Brian Stone 0:03
Hello, and welcome to The Buzz, a Bank Automation News podcast. My name is Brian Stone, and I am the Associate Editor at Bank Automation News. Joining me today is Shane Berry, Head of Customer experience at Mission Lane. Shane discusses the ways retail clients extract value from automation, how banks can best deploy messaging to help consumers and where banks have disconnects between their digitization and the customer expectations.Shane Berry 0:30
What I think is interesting is is the use cases for the consumer themselves and, and the sort of the increase in prevalence of of automated approaches to solving some of those problems, I’m thinking things like budgeting, or financial planning, right? You know, things that take in information and use machine learning models, or algorithmic automation to kind of help solve those problems, algorithmic savings accounts, things like that, or even, you know, customer service or chat bots. But there’s actually a really simple use case that I like to use that illustrates how we might think about automation to real value to our customers. And its limitations, the use case of automatic payments. So in our case, automatic credit card payments, right? It’s hard to think of a more cut and dry use case with obvious value to our customers in the company. So for the customer, auto pay ensures on time payments and prevents that downward cycle of debt and fees. And default that sometimes, you know they’ve experienced in the past has led them to the place that they are. But for many of our customers, it’s actually not that simple. Because they have dynamic finances. And it makes enrolling in auto pay and ceding control kind of challenging for them. They might have income that comes in at different amounts, at different times of the month. They might the long wait times of payments, when the money’s like left their account, but hasn’t hasn’t paid their credit card can actually lead to a lot of sort of uncertainty in their liquidity. And then you know how much money they have left over there after their other expenses can really vary greatly from week to week, and month to month. So those are some of the challenges around something as simple and automated as auto pay. When you look at that, I have a person that we met in one of our user research studies that sort of exemplifies this, we’ll call her joy. It’s not her real name, but it works. She was a 30, something mother of five, from Florida that we met in a user research session, she had multiple sources of income, she worked part time for Home Depot, but also did a bunch of gig work, home cleaning DoorDash, things like that. So she had a complicated set of bills and expenses. And because of that she had a really like bespoke system for budgeting that she had developed that involves handwritten notebooks with lines that were drawn connected specific income sources to specific bills, you know, really kind of complex and, and unique way of managing this budget. And she was really comfortable with that approach, when she used auto pay in the past that actually bitten her when the bill came out earlier than she thought. And it caused overdrafts and missed payments on other bills. So I, I often think of joy when I think of, you know how we might consider a customer’s point of view when thinking about automating things. So there’s a lot of ways in which we can sort of mitigate the ways that sort of reduce the sense of a customer’s control, giving them the ability to shift their due dates, so they match better with the income. Having them, you know, if they need, like separate accounts for bills, so they have some assurance that the money won’t get pulled out of the wrong spot for the other bill. You know, basically splitting splitting accounts, the ability to change these payments, and the amounts that they spend each month. Because a large percentage of our customers, you know, and customers in general, they, you know, they need to pay unique amounts each month based on what’s coming in and going out.

Brian Stone 4:17
When, when it comes to financial institutions, banks, credit unions, what can they do to sort of help bridge this gap that we’re seeing between automation and use billpay as an example, and attempting to augment the customer experience to I guess, you know, not only make them more comfortable, but you know, to sort of like you said meet them where they are.

Shane Berry 4:44
Yeah, I mean, it sounds really odd. It’s a great question, and it’s not always cut and dry. I think when I think about solving this, and it sounds really obvious, but it’s staying in touch with the customer. It’s understanding what their expectations are what Their needs are, and really having having, you know, conversations with them to really identify areas that were ceding control, it just isn’t going to work for them, or where it where it might. And so there’s this sort of two pronged, I would say one is, it’s just that it’s the research and conversation with with your actual customer base to really identify those areas. The other thing is to just really keep in mind that the, the approach to automation should often feel fairly invisible to the customer. Customers really, you know, one way to have them maintain a sense of control is to really layer in an idea of humanity, if they feel as though they’re interacting with another person. You know, that’s, that’s an understandable interaction, if they feel as though they’re, they’re interacting with an algorithm, or, or, you know, a digital experience of some sorts, the sense of lack of control can be exacerbated can be multiplied. And so the idea, I think, is to say, stay in touch with your customers, and understand what their experience has brought in terms of their expectations, and their context. And be, always think about adding that layer of humanity, between the algorithm between the automation and the customer themselves. You know, a simple example here is, is, you know, customer, customer service. chatbots. Right, thinking about the use of those is a really fantastic way to maintain consistency in your customer experience, and also reduced costs for us, you know, for us operators, but when they’re when the interface is between the customer directly on the chat bot, that’s a recipe for frustration at this point in time. And you know, and, you know, until, until they become sufficiently advanced that we can’t tell the difference, and they’re all passing the Turing test. But In the meantime, one way to sort of solve that is to actually have, you know, kind of aquarium will knowledge bases that are natural language, I’m thinking Chet GPT, that’s actually being used by your customer service agents to act as the intermediary between what’s coming out of the of the automation, and being being given to the customer. So really, it’s just wrapping that automated experience and a level of humanity and control.

