.jpg)
Build What’s Next: Digital Product Perspectives
The process of developing digital products and experiences can be a daunting task organizations often find themselves wondering if they are solving the right problems the right way hoping the result is what the end user needs. That’s why our team at Method has decided to launch Build What’s Next: Digital Product Perspectives.
Every week, we’ll explore ways to connect technology with humanity for a simpler digital future. Together, we’ll examine digital products and experiences, strategic design and product development strategies to help us challenge our ideas and move forward.
Build What’s Next: Digital Product Perspectives
How AI is Transforming Design Teams in Financial Services
Join Jason Rome and Mike Adam in this episode of "Build What's Next" as they explore how AI is revolutionizing design teams in financial services. They discuss the impact of AI on design, research, and product teams, covering topics like psychological safety, process changes, the downsides and upsides of AI, and how to integrate it into team workflows. This conversation is highly relevant for design leaders navigating the evolving landscape of AI.
Episode Resources:
Jason Rome on LinkedIn: /jason-rom-275b2014
Michael Adam on LinkedIn: /in/mikeadam/
Method Website: method.com
JPMorgan Chase: jpmorganchase.com/
You are listening to Method's Build what's Next digital product perspectives presented by GlobalLogic. At Method, we aim to bridge the gap between technology and humanity for a more seamless digital future. Join us as we uncover insights, best practices and cutting-edge technologies with top industry leaders that can help you and your organization craft better digital products and experiences.
Speaker 2:All right, everybody, welcome back to another episode of Build what's Next. I'm your host, jason Rome, really excited for the conversation we're going to have today. Mike Adams, who I've known for a long time leader in the design space, he and I recently ran into each other at a networking event and got into a pretty deep, nerdy conversation about the impact of AI on design teams, on research teams, on product teams, and I said, hey, this seems like a great podcast. So excited to talk to Mike, whose teams he's been using it personally, his teams have been using it. We're going to get into kind of a culture, psychological safety process, the downsides of AI, the upsides of it, how he's teaching his team. So I think it'll be really relevant for design leaders today and a bit of a different topic than we've talked about. So, mike, we'd love you to just give a little bit of background on yourself and then we'll jump right into that. Thanks, jason.
Speaker 3:Well, I've been in UX for about 18 years at this point in time, so mostly around financial industry. So I've gotten a chance to do every kind of banking, wealth finance, wealth management, as well as business side of the house, so it's been a pretty broad opportunity to learn about the customer from every angle. Just a little pre before that, I actually started as a graphic designer, found my way through 3D animation, but didn't want to work 80 hours a week or 100 hours a week, so I found my way into UX after, that.
Speaker 2:Yeah, it's funny. You mentioned graphic design. I was talking to another designer recently and he said you know illustrators, graphic designers, you know they're the ones that have kind of taken some of this AI stuff the hardest in terms of like, where people are like, oh, I can just get an icon, I can get an illustration, I guess. Starting generally, where are you and your teams starting to adopt AI across the design landscape? Maybe some of the tools or other tools you've seen with, whether it's at work or playing on the personal side, like what's your thought on the state of it and how's the team going so far?
Speaker 3:A lot of what we're doing now is experimentating, you know, like getting an opportunity to try it out more than it is like applying it to the process completely. But I'd say the biggest opportunities are in the process. Like we have a system right now to create a product brief, a project brief, out of all the meetings that we've captured or all the teams chats or whatnot, rather than having to spend weeks creating one where like build one and then tell me what I'm missing. So, yeah, it's great, like in the process moments. Other spots are a little on the outside edges of what's allowed inside of a, of a restricted bank, but we'll use it to generate an image for a PowerPoint so that we can get a consistent image without having to do a photo shoot. Yeah, that's not going to go to a live experience, but it gets us the influence or the stories that we want to make.
Speaker 2:Yeah, there's a. You know there's an interesting piece there, right? Because when I work on design, I always talk about, you know, designing to build something, which is what a lot of people think of. You know the final pixel pushing we have to do right. Designing to learn something, to validate, but then designing to tell a story and to share a vision, and I've always thought I've gotten resistance sometimes on designers. I'm like, hey, we don't know the requirements yet, we haven't talked to users yet. I don't want to put pen to paper. But you mentioned using design to tell a story. Like, in your mind, how does AI start to open up? You know more opportunities to use design and visuals as a stimulus to tell stories and get people engaged. Like are you seeing teams embrace it more now that it's a little faster to do some of these things? Like you mentioned creating an image, you said embrace.
