- Where banking leaders see the biggest opportunities for automation in the banking industry and the benefits banks expect to realize from their automation initiatives
- How banks are performing on their automation initiatives and key operational concerns
- Adoption of AI-enabled and RPA-enabled automation technologies, and barriers to new tech implementations
- The expected impact of automation on workforce size and composition and organizational readiness to govern expected workforce transformation
Janet King (00:06):
Good morning, good morning, good morning, good morning. Thank you all for joining us today and for coming so early in the day. We really appreciate that. Hope Breakfast is good. It's always a nice draw to get everybody around the table. So my name's Janet King. I am the VP of Research for American Banker, the host of today's this week's conference. And this is Bill Hincher joining me, who is the director of banking and financial services industry for UiPath. Thanks for being here, bill.
Bill Hincher (00:35):
Good morning. Thanks, Janet.
Janet King (00:38):
So we're excited to be here today to present some research every year at American Banker. We conduct a number of original research initiatives on a wide variety of topics. And this particular piece of research on the fully automated enterprise is actually an annual initiative. And UiPath partnered with us on the initiative this year. Our goal really was to explore how banks are automating and utilizing internal resources and external partners and vendors really to drive automation within the workplace. So this was a piece of research that we conducted during March. We had a hundred senior, mostly senior level leaders from banks participate in the research, and all of them were screened for involvement in our knowledge of their organization's enterprise automation initiatives. So this is a mix of banks. We had a good portion of medium to large size banks, but we also had a mix of some credit unions and retail banks. So a good balance sample I would say. One last thing I just want to mention just kind of from a housekeeping standpoint is what we mean by enterprise automation. So we kind of framed this up for the people participating in the research, so everybody would be answering the questions from a singular point of view, if you will. And so this is how we defined it as the application of technologies including AI and RPA to empower end-to-end business process automation. So really thinking about it is going beyond single point to point integrations and really looking at an organization's ability to automate and flow operational and post operational data between multiple applications and systems so that they can automate complete business processes. So it could be on premise tech in the cloud hybrid or both. And it covers everything from application integrations to data integrations. Anything to add there? Anything I missed?
Bill Hincher (02:38):
Nope, you covered it.
Janet King (02:39):
Great. All right. So we wanted to start out by talking about AI and chat GPT because it's been such a frequent topic in the news and it's been a frequent topic here at the show as well. Everybody's talking about that these days. And if you Google Chat GPT and just see how many things searches come up, it returns 771 million results. At least that's what it did when I googled it, but it's probably even more now compared to a couple weeks ago. And if you look for how many people have tried Chat G P T, it returns according to the latest available data Chat GPT currently has over 100 million users, and the website currently generates 1.8 billion visitors per month. So it's something certainly we've been writing a lot about at American Banker. I think I came up with something like 32,000 results just on our own brands. So it's a really frequent topic. And even this morning when I was thinking about how can we engage an audience at 745, we tried out chat GPT, and luckily it said we have seven of the 10 recommended elements in our presentation today. So I think we're going to hopefully keep you guys engaged. We are missing three, we're missing music, but we didn't think you guys would want to hear YMCA this early in the day. We're missing icebreakers, but two truths and a lie seemed like a little much for people you don't know that well. And then of course I tried to get Bill to tell a cheesy joke, but
Bill Hincher (04:08):
I'll spare you from the bad jokes.
Janet King (04:10):
So we decided not to do that. So just show of hands, how many of you guys have tried chat GPT, either personally or professionally? Almost everybody in the room. Have you used it? How many people pers for personal use just to have fun and how many people are actually using it professionally at work? Interesting. There's been a big debate about how to use it in editorial rooms and newsrooms and media brands. So it's something we certainly talk a lot about a lot at American Banker. Bill, how about you? How have you tried chat GPT?
Bill Hincher (04:41):
Yeah, professionally, I'd say probably more than I would like to admit. It's doing a lot of, I think the legwork for me. But yeah, I think I'll pull the audience. Is there any preference for Google versus chat GPT, Google Show of hands, Google Bard? Nope. Going once chat GPT, everybody. So yeah, the Coke versus Pepsi debate I think is where we're at.