Brian Stone 7:34
Yeah, um, so one of the things I wanted to ask about is, you know, all of that makes sense. And we’ve done a number of articles on use of AI, such as chat GPT. But when it comes to, like, let’s say, a larger financial institution, if you’re one of the big banks, you know, Bank of America comes to mind, Wells Fargo, things like that, you know, you can use that chat GPT. But it still has its limitations as well. How would you recommend like a large bank approach, injecting that layer of humanity? Whereas, you know, chat, GBT, I was using it the other day, has trouble pulling, you know, recent information for my thinking, you know, after 2021 is where it starts to get a little fuzzy with new information. So in a larger financial institution, how, what what tips could you give to sort of bridge that gap and, and sort of meet automation with the customer experience there?

Shane Berry 8:36
Yeah, that’s a great question. And honestly, I don’t have a cut and dry answer. I think because of how sort of nascent this this sort of approach is, right? How new this is, for a lot of us. I go back a little bit to what I was saying a little bit before in that a human intermediary between the automated outputs and our customers, I think is absolutely critical at this point in time, do I think it will be necessary forever? Probably not. But for now, I think it is in order to solve you know exactly what you just brought up the gaps in information. And more critically, not not gasp but but language learning model may just not actually have context into our customers or what it means to be human in these particular cases, empathy, right? Those things are absolutely critical. For our helping our customers maintain, like I said, a sense of control in managing their finances. So I think in the near term, it’s, it’s not about using automation, be it language learning models, or, you know, simpler machine learning models or, you know, simpler algorithms to really make customer decisions for them. It’s not about narrowing customer choices via these automations. Right? It’s about broadening their choices and giving customers context. It’s about getting the right information in front of the customers at the right time so that they can make the info choice with minimal cognitive load. So it’s really like it this point that the tactical way, I think, you know, to really do that is to is to have that human layer in between. In the near term, the idea is to retain a certain degree of control, and enable the customers to get the value from the automation that they otherwise wouldn’t. If they’re spending if customers are spending hours a day thinking about this stuff, and we can reduce that to minutes, you know, not eliminate it entirely, but reduce it and make those minutes matter more in terms of confidence about the decisions in terms of the correctness of their decisions, that’s where we’re providing real value with automation.

Brian Stone 10:40
Let’s go back to your your Julie example. She suffered from, you know, overdraft due to a bill, you know, that came out earlier than expected. How important is it for banks to have different layers of alerts in place to let customers know, you know, this might happen? How customizable should banks make their you know, controls, say card controls? You know, overdraft protection, things like that, how do you balance those alerts with keeping sort of that human layer that you had talked about? Because, obviously, you know, it’s as easy as a as a push notification, you know, someone might have additional questions. So, what can banks do to sort of meet their customers there as well,

Shane Berry 11:27
you know, I think it comes back to some very, like fundamental UX principles. You know, user experience principles are really around, pads out of any any alert, right? Like, if you think about the case of joy and AutoPay, it really comes back to kind of really fundamental UX principles around how you offer customers choice in any interface that they’re interacting with. If enjoys case, she got an alert that says, we’ve taken out your auto payments, there’s nothing for her to do is nothing actionable there. And that may already be too late for her to avoid overdraft charges. If instead, your alert says, you know, you’ve set your automatic payment, or even in this case, you know, maybe we could we could brand it something different so that the real value was clear to you know, your backstop payment is about to be withdrawn. Would you like to change it or edit it? Is this still the information that you want to be submitted? Is this still the amount from the right from this account? And then, you know, take that and allow that to be a place where it could be a simple confirmation of yes, that’s exactly right. Or, you know, this month, I need to do something different, I need to move the date, I need to switch the account. It comes from wrapping that all in I know, that sounds very simple and very obvious, but in human language, right? Wrapping it all in, in, in in copy that that sounds like a person wrote it is these little simple fundamental things are important. The use of you know, I talked about this all the time. And it’s such a simple thing. But the use of contraction saying we’re with a with an apostrophe instead of a we are little things like that, change that copy to humanize it. And then you give folks, like I said, pads to edit or maintain a sense of control over the the subsequent actions. And that’s at least the starting point. You know, it doesn’t solve all the problems, but that’s a starting point.

Brian Stone 13:32
You’ve been listening to the buzz of bank automation news podcast, please follow us on Twitter and LinkedIn. And as a reminder, you can rate this podcast on your platform of choice, be sure to visit us at Bank automation news.com.