Speaker 3:I think it's been really, really slow to embrace. To be fair, the finance industry is slow to any software increase, so it's a little slow in that category, but I do think it's great. In things like I want to put a bunch of data into a place and figure it out. Data into a place and figure it out. If I'm telling a story around the legal ease of a product, the LLM we have internal at Chase right now is a. It was taught by all the all the legal documentation, regulatory environment, so you can get that stuff from it. So it's a great more of a buddy right now to get your data correct and get it organized.
Speaker 3:I think the next thing we're going to be doing is like being able to tell better stories. I used it just recently to tell a story about the value of UI inside of the inside of a product that we're working on, and it was not because I needed it, but more because 12 or 13 people were trying to write this one story. I'm like, well, rather than edit that, why don't you write your own version and I'll let the LLM mash it all together and figure out where it ends up, and I'll let the LLM mash it all together and figure out where it ends up.
Speaker 3:And it got us, I'd say, 75, 80 percent of the way there, after just that mashup, where in previous times that would have taken I don't know weeks of us kind of what did you mean by this word and what do you mean by that?
Speaker 2:So yeah, normally in a financial services institution that's at least like 10 PowerPoint decks and 37 meetings to be able to get to that. You know you mentioned kind of slow to embrace. I think some of that, you know, when you and I are talking, some of that's maybe uncertainty, some of that's resistance, some of that's fear of people I've talked to. But you know, for you as a leader, how do you view the balance of? You know, hey, we're, we're experimenting, we're taking it slow to stay compliant, we're learning how to use this. I mean, man, it feels like every day that there's something out about. You know all the security concerns with it, right. But how do you balance like, hey, let's take it slow, let's be thoughtful about it, versus are we missing opportunities because we're not embracing it? Like, how do you think about that? And then just talk a little bit more about the mindset of the team and like the emotional side of this as well.
Speaker 3:That's good. In the space I'm in right now, I'm in zero to one, so meaning like it's all net new and we're supposed to move fast. Yeah, so when I think about my answer to my current role, it's a little different than a lot of the bank industry that's not moving as fast. Like if you're managing a product that's been around for a long time, you may not feel as much pressure to change or like I can get there, I can take my time. New account open to you. I can get there, I can take my time. New account open to you. But in like a zero to one space.
Speaker 3:A lot of the challenges that we're having is I want to move faster, I'm being asked to move faster, but we don't have access. So we're looking for pressing those moments. Where can we go a little bit farther? So a lot of that is giving them, the team, permission Like, hey, it's okay, go try it out. We okay, go try it out. We're not going live with this thing. Go a little farther. Maybe we can go find and talk to a group that has got special access and find out what's the wink and a nod to let us use it. So a little bit of a permission to go find the next thing.
Speaker 2:And then just for how the team is set up now, do you guys separate UX UI designers from UX researchers? Is that hybrid or are those pretty separate roles today? Ux UI designers from, like, UX researchers Is that hybrid or is that are those?
Speaker 3:pretty separate roles today. They're design, research and content are the three separate roles? Okay, Though they're all reporting to the same group, Like they all report to one leader.
Speaker 2:How I think a two-part question, because I want to explore, like, how do these roles start to evolve and change into more hybrid roles with the advent of AI, or do they at all? But, like, across those different groups, like walk me through, like the emotion, the uncertainty, like which of those groups feel helped the most by AI. Like which of those feel like, hey, you know, is this going to replace what I do? Or like, hey, it's going to do a bad job at what I do. Like what's the just, what's the emotional sentiment?
Speaker 3:And whether it's your designers or other designers in the community that you talk to about these tools. It's good. I actually think the content teams have taken the most out of it. Like I think they'll like the opportunity of exploring through the tool a little bit more. Maybe because it's words already. They're more comfortable with putting something in there and putting it back out, asking for it to be a standard.