Janet King (05:02):
Absolutely. So I think what's interesting about the whole AI an automation conversation is that many people don't realize that the two are really very closely linked. And this is just a quote from McKinsey, which kind of captures that, right? And you can read it, but I'll read the first part of it, which just says few would disagree. We're now in the AI powered digital age facilitated by falling costs for data storage and processing, increasing access and connectivity for all and rapid advancement in AI tech. These technologies can lead to higher automation and when deployed after controlling for risk can often improve upon human decision making and speed and accuracy. And they go on to say that AI can potentially unlock 1 trillion of incremental value for banks annually. Huge opportunities. So can you talk to us a little bit about that bill, that link between AI and automation?
Bill Hincher (05:59):
Yeah, I mean, as we see the market, right, AI and automation are invariably kind of drifting together, and this has been true for a number of years. When you think about the way they come together in the same context, you have AI acting as the brain automation, acting as the muscle. If you think about brains without muscle, both are effective in kind of what they do, but the combination of the two are a pretty powerful combination. And that's really what we're seeing in the market as well. When a lot of companies have been focused on automation, purely on the automation side for a number of years, they've achieved significant benefits early on. But now that they're introducing AI as well, you really see those benefits and those results starting to supercharge. So they are a natural pairing and we continue to see the convergence of the two topics in the market.
Janet King (06:47):
And I mean it's certainly backed up. It's certainly backed up by a lot of value. The value proposition for AI and automation is pretty clear and pretty compelling. So we wanted to take a look at some of the data we collected, which you can see here on these slides to see where leaders in the banking sector see the biggest opportunities for automation. And what we found is that fraud detection and prevention in management tops the list at almost 60%. And that's been the most cited use case for automation for a few years that we've been conducting this research. And it makes sense, right? Technology has become quicker and more efficient, and AI offers greater flexibility and control to address what is really a top challenge for banks. It's a huge challenge, big security landscape out there, and I expect banks will continue to focus on that use case moving forward. I would expect that to continue to rise to the top. But we're also seeing other use cases that banks frequently identifies top automation opportunities, and one of them is loan application processing and underwriting, which was mentioned by almost half of our respondents this year, which is up from 34% last year. So statistically that was a pretty good jump making it number two in our list here. So I'm curious, bill, what you see as some of the reasons why we might be seeing an increase around interest in loan application processing, underwriting.
Bill Hincher (08:14):
Yeah, I really think the answer's twofold, right? If you look at the macro perspective of the last few years, we've seen loan volume application through the roof. Banks were struggling to keep up with the volume, they simply just couldn't hire enough people, quickly enough skilled people quickly enough to handle that volume. So that's really been the macro perspective of the last couple years. And I think that now that loan volumes are continuing to subside a little bit, especially on the mortgage side, you see bank executives recognizing now is a great time to prepare for the next wave of volume. So they're almost looking at it from a lull in the market perspective of, hey, now might be a good time when we're dealing with subdued volumes to look at this problem so we can better position ourselves down the line and prepare for the next wave and to allow us to capture more of that volume quicker and more efficiently. I think the second piece of that answer is really on the technology side. Document processing within a bank is not something necessarily new. It's been around for a number of years, but what we are seeing is a increasing capability from a technology perspective. You have the ability to to process increasingly complex documents that have previously ever been achievable, and you can do it within a single platform like UiPath. So you have the introduction of document processing, really intelligent document processing that allows you to handle multiple variations of complex documents and then inject that into an automation workflow so that you're connecting both upstream and downstream, and you can do it within single platforms such as UiPath and some others that exist in the market. So I think when you combine this macro perspective with technology enhancements, it's really kind of a ripe opportunity for the application of both technologies.
Janet King (10:02):
Absolutely. And we also see another frequently cited opportunity here for automation is around personalized customer experiences, challenging area to be sure. But what we saw is that a larger number of banking leaders this year talked about personalized customer experience being what they see as a key automation opportunity compared to last year. So it was like one in three this year compared to about what, 24, 25% last year. So how do you see banks specifically using AI and automation to help with that specific use case?
