Europe, I think, a thing coming out where banks have to provide that transactional data via API. You’re going to see a whole slew of the data analysis solely that can’t be held in secret within the bank. Other FinTech providers will be able to provide the experiences.
You can see one of them that’s interesting on the deposit side. digit.co. I think it’s digit.co. We’ll put the link in the … They actually monitor your checking account and automatically decide to save money for you.
Jim Young: Interesting.
Carl Ryden: Will text you and say, “Hey, I just transferred $100 over to your savings account.” You can …
Jim Young: You can opt in on that sort of thing.
Carl Ryden: You can text back and say, “Undo. I’m saving for this.” Got it. The AI learns from this, but the AI’s managing. They see your spending, manage your savings, help you make the most of it. I think that we’re going to have to find ways of creating value in that customer interaction and really designing to the edges because what you save will be different from what someone else saves. It’s not a simple Excel formula.
Jim Young: Sure.
Carl Ryden: It is learning based on your behaviors. I think we’re going to see more-and-more of that, both on the consumer side and in the commercial side throughout the bank.
Jim Young: I mean, as you’re telling me, I’m thinking about going back again to Amazon and these delighters as you go out and you buy a book. You look down there and you see, “People that bought this book also liked this book.” You go, “Wow! That’s kind of cool. I might be interested in that thing.” It’s, from a bank’s perspective, that’s just opening up more opportunities.
Carl Ryden: Amazon sends me … My problem now is …
Jim Young: You are a total sucker for Amazon.
Carl Ryden: I’m a sucker, yeah. My list of books that I’ve purchased, my backlog is huge now. The reason that keeps me from buying more stuff on Amazon is not that they’re not good or they haven’t figured me out, is that I can’t keep up with them. I haven’t finished the first book then I go, “How did you know I wanted to read that?” Bought it and started reading it on the Kindle. The next, I get, here’s another one.
“If you liked that, I bet you really like this.” I go, “Yeah, I do.” I’ve got a backlog of ten books now. That’s a great place to be, I think, from a customer supplier, customer relationship standpoint to be.
Jim Young: As opposed to saying, “You are a 40-something person from eastern North Carolina. Therefore, an average person …”
Carl Ryden: Only average folks.
Jim Young: Exactly.
Carl Ryden: Folks with your age from where you’re from.
Jim Young: Yeah.
Carl Ryden: What you’ll find is that they’re all unique. When we talk about designing for the averages versus designing to the edges, it’s really about designing for humans, right? Individual humans. Folks are consumers of stuff. We’re training Amazon, but Amazon’s also training us on what sort of experience to expect.
Jim Young: Yes.
Carl Ryden: When the bank doesn’t get me, and they don’t understand what I’m trying to achieve, when they should know them, “You know me. You have all of my accounts. You see all this stuff.” My expectation of you is becoming framed by my expectation of what Amazon and other folks make seem magical.
Jim Young: That’s a good point because we talk about you’re while you’re not competing just against the bank, you’re competing against everyone else and all the experiences they’re giving because you’re going to be judged based off of this.
Carl Ryden: I think there was a quote. I don’t know if it was in that article, but another one is, “Your bank needs to become a brand before these big brands become banks.” Googles and Amazons and other folks, it’s coming. I think you have to find a way to move closer to that.
Jim Young: Absolutely. All right. We’ll stop there for this episode. Carl, thanks for coming into the studio to chat. Let’s do it again soon. Remember, you can always find more information about today’s episode at precisionlender.com/podcast. We’ve got a lot of links to put in for this one. If you like what you’ve been hearing, make sure to subscribe to the feed in iTunes, SoundCloud, or Stitcher. We love to get ratings and feedback on any of those platforms. Thanks for listening. Until next time, this has been Jim Young and Carl Ryden and you’ve been listening to the Purposeful Banker.