Speaker 3:I think researchers are the next on the list because they're finding little ride alongalongs like hey, we put the information through it as almost like a second opinion to see if it comes out in the same spot. But they're also paused at finding out a research AI could do the research for them. They like going into the field. They don't want to lose going in the field. They don't want to have an AI replace them. So I think there's a little resistance there. I think the top right now is the designer, but them. So I think there's a little resistance there. I think the top right now is the designer, but I feel like designers are like two animals there's the one that likes to do the craft of design and one that is using design to solve problems. I think the designers that solve problems like great new tool. I'll just throw it in there next to everything else, and I will. I'll just be that much better, I think, designers that are more craft focused or like I'm.
Speaker 2:I like doing the delivery work. I think they have the most kind of fear of losing their, losing their place. How, um, I mean, how are you navigating those conversations with folks? And you know you mentioned we're experimenting a lot, we're taking things slow. You even said they're like, hey, this is just another new tool, but, like you know what's? Uh, what are those conversations look like? What do those feel like with folks?
Speaker 3:You're going to make me tell my secret. I can listen, so no, I'm not. So, yeah, there's a lot of people have started using Loom or Stream to record their material, and I think it's actually the best thing for designers, because designers aren't great at explaining themselves, and how are you going to hop from doing a design and not being able to explain it all the way to telling an AI to help you? And so I'm encouraging them to create little streams, seven minutes of like, hand this off to the tech partner, or hey, if this is the second time you're reviewing a leadership, see if they'll take the stream so that they can understand it from the second time. Yeah, so it's just an interesting step that they're having to take to actually practice articulating stuff, and they're really embracing the idea, but I hadn't told them that that was because they're going to have to start, you know, basically being able to tell the LLM what to do. Interesting.
Speaker 2:Yeah, it's funny you mentioned that story because I think it's always such a balancing act of like how much you know you mentioned the craft right, but how much of a designer's job is solving the problem versus how much is the craft, versus how much is explaining the solution they got to and getting people bought in? And you know I both worked in financial services. That's one of the hardest things sometimes is getting people bought in. I remember I had a wonderful designer on my team earlier in my career and we're working on a project together and you know absolute whiz at design and could get to the answer really quickly. Absolute whiz at design and could get to the answer really quickly. And I could look at it as someone that you know new design and I could appreciate the heuristics and the thought he put into it. But the designs always fell flat and I told him I pulled him aside. I said, hey, how much of your job do you think is designing versus explaining the design? And he said 90% designing, 10% explaining. I was like it's probably closer to like 30% designing and 70% explaining, especially as a consultant and being able to defend that and it was a. It was kind of a light bulb moment.
Speaker 2:I love the book Communicating Design Decisions, but I think that's that's such a key thing, especially because you know, if there's anywhere, I would guess an AI would hallucinate. It's probably trying to explain a design and the decisions that it's making and we'll talk about craft in a minute but I think forcing people to be able to articulate it because the best designers I know, you know they can put just enough stimulus in front of a stakeholder to have a really good conversation. You know it's like, hey, this is a solid C plus, but I know it's going to at least get people to start giving feedback and start talking and I've separated my ego completely from this output and we're just going to have a great conversation about it. I think that's a difference between senior and junior designers that I see a lot of times. So using Loom is a really I hadn't thought about that. By the way, I love when designers use Loom to send something to me, because I listen to a lot of podcasts and books and then I can just watch the explanation on like 1.5 or 1.75 speed. It saves me like a 30 minute meeting down to like three minutes, but it's wonderful because I can go back and watch it and get that.
Speaker 2:You used the word craft earlier. I kind of want to come back to that. What do you see as the risk? Potentially, where does AI like, I guess? What are the risks of AI when it comes to craft?
Speaker 3:And then, I want to take this a couple different places. I was thinking about this the other day that there's two different crafts. There's like the real artist side of craft, which most of them are trying to create a style, and I think we've seen that AI can copy style. So there's a little bit of fear on that side of the house. But a lot of what the UX world is is not that type of craft. A lot of that is more like getting stuff delivered and I was.
Speaker 3:I don't remember who I was talking about this with, but we were having a talk about how, over the course of UX maturing, we've got a lot of not end deliverables but like outputs across the project or the process. I think putting more energy into those moments along the way versus investing in the end output would be a really good way to like think about craft right now for us, because that end output in the near future probably could be you could be replaced for that part but then the stuff in the middle, the thinking that goes into the product, is where I think there's a lot more opportunity for us to kind of like grow our craft around understanding the customer better, influencing the decision that's going in there. Maybe it's like here's how you can experiment, because now that tool can make 12 of these tomorrow, so why not try it out?