Brian Stone 0:03
Hello, and welcome to The Buzz, a Bank Automation News podcast. My name is Brian Stone, and I am the Associate Editor at Bank Automation News. Joining me today is Shane Berry, Head of Customer experience at Mission Lane. Shane discusses the ways retail clients extract value from automation, how banks can best deploy messaging to help consumers and where banks have disconnects between their digitization and the customer expectations.Shane Berry 0:30
What I think is interesting is is the use cases for the consumer themselves and, and the sort of the increase in prevalence of of automated approaches to solving some of those problems, I’m thinking things like budgeting, or financial planning, right? You know, things that take in information and use machine learning models, or algorithmic automation to kind of help solve those problems, algorithmic savings accounts, things like that, or even, you know, customer service or chat bots. But there’s actually a really simple use case that I like to use that illustrates how we might think about automation to real value to our customers. And its limitations, the use case of automatic payments. So in our case, automatic credit card payments, right? It’s hard to think of a more cut and dry use case with obvious value to our customers in the company. So for the customer, auto pay ensures on time payments and prevents that downward cycle of debt and fees. And default that sometimes, you know they’ve experienced in the past has led them to the place that they are. But for many of our customers, it’s actually not that simple. Because they have dynamic finances. And it makes enrolling in auto pay and ceding control kind of challenging for them. They might have income that comes in at different amounts, at different times of the month. They might the long wait times of payments, when the money’s like left their account, but hasn’t hasn’t paid their credit card can actually lead to a lot of sort of uncertainty in their liquidity. And then you know how much money they have left over there after their other expenses can really vary greatly from week to week, and month to month. So those are some of the challenges around something as simple and automated as auto pay. When you look at that, I have a person that we met in one of our user research studies that sort of exemplifies this, we’ll call her joy. It’s not her real name, but it works. She was a 30, something mother of five, from Florida that we met in a user research session, she had multiple sources of income, she worked part time for Home Depot, but also did a bunch of gig work, home cleaning DoorDash, things like that. So she had a complicated set of bills and expenses. And because of that she had a really like bespoke system for budgeting that she had developed that involves handwritten notebooks with lines that were drawn connected specific income sources to specific bills, you know, really kind of complex and, and unique way of managing this budget. And she was really comfortable with that approach, when she used auto pay in the past that actually bitten her when the bill came out earlier than she thought. And it caused overdrafts and missed payments on other bills. So I, I often think of joy when I think of, you know how we might consider a customer’s point of view when thinking about automating things. So there’s a lot of ways in which we can sort of mitigate the ways that sort of reduce the sense of a customer’s control, giving them the ability to shift their due dates, so they match better with the income. Having them, you know, if they need, like separate accounts for bills, so they have some assurance that the money won’t get pulled out of the wrong spot for the other bill. You know, basically splitting splitting accounts, the ability to change these payments, and the amounts that they spend each month. Because a large percentage of our customers, you know, and customers in general, they, you know, they need to pay unique amounts each month based on what’s coming in and going out.

Brian Stone 4:17
When, when it comes to financial institutions, banks, credit unions, what can they do to sort of help bridge this gap that we’re seeing between automation and use billpay as an example, and attempting to augment the customer experience to I guess, you know, not only make them more comfortable, but you know, to sort of like you said meet them where they are.

Shane Berry 4:44
Yeah, I mean, it sounds really odd. It’s a great question, and it’s not always cut and dry. I think when I think about solving this, and it sounds really obvious, but it’s staying in touch with the customer. It’s understanding what their expectations are what Their needs are, and really having having, you know, conversations with them to really identify areas that were ceding control, it just isn’t going to work for them, or where it where it might. And so there’s this sort of two pronged, I would say one is, it’s just that it’s the research and conversation with with your actual customer base to really identify those areas. The other thing is to just really keep in mind that the, the approach to automation should often feel fairly invisible to the customer. Customers really, you know, one way to have them maintain a sense of control is to really layer in an idea of humanity, if they feel as though they’re interacting with another person. You know, that’s, that’s an understandable interaction, if they feel as though they’re, they’re interacting with an algorithm, or, or, you know, a digital experience of some sorts, the sense of lack of control can be exacerbated can be multiplied. And so the idea, I think, is to say, stay in touch with your customers, and understand what their experience has brought in terms of their expectations, and their context. And be, always think about adding that layer of humanity, between the algorithm between the automation and the customer themselves. You know, a simple example here is, is, you know, customer, customer service. chatbots. Right, thinking about the use of those is a really fantastic way to maintain consistency in your customer experience, and also reduced costs for us, you know, for us operators, but when they’re when the interface is between the customer directly on the chat bot, that’s a recipe for frustration at this point in time. And you know, and, you know, until, until they become sufficiently advanced that we can’t tell the difference, and they’re all passing the Turing test. But In the meantime, one way to sort of solve that is to actually have, you know, kind of aquarium will knowledge bases that are natural language, I’m thinking Chet GPT, that’s actually being used by your customer service agents to act as the intermediary between what’s coming out of the of the automation, and being being given to the customer. So really, it’s just wrapping that automated experience and a level of humanity and control.