Bill Hincher (10:36):
Yeah, I think when you look at the customer experience, you first have to start with the modes of communication. So everybody knows that the primary driver of customer communications are coming in via calls, and there's a lot that automation can do from a benefit standpoint on the call side. But if you look at some of those second and third tier drivers of customer communications, you start talking about social media interactions in the form of messaging, you talking about emails that some banks are dealing with millions of emails inbound from customers on an annual basis. And when we get into all of these written forms of communication, you need specialized AI to start to be able to process that, right? So if you look at a freeform text, email or social media interaction, everybody in this room is going to write that message in thousands of different varieties depending on the day. So variation becomes very challenging. We're dealing with highly unstructured text in those communication channels, and that's where you need some specialized AI. UiPath has a product as well to help you start to make sense of all of that messy data. So you're talking about understanding the intent, what is the customer actually asking about things like sentiment analysis, are they particularly frustrated? All of those are sort of inputs at the beginning of that journey. And then you can start to extract the data from that email to then process in an automated or more efficient manner downstream. So you kind of combine all of these intricate data elements to then frame a picture around what the customer is actually trying to engage the bank for. And that's leading to improved customer experience. You're applying automation as well as some of these advanced AI capabilities to really reduce the response times on customer experience type of scenario. So we see continued interest from clients on that front and a lot of benefit being delivered as well when you start to introduce multiple different technologies for specifically the customer experience.
Janet King (12:30):
Yeah, that's an exciting application I think for AI and automation that I'm really excited to see where that goes. So much data and so much unstructured data. To your point, another top automation opportunity that our respondents this year cited was risk mitigation. So we had about 35% who also see risk mitigation as a key opportunity for automation within banking. Can you share some examples for us of how banks might be using automation to manage and mitigate risk?
Bill Hincher (12:57):
Yeah, I think, again, you have to look at the macro picture. You see obviously the regulatory environment continues to place increased scrutiny amongst banks, especially with war in Ukraine and sanctions screenings, things like that increasing on Russian entities. Banks are facing a lot of challenges in terms of maintaining pace with those increased workloads in terms of customer screenings and things like that. And if you consider even a lot of the banking upheaval we've seen in the us, a lot of deposits have changed institutions very quickly. So you have an increased burden on those new account opening processes of which compliance checks and screenings are part of, right? And we want to make sure that there's not a overburdened compliance team processing through those compliance checks because as we all know, they're central to how a bank manages their risks. So really using automation as a capacity enhancer to help process through backlogs, help the compliance team become more efficient, allow them to spend more time on analysis and less time on actually data gathering, data collection, things like that. So again, banks are continuing to face challenges on the screening side, the compliance side, and that's kind of the natural fit for automation as well.
Janet King (14:14):
Thank you for that. So we wanted to then take it down to the individual bank level, so we have a pretty good sense of where they think automation is going to be leading the industry, but why are banks automating at the institutional level and what do they see as the business benefits really of automation? So three things really rise to the top of the list when asked about those business benefits. So it's reduced operating cost, improved customer experience, and of course better workforce optimization and efficiencies. So three kind of meaty opportunities there. And how do you think banks are quantifying the ROI there?
Bill Hincher (14:54):
I think when you look at the first and the third, they're really kind of linked. And if you consider periods such as this where you're seeing economic slowdown, automation is a natural lever to reducing or impacting costs from an operational perspective, when we're not seeing growth, obviously we need to manage costs. But I think from a workforce optimization standpoint, we're not necessarily talking about FTE takeout or headcount reduction at many of our clients. I think the opposite is typically true that banks, even despite economic slowdown periods, still have ambitious goals. And one of those goals is often growth and they want to enter new markets, but they want to do it without necessarily needing to continue to increase their workforce. So really it's about a productivity play. How do you make your current staff more productive so that you can A better manage costs, but B, then be more effective in terms of successfully scaling your business without necessarily scaling your operations.
(15:52)
So that's kind of the cost reduction as well as the workforce optimization play. And then obviously customer experience. I think this is kind of the no-brainer, right? I think if you pull the audience across the room, every single person in this room is probably going to have knowledge of some customer experience initiative taking place within their enterprise today. And when you look at how customer experience often is a differentiator in products, we're often seeing highly commoditized product types across different banks, different financial services organizations. Customer experience is often the reason that people continue to do business with your bank. So again, capturing the hearts and minds of executives tying into initiative automation or program initiatives at a banking level. Again, I think that's kind of where the focus becomes how are we delivering value at the organizational level and using automation to achieve some of those goals.