Speaker 2:Yeah, Do you see any risk with AI? You know, over homogenizing designs and so you end up with like everything kind of looking the same at all, and is that a good thing or a bad thing? Like how do you think about that?
Speaker 3:I think, yes, I think it's going to be very likely. So probably a little bit of what we've been doing as an industry in UX is probably getting there already, because there's so much that's like, why would we reinvent a radio button? It's been established for a long time.
Speaker 2:We're all using the same Tailwind components and we're all on Angular React now, anyways, and we're, you know, using the same five or 10 base design system. So there's a lot of that happening right now.
Speaker 3:But I do think that's where okay. So let's let the AI do that. But the designer space is that 20% where you still need to show value of your product and make people be wowed. I was having this conversation the other day. When was the last time on a project you got to spend time in the wow which? When was the last time on a project you got to spend time in the wow, which is what designers want to do? Yeah, with AI doing the 80% of the lifts, like that's where we want to invest our money. I'm going to make that amazing versus I have to deliver everything. I just need to make sure that that part is the amazing part.
Speaker 2:Yeah, there's so much there. I mean the other thing and you and I talked a little bit about this with product teams where I've seen, you know, product folks start to use AI to write PRDs or write epics and stories. I've played with it myself and I've, you know, played with vibe coding and design tools. You know, I think one of the big risks I see is these tools tend to they have some behaviors. That's almost like a junior designer, junior product person. In some ways is they confuse quantity with quality. So there's way too many features. The epics are too long, the PRDs are too long or, in the case of design, there's just too much going on. And one of my favorite quotes about design has always been you know, a design is not done when you've added everything to it. A design is done when you've taken everything away. And so have you seen AI potentially like overdoing it yet? And how do you make sure that teams don't just accept that, that they still have that like critical eye for like streamlining and taking things?
Speaker 3:out. I'm really hoping that AI is in that earlier phases, still right now. Or, and if you think about how AIs would be used in the way you were talking about being a vibe code early on, early on would suggest, okay, we'll go do that 12 more times. There was a point in time where I would take the product partner's PowerPoint presentation at like, thank you for doing the one thing, and I'll take that and figure out how to make it work. But if they can go all the way through a vibe code and making it work like build 12, 13 of them, let's have, I can bring in 12 users and you guys can play with them. Uh, it kind of goes back to the usability test where you had your product partner who wanted thing in this exact spot but nobody would find it and you're behind the mirror watching the product partner realize they're wrong. So you kind of got back to now.
Speaker 3:I didn't have to design it. Isn't there a? Isn't there like a little story yet about if you, if you um, pay me, you pay me to do the design, versus I, I pay me to watch you do the design?
Speaker 2:it's like we're in that space somewhere yeah, um yeah, I, I, uh, you haven't. You haven't had a product person show up with like a stitch together like miro or figma or a vibe coded prototype for your design team, yet.
Speaker 3:Not vibe code, but we've had some, some product partners bring some stuff together.
Speaker 2:Yeah, I think that that's one of the interesting things with just hybridization of roles, right Cause it opens up the ability to do this stuff, but I think still having specialties and still having people know where their final strengths are and where they can optimize. But I think, again, it goes back to you know, I would much rather look at a early clickable prototype to convey a concept versus a 10 page word document at the end of the day because, like I still see a lot of companies you know doing long PRDs, brds I sit and I observe and I watch the teams talk about these documents and there's this, you know, illusion of alignment. That's happened and I can look at all these teams and be like, no, they're not aligned on this at all. But I think when you put a prototype in front of somebody early, it's easy to get like, oh what, this is what was in my head, this is what's in my head, those, this is not bad at all, but now we can at least have a conversation versus a document.
Speaker 2:I think it's easy for people to walk away from that meeting with multiple interpretations. So I think that's a huge opportunity for teams. I want to go back to this 12, 13 options. Parallel prototyping because I find that's always been one of the hardest things for teams to want to do is, yes, people want to AB test, but especially early. I find most design teams want to take one design concept and iterate on that, versus take three design concepts and battle royale them against each other and kind of compare them. Are your teams already doing parallel prototyping or are you seeing AI as a new opportunity to maybe enable this parallel prototyping? And how do you get teams to think that way, or do you already have teams thinking that way?