Brian Stone 7:34
Yeah, um, so one of the things I wanted to ask about is, you know, all of that makes sense. And we’ve done a number of articles on use of AI, such as chat GPT. But when it comes to, like, let’s say, a larger financial institution, if you’re one of the big banks, you know, Bank of America comes to mind, Wells Fargo, things like that, you know, you can use that chat GPT. But it still has its limitations as well. How would you recommend like a large bank approach, injecting that layer of humanity? Whereas, you know, chat, GBT, I was using it the other day, has trouble pulling, you know, recent information for my thinking, you know, after 2021 is where it starts to get a little fuzzy with new information. So in a larger financial institution, how, what what tips could you give to sort of bridge that gap and, and sort of meet automation with the customer experience there?

Shane Berry 8:36
Yeah, that’s a great question. And honestly, I don’t have a cut and dry answer. I think because of how sort of nascent this this sort of approach is, right? How new this is, for a lot of us. I go back a little bit to what I was saying a little bit before in that a human intermediary between the automated outputs and our customers, I think is absolutely critical at this point in time, do I think it will be necessary forever? Probably not. But for now, I think it is in order to solve you know exactly what you just brought up the gaps in information. And more critically, not not gasp but but language learning model may just not actually have context into our customers or what it means to be human in these particular cases, empathy, right? Those things are absolutely critical. For our helping our customers maintain, like I said, a sense of control in managing their finances. So I think in the near term, it’s, it’s not about using automation, be it language learning models, or, you know, simpler machine learning models or, you know, simpler algorithms to really make customer decisions for them. It’s not about narrowing customer choices via these automations. Right? It’s about broadening their choices and giving customers context. It’s about getting the right information in front of the customers at the right time so that they can make the info choice with minimal cognitive load. So it’s really like it this point that the tactical way, I think, you know, to really do that is to is to have that human layer in between. In the near term, the idea is to retain a certain degree of control, and enable the customers to get the value from the automation that they otherwise wouldn’t. If they’re spending if customers are spending hours a day thinking about this stuff, and we can reduce that to minutes, you know, not eliminate it entirely, but reduce it and make those minutes matter more in terms of confidence about the decisions in terms of the correctness of their decisions, that’s where we’re providing real value with automation.

Brian Stone 10:40
Let’s go back to your your Julie example. She suffered from, you know, overdraft due to a bill, you know, that came out earlier than expected. How important is it for banks to have different layers of alerts in place to let customers know, you know, this might happen? How customizable should banks make their you know, controls, say card controls? You know, overdraft protection, things like that, how do you balance those alerts with keeping sort of that human layer that you had talked about? Because, obviously, you know, it’s as easy as a as a push notification, you know, someone might have additional questions. So, what can banks do to sort of meet their customers there as well,

Shane Berry 11:27
you know, I think it comes back to some very, like fundamental UX principles. You know, user experience principles are really around, pads out of any any alert, right? Like, if you think about the case of joy and AutoPay, it really comes back to kind of really fundamental UX principles around how you offer customers choice in any interface that they’re interacting with. If enjoys case, she got an alert that says, we’ve taken out your auto payments, there’s nothing for her to do is nothing actionable there. And that may already be too late for her to avoid overdraft charges. If instead, your alert says, you know, you’ve set your automatic payment, or even in this case, you know, maybe we could we could brand it something different so that the real value was clear to you know, your backstop payment is about to be withdrawn. Would you like to change it or edit it? Is this still the information that you want to be submitted? Is this still the amount from the right from this account? And then, you know, take that and allow that to be a place where it could be a simple confirmation of yes, that’s exactly right. Or, you know, this month, I need to do something different, I need to move the date, I need to switch the account. It comes from wrapping that all in I know, that sounds very simple and very obvious, but in human language, right? Wrapping it all in, in, in in copy that that sounds like a person wrote it is these little simple fundamental things are important. The use of you know, I talked about this all the time. And it’s such a simple thing. But the use of contraction saying we’re with a with an apostrophe instead of a we are little things like that, change that copy to humanize it. And then you give folks, like I said, pads to edit or maintain a sense of control over the the subsequent actions. And that’s at least the starting point. You know, it doesn’t solve all the problems, but that’s a starting point.

Brian Stone 13:32
You’ve been listening to the buzz of bank automation news podcast, please follow us on Twitter and LinkedIn. And as a reminder, you can rate this podcast on your platform of choice, be sure to visit us at Bank automation news.com.