Janet King (16:45):
Do you see automation occasionally? Is it helping banks to scale out some of those initiatives more quickly too? Go to market faster?
Bill Hincher (16:54):
Yeah, absolutely. I mean, speaking from a UiPath lens in an RPA lens, right? RPA typically is delivering quick time to value in relation to traditional I technology products. So if you think about the automation delivery time on sometimes, or I'm sorry, the development delivery time on a I development, things like that versus RPA development, you're typically going to see quicker time to value on some of those RPA components in comparison to traditional LED tech technology led development, doing things like API. So yeah, absolutely, you can think about quicker time to value and being able to enter new markets quicker through automation or low-code automation.
Janet King (17:35):
That's great. So given the various benefits and opportunities bankers see for automation, and there are certainly plenty, we wanted to understand what kind of progress they're making on the journey, and we may not have yet completely reached the tipping point on where we are, but we're making substantial progress across the industry. So what you can see here is that four in 10 are telling us that they're well on their way and they're starting to see some measurable impacts in some areas, if not a majority of areas of their business. And then we have an equal number about four in 10 who are still early in the journey with the balance, actively learning and gathering information. So what do you think banks who are maybe a little earlier in the automation journey can learn from their more experienced peers?
Bill Hincher (18:27):
Yeah, I think they get out first. And for foremost, you get out what you put in. So I think if you treat it as a strategic tool, then oftentimes the output is something that impacts the organization from a strategic standpoint, for lack of a better term. If you treat it as a science project, in many instances we find that the returns on that investment are traditionally difficult to quantify or fairly small. So after a number of years it can be challenging. So I think for those who are starting the journey, it's really important to tie it to strategic initiatives so that you are aligned and clearly can measure what some of those expected outcomes are. I also think that from an executive alignment standpoint, it really needs to be a top down initiative. We need all members of the organization kind of pulling in the same direction. So I think the bottom up is bottoms up process is not to be overlooked. It kind of needs to be a combination of the two, but without that executive alignment and support, I think typically we see that programs will continue to perhaps struggle to gain their footing after a number of years. I think early returns will be found, but then true scale comes through a really an enterprise-wide type of program.
Janet King (19:50):
So I think we're seeing a lot of active interest and initiatives going out. And I think the good news is that a majority of the banks, at least that we spoke to, are investing in automation with three and four telling us that they plan to implement new technologies in the next 12 months. And I think to your point, that really underscores that even in times of market uncertainty or economic uncertainty, banks are really leaning into technology investments. And we see that in all of the research that we've been conducting this year, the majority of banks are increasing or at least holding those investments steady as we go through the next 12 months. So especially investments that they really understand are critical to supporting the continued transformation of their business. And it validates I think, the importance of automation in that journey. But with any technology implementation, there's always some concerns, there's always some challenges that go along with it. And some of the things that floated to the top of this survey were concerns about data privacy, talent or lack of skills to see these things through to fruition cost and then trying to find the right partner or solution. So with that in mind, we wanted to understand how are they moving forward, what are they doing? What approaches, what technologies are they leaning into in the journey to really become a fully automated enterprise? And one data point that I think is particularly interesting is this one here, which looks at the extent to which banks are leaning into end-to-end solutions versus single point solutions. And we see a bit of a bell curve of sorts. If this was a bell curve, you could kind of imagine that with a minority falling into the buckets on either end and with most acknowledging a reliance on a single point solutions for many, but not all. Notably though, two intent told us they expect to replace their existing point solution with an end-to-end automation solution in the next 12 months, which I think suggests that this balance might be shifting. So we'll be keeping an eye on that in future years. But Bill, what advice would you give banks who are looking to make investments in AI and automation to ensure that they're getting the maximum return from those investments?