Speaker 3:I don't think I'm seeing it as much as I'd like, and this might be a factor of so much of what we do already kind of has been decided that by the time a designer's thinking about it, there's only one or two options to go down. Actually, we did a project with Method, where the goal was to create as many or exhaustive ideas as possible in a short amount of time, and so humans did it, not AI did it, but some of the jokes were this was like a whole year ago, before AI existed, you know.
Speaker 2:So it was a long, long time ago.
Speaker 3:But the great thing about that was we were pretty I don't even want to call it blue sky Like we had good requirements, we knew we were trying to accomplish, but it was like how can you solve this? How many different ways can you solve this? And I think, having once again giving the designers permission to go explore, we would. I would go into a room and say I didn't see any that had giraffes or balloons inside, Like I was really pushing where you're going with this thing, yeah, and pushing where you're going with this thing, yeah. And I think that's OK, because sometimes the best stuff comes out of those 12 or 13 mistakes. And in design school you were expected to do 50. And there was a story of the first 10 were the obvious ones and the last 10 were the ones that were like you were exhausted of ideas and those were the greatest ones came out of. So it wasn't that that you needed quantity, you need to get all the bad stuff out.
Speaker 2:Yeah, and I find that's a really hard thing, especially in financial services, right, when you're trying to do some of those forced variations because you're going to run into somebody that says, oh, that's never going to pass legal compliance, right, oh, we would never do that, that's not on brand for us. And that's been a tough conversation. I'd be curious how you've navigated this, but that's been a tough conversation for me to navigate of, hey, this design isn't for development, this design is to optimize learning. And so, yes, it's not technically feasible, like, no, we're not going to give someone a cash bonus for this behavior to their account. But we're trying to figure out. You know, I've done this a couple of times, right, where I forget what, what journey or flow we were trying to optimize for a bank. But you know, we were trying to get people to sign up for something like auto pay, right, and we're like, well, you know, could we pay them five bucks, you know? So we avoid all the calls to the call center because people are sloppy payers and they forget. Someone was like, well, we would never do that. And I was like, what? Like? Because then that gets us to like the underlying emotion. Like is not using it because someone has a feeling of control or is it because of a lack of incentive? And that's a really valuable feedback piece to learn.
Speaker 2:But it was really hard to get stakeholders to test something that we wouldn't do to pursue the value of learning necessarily, you know, is another one. We did a bank website redesign. Know is another one we did a bank website redesign. Uh, and that was fun because we person, we purposefully gave the stakeholders um a mild, a medium and a spicy version to show them, like how the vault, the brand, could evolve and this was a big bank. So, yes, they chose the mild version. That's fine, but you know, at least showing that and getting feedback showed like different variations of it. So how have you gotten, or how have you been able to get stakeholders to accept like testing something that we won't do in the sake of learning um about?
Speaker 3:the user stakeholders are still a little, a little challenge with this. They still want to have something good. The mild, mean spicy does work for us. I actually use 1990 movies are great. I have great analogies in them.
Speaker 3:So faster to the loan over the top truck driver arm wrestler he. He's a scene where he says he turns his hat around and it's like a switch turns on his head. And I used to use this at the point of when you go from discovery to design. So when you're in discovery, you got your hat forward, you're a truck driver, you're explore, do whatever. Nothing's wrong, like the words somebody won't allow for.
Speaker 3:This should not be in your repertoire because we're so early in the process, you have time. But when you hit the design process, you've got final requirements we should be moving into like real serious. So you turn the hat around, the switch turns on and it's. Are you going to deliver on time? Is it going to meet the user's needs? You get ada requirements, but the idea being that first half, with the truck driver hat forward, you, you should be able to do those explorations and have the freedom. And it might feel weird, but finding different analogies to tell product partners similar things is how I've been like let let the product process flow, let me know what your ideas are. I want your ideas too. So it's kind of given a creates a space to get um people to feel like it's okay to be wrong or okay to fail in that space yeah, and that's interesting, you know, because you know 2010s, right?
Speaker 2:um, you know, being here in charlotte working with a lot of banks, right, we did a lot of design thinking workshops, we worked in a lot of innovation labs and I feel like the you know the that we've got two weeks to really diverge is is no longer a luxury that anyone has, and so I think, to your point, finding meaningful ways to bring those moments of freedom into the design process is really hard and I really like that. You know AI. One of the hardest things about design sprints back in the day was the idea that we're going to get to a prototype really quickly to be able to test with users. It's the whole sprint idea and people were exhausted after five days of being able to do that. But I think AI opens up the ability to do these mini design sprints and think differently. But I do think it is a. It's a stepwise change because I think, to your point, with organizations and budgets and the focus on productivity and efficiency metrics, people have the hat backwards most of the time, if you will, and they don't have permission to like switch it around. So I think that's a really I like that analogy. I'm going to steal that from you, probably. So I like that a lot.
Speaker 2:I don't recommend watching the movie. It wasn't that great. Okay, no, I am, I am. I am all one for metaphors. You talked about AI and research being a helper and something that you know. You like to use that and it brings in kind of different ideas. How are you using AI as, like, a helper in research? How are your teams using it and what's the upsides Like, what are the benefits it's bringing? But then are there any gotchas or any places where it seemed like it was helpful but it actually? Oh, we missed empathizing with the user.
Speaker 3:Here there's the biggest one and there's a couple of spots in the process of research, but it's at the end right now we're looking at. So we've done a lot of research at a bank. The forging experience of going and finding the previous research before you starting new is like a really real step in the process now. And I think AI is a spot where if you can get we're working on now to get a use case in our AI where all the research is there so you can ask it questions and find the information you need. But there's a I think it's really going to be good at the beginning of a project to say what has already been answered.
Speaker 3:My concern for that is will it replace the time we would normally spend with, hey, why don't you come in stakeholders and let's review and like really fall in love with the problem statements and the challenges that our customers have and so we can solve them? Like I think you can easily just dump it in there and say, oh, the AI has it, I don't need to know it. Yeah, if you think that you're solving, if you're solving that problem like we're not, we're not solving the problem of getting you to understand it. We're solving the problem when you need it later and I could see that falling into it's in there, so I don't need to join the.
Speaker 2:And because that is a huge issue and you know, for me as an outside partner to some of these companies, where some of the companies I've worked with, like I've you know, I've worked there longer than some of the stakeholders sometimes and so they'll be an event and whatever comes out of that event is the truth, whereas they don't have strong like baseline hypotheses of like what do we already believe about the user and what are we hoping to learn and what will change our mind.
Speaker 2:And I think that's one of the hardest things is almost this more like Bayesian thinking of hey, here's our priors, here's how the research changed those priors, versus treating the research as these 12 people we talked to said this, and so now that is the ground truth and it's like, well, no, we're a pretty big company, we work with a lot more than 12 people, like there could be bias in that, and so it's I really like that that you guys are doing that. How is it? How are you changing the upfront part of the research process, or discovery? Like how are you helping teams write better briefs? I guess going in discovery to be able to think about that and work with stakeholders?
Speaker 3:The biggest thing is this process we're doing now. We have a project brief that we do. Sorry, there's. We have a lot of different processes for intaking a project, but often it starts with just conversations and a lot, of, a lot of stuff doesn't make it into the document that we're going to be using, that everybody's going to share. And it's been my favorite part of adding AI uh, cause I'll gather all the emails, all the texts, everything about it in there.
Speaker 3:Uh, the nice part about it is it creates a good template of the project brief. It in there. The nice part about it is it creates a good template of the project brief. But the second thing up is what are my designers going to ask? And it can fill in with. Like it doesn't have this, it doesn't know what strategy, like it'll fill in the gap so we can start to get those answers. It's really gone from something that we'd kind of craft that over time to you can use the conversations that are going on throughout the day and throughout the week about the topic to pull it in and get a straight answer for you or a straight, structured project pretty fast.