Bill Hincher (22:07):
Well, I think first, if you consider this data point side by side with the prior data point about where banks are at in terms of their maturity, I think they kind of naturally dovetail. I think what we typically see first is those banks who are newer on their automation journeys are typically starting out with point-to-point automation solutions, you know, can put this in the bucket or category of task-based automation. They're focused very specifically in certain corners of an organization as opposed to looking at the larger picture of where that task sits within a larger end-to-end workflow. So I think the banks that we've engaged with from a client perspective who are further along or responded in a more of the maturity side of the scale are now looking at more of those end-to-end automations. And that's really where we've seen continued value being driven across the organization is around end-to-end automation that combines multiple capabilities, multiple technologies to facilitate different customer journeys. We talked earlier about loan application processing and document processing. That's reminiscent or kind of a good example of what we see from an end-to-end automation perspective where you have the need for automation to handle much of the process, but you need to introduce things like specialized AI to facilitate some of that document processing. So that's where we see really that supercharged ROI and that return on investments start to be realized is through the introduction of multiple technologies to handle or focus more on an end-to-end as opposed to just point solutions that are focused on parts of the business. We want to impact a larger portion of the organization and also customers when you look at the customer journey, we want to really make an impact on some of those per scenarios as well. And that's really kind of where you need more of a end-to-end automation focus.
Janet King (24:05):
That's great. Thank you for that. So we want to take a few minutes and open this up to questions from the audience. Abby will be walking around with a microphone if you have a question, so we can hear you, but let's give that a minute. Great. Abby, we have a question in the corner over here.
Audience Member 1 (Michael Bernard) (24:35):
Hi, Michael Bernard. I'm a industry Analyst. This is great and well presented. And I know the presentation focus main mainly on automation in processing and operations. Were there any insights uncovered in the field of software testing automation or QA testing tools or automation processes or more generally agile processing agile frameworks in general that are helping banks to bring improvements to market faster?
Janet King (25:13):
We didn't specifically ask about the software development process in this piece of research, but can you speak to maybe how you see Yeah, automation really working there?
Bill Hincher (25:22):
Yeah, when we talk about I think the multiple technologies within a single platform, UiPath actually is uniquely positioned within the testing market as well. So you think about the natural correlation between RPA automation and what testing functions within a bank do, right? They're essentially clicking through a process to make sure that it works. So UiPath is taking a different lens. I think the testing market is sort of a legacy market, hasn't seen a lot of innovation in the last number of years, and UiPath is one of the newer testing tools that does exist in the market. So when we take that story to our banking clients, they're typically very receptive because they're kind of now looking at it through a low code approach of how can we redefine this testing function within the bank? And those typically will, the benefits are typically quicker time to be time to development on the testing side as well as reduce cost and as well as increased coverage. So testing has been traditionally very limited. You're talking about testing limited sets of data, and now through quicker time to value can effectively test more scenarios. So those are the benefits that we're seeing in the testing side of the market.
Audience Member 2 (Kate) (26:40):
Hi, I am Kate Davis Green, I'm in the fraud and identity industry, and it was really interesting in your studies how fraud prevention and mitigation really rose to the top when you think about ways to improve and automate. But it was interesting in the slide that says why automate it really dropped to the middle of the pack and things like cost prevention went to the top. And I was interested to hear your theories behind that as to why that happened and if it's because when compared to things like cost mitigation that just accelerates.
Bill Hincher (27:18):
Yeah, I think when I speak to customers about benefits and value, the tougher things to sometimes explain are those areas of things like fraud and risk avoidance where it's a little bit of a let's prevent what's going to occur. It's a little bit hard to realize on the balance sheet as you can imagine, because talking about a future event and may or may not occur. So when you think about pure cost takeout, of course the conversation gravitates in those directions because we can measure them today. It's a little bit of a intangible type of conversation. You can't necessarily feel fraud prevention all the time. So I think that's one of the challenges and perhaps why you see the distinction between the two data points.
Audience Member 3 (Alberta) (28:08):
Good morning, Alberta Monte far with SoftTech. We're a nearshore service provider. Question on, have you seen any significant differences in organizations that are leveraging products like yourself to do citizens development type activities outside of the technology, traditional technology organization and the value that they're providing to that or the challenge that they're bringing about right to the enterprise versus organizations that are strictly driving this from a technology with a partnership with a business to drive automation? What are you seeing in that?
Bill Hincher (28:49):
I think in order to be successful in citizen development programs, there needs to be a traditional automation foundation that program builds upon. So I think about citizen development as a way to scale the benefits of automation across the business. It's not necessarily the driver of that scale, I'd say. So I think the foundation needs to be set by your traditional automation development programs. There needs to be the muscle developed on that side. The operating model needs to be developed from a centralized standpoint so that you can then manage and roll out that citizen development program in a governed way with a lot of efficiency. So I think it's you got to start there and then kind of gravitate or see the business mature into citizen development. So that's kind of the, I'd say a natural maturity that I typically see the most successful automation programs taking.