Speaker 2:That's cool. We're using it. Similarly, especially, we'll have, you know my team we'll have a lot of conversations, especially with a new client right, and new clients. They send us a lot of information. And so figuring out, like taking all those conversations, especially with a new client right they're in, new clients, they send us a lot of information and so figuring out, like taking all those conversations like that's where a lot of the data gets lost, like there's so much context, and making sure you know securely that we can capture all that information and figure that out and how to play it back to the team of it's like hey, yes, here is the the project brief for us, but here's all this additional context that that people have shared be able to do that is so key. Um, the I guess you know we talked about content versus research versus design. Sorry, um, does does it start to open up? Who does research? Like? Does ai make it so more people can participate and lead that work?
Speaker 3:not in the bank bank. There's pretty they're pretty rigid about. It's not that anybody can't do research. The way you access a customer is really really rigid. So we keep it to the researchers who are trained how to do it, so they're really particular about that. But that's when we're talking about field study. So if it's outside of field study, especially around competitive analysis, it's one of the biggest spots where we have a tendency of having a free-for-all competitive analysis. Anybody can go do it. So it's actually done the reverse where, because anybody would go do it, nobody was centralizing it. Now we're having an opportunity to say well, how many times have we gone? Look at the Apple wallet. Why don't we just pull that together rather than do it for the 13th time? Why don't we just see if we already have the answer from the first 12? So trying to centralize it that's been hard as humans, but with an AI kind of helping track that stuff it's going to, it'll be a lot easier to say have we already gotten those answers?
Speaker 2:Yeah, have you guys started looking at or talked about like synthetic personas, synthetic users, like synthetic research or anything to like? Help practice build that out? You know where is that on the horizon it has been talked about, but it hasn't no action yet.
Speaker 3:Yeah, I'm looking forward to seeing it, not because I think it'll replace a lot of stuff, but I think that we do need to have, like some standard ways. People think, especially in banking, my favorite is you don't know how crazy you are until you sit in a room and watch how everybody else has learned how to do finance. So every single customer does it a little different. So it would be hard to like have to replace that with AI. But I do think a lot of the things that we're trying to solve do need to kind of narrow down to what kind of customer are we trying to solve for, and that's where I think it would really be. Help would be like this is the type of customer we're solving for and we can have the AI test for it. Yeah, harder to find that person.
Speaker 2:No, it's interesting, right, because you know I've done a fair amount of, you know, financial health type projects, right, and it is the project of all projects where everyone becomes the user, no matter what, and everyone's got their unique idiosyncrasies of how they and their partner manage their finances, how they manage it. And it's one of those things I've found right, people working at banks are usually at a certain socioeconomic level. A lot of them tend to be a little bit more analytical and so, like I've constantly have to remind people like, hey, you're not the user, your way of looking at the world is not the user. Also, the amount of money you have, it's not typical of our users and the money problems you have not typical.
Speaker 2:But it's so hard for people to like remain anchored on someone outside of themselves and it's a classic like people end up building for themselves because everyone has a bank account, right, everyone, everyone has money and it's it's. It's it's a little different, like when we're working with like a, you know, industrial air compressor manufacturer, like a utility company, because, like, not everyone goes and, you know, has an experience with a transmission line, so you're like deeply ingraining yourself, but it's really hard with banking to maintain that, and I guess even with personas because I feel like personas, you know personas, jobs to be done, archetypes, they've kind of all gone through their phases of like usage, and so are you guys even using personas anymore? Are you more using like segments, like how has that even evolved for how your teams use those?
Speaker 3:If the products that existed for a while a lot of them it's better to use the segments because you have the data there. We know how people are functioning or how many accounts they have, how often they're pulling money out, whatnot. So segmentation has been a better approach to that. But with the new projects we're starting, we went through the process of personas to find design targets. So if we're going to make a new product, we don't want to go out there and say we're going to solve all problems, but we don't want to go out there and meet nobody's needs with the first generation.
Speaker 2:So mass market, mass AF a little too broad, yeah, a little too broad.
Speaker 3:So figuring out yeah, is there a security play with this? That might be a design target, a persona that we want, or is it? They're slow to adopt but they will really embrace it or they only will do it with new features, like knowing which path we're going down, has been helpful. So looking at the personas for which ones the right design target has been really, really interesting. Yeah.