Janet King (29:45):
Excellent. Any other questions from the room? Any final thoughts?
Bill Hincher (29:53):
I'm just surprised we didn't have a Chat GPT question.
Janet King (29:58):
Wait, maybe we do have a question spoke to soon over here.
Audience Member 4 (30:01):
So this is the tough love category. So I sometimes hear conversations like RPA is an expedient, it solves a spot problem, it doesn't get to root causes of systems, so it might be a band aid on a structural problem. How do you respond to that kind of chatter?
Bill Hincher (30:20):
Well, there's a lot of ways to respond. I'll say that if you just look at the pipeline of any IT department, the pipeline is long and there's always projects to be that the business is asking for. At some point, there's going to be a threshold at which you cut off those priorities. So you have projects that are going to get IT funding, and they're going to want to be done a certain way. They want to create APIs and things like that. There's going to be projects below that line that aren't going to be delivered by IT in the current fiscal, but the business is still demanding it. So there's this need of, is there a quick solution if you want to call it a Band aid, perhaps there's an opportunity to apply. It's not necessarily a bad term. I think there's opportunity if you want to take look at RPA to address a problem for a year or two before it can actually get to it, that's an opportunity there because the development time on RPA is quick, quicker than your traditional automation. So I think that's kind of one element of it. It's just looking at those IT priority lists, and then it's about giving the business what they need today as opposed to 12 months down the line. So you can do both, right? There is room for both.
Audience Member 4 (31:48):
So the paraphrase is perfect, is beginning.
Bill Hincher (31:52):
Exactly. And if you want to take a perfect approach, how long have banks talked about core modernization and integrations and how much of those programs have actually been delivered? So there's always this, it's a sort of ever present problem of we're always chasing this goal and the IT priorities, the asks from the business are never going to subside, so maybe there's a quicker way to do it. And we compare that with more traditional approaches. So there's opportunity for both in that conversation.
Janet King (32:27):
Excellent.
Audience Member 5 (Rakesh Sharma) (32:28):
Rakesh Sharma, So on the use cases you had up there, the upstream type of use cases like loans, fraud prevention and personalization. But before you can do any of that stuff really well. So those were downstream, upstream is the data issue. So none of that, those use cases work very well without good quality data. So can you talk a little bit about automation in terms of getting good quality data?
Bill Hincher (33:04):
Yeah, I think when you look at data, the data picture, typically one of the biggest challenges is combining multiple different data sources. One of the biggest applications of UiPath and RPA in general is really connecting into some of those disconnected data sources that exist within an organization. So historically, your legacy based applications, your mainframes, things like that have been a bit difficult to ingest into any sort of workflow because they're very difficult to pull data out of. So I think from my perspective, that upstream process can be better facilitated when we're able to connect different data sources, some of which are legacy and difficult to interact with. I think there's kind of an easy way to do that through RPA and automation and then downstream kind of moving down that data chain a little bit more. There's opportunity. I'm not saying that RPA is the necessarily the only tool that does it, right? But if you think about data transformation, data reconciliation, that all can effectively be done with automation or RPA as well in addition to other tools that exist within the market. So it's just a matter of choice or preference in terms of what do you want to do with that data once you have it from some of those difficult data sources, including unstructured data sources, right? So we talked a little bit about documents, emails, those are some of the most challenging data types and often the data types that banks haven't even been able to get to yet, right? So you think about 80% of this organizational data residing in essentially stuck or kind of trapped within unstructured data sources as well. So opportunity on that front, I'd say.
Janet King (34:51):
Thank you. Great question. Any other questions in the room? Going once, going twice. I just want to thank everybody. Remind me, what's your booth number here at the event? Do we know?
Bill Hincher (35:05):
We just had the breakfast.
Janet King (35:06):
Just the breakfast. Sorry. Thought you had a booth. You can come and find us later. He'll be around. Look for Bill. Look for me. This full report is also going to publish on American Banker next week. So if you are a subscriber, you can find that in our research section on American Banker. And I just really want to thank UiPath for being here and for hosting this breakfast. I hope you guys learned a lot and I hope you have a great day. So thank you so much. Thank you.