Speaker 2:I did a similar podcast with someone from the PE space and we talked about, you know, using AI to contradict yourself and test your own thoughts and I think that's always been one of the hard things with personas right Back in the day. You know, people build personas and the posters are all like hanging up on the wall and they've got names and how many dogs they have and like all this information. But I would always find, you know the number of personas an organization created compared to the number of personas people remembered. It would kind of pick their one or two favorites and probably the one closest to them or someone they know, and they kind of become that one, become that one.
Speaker 2:I think that's something AI can open up is like pressure, testing your ideas and against different personas, different archetypes, different jobs to be done, usually a little bit too hard or a little bit too expensive or like not a natural step, as long as people can kind of put it in their workflow. You know, maybe this is when the hat is sideways and maybe it's not front or back yet, but as you're transitioning, like pressure, test yourself against that and think about, especially in banking, where there's just such a diverse. I just think about the things I've learned, you know, like working on overdraft and finding out like hey, some people use overdraft as a mini loan, basically as a mini fee. Like such an insight of like how someone lives, like what that means, like, ok, what's the gap in our product design? There's so many interesting like little tidbits and learnings like that over the years that I've found working in banking that it's so easy to like lose sight of some of those users and start designing for yourself.
Speaker 2:I think that's a huge AI unlock is hey, run this through different segments, different archetypes, pretend you're X, y and Z different people and give me some feedback. Now, it doesn't replace actually testing with real people, because I think that's one of the key things is, you still need good folks that have that ability to go deep and get like really deep empathy and like really understand the why, because you know AI has a tendency to, you know, bring things to the middle. It's going to miss edge cases unless you like force it to. So I think really learning how to prompt it to test yourself and become that helper that you've talked about is going to be a key unlock for teams to really sharpen their thinking. So this is a I think it's a fun space. Are your teams getting more advanced with how they think about using it, or is it still like it's more of an admin, research assistant, of like aggregate, my notes type of thing, different levels?
Speaker 3:you're bringing up something I hadn't thought about, sorry, so I gotta go back to your question. Let's do it so right now. If you think about a persona, the best thing you can do with it is in the middle of a meeting. Say, will jane ever actually use that thing? Like really, and it's hard to pause and halfway through talking about adding this new feature that everybody's excited about trying to solve, really pause and be like are we actually solving the problem of this persona? If the persona is going to do its job, they need to be in the room, and I wanted to extend your idea of the AI. Could the AI be the persona in the room, rather than having to have somebody have an epiphany moment, just like end of meeting the personas answering that question for you, because it could physically be there in the meeting with you somehow. So it'd be an interesting.
Speaker 2:I mean, I, I think you know I'll try this today and see how it goes but I think it's pretty easy to be able to just simulate a focus group and and that's where, um, I've always really liked archetypes, because they're there's, they're standard personality types because, you know, I always like to draw the distinction between you know, segment, which is more of a quantitative view, and then you have your traditional marketing persona, which is a little bit more about the Val prop. If you think back to the history of personas, it was really looking at like what are the different reasons someone's going to buy this product? Because you know, so much of our focus on the customer in the past was around that acquisition event. But now you kind of have this like UX persona that I like to think about, where you have very different ways. People engage with technology and think about technology and their actual preferences for how they want to accomplish things, and some people spend more time exploring, some people are more functional, some people want to kind of streamline clicking through, and so you can balance multiple of these factors and give that to a GPT and say like hey, imagine these five scenarios, take this design, run a fake focus group like, hey, imagine these five scenarios, take this design, run a fake focus group, like what might people say?
Speaker 2:And, going back to your point, it's another one of those hat forward moments of what haven't I thought about. Like, challenge my thinking and bring it in. And that's one of the things I love about AI is it's no longer about the answers, it's about the questions. And like, who asks the best questions and sharpens their thinking the most? Um, and I guess, going all the way back to your point around the designers that love the craft versus designers that love to solve solve the problem. You know it really leans into that designers who love to solve the problem aspect of it. So, um, maybe we'll, maybe we'll do a test of this one. Um, as we go through I like what, um, you know, a year from now, what do you think has changed most in design organizations when it comes to ai and and and how they're working and evolving?
Speaker 3:well, I have a, I have a hope and I have a potential. So okay, especially in where we are uh, in finance industry, that might. The potential is that it'll be just like more of a day-to-day assistant. It'd be great if it started to say here's the first draft of it and make it better. That would be great. Where we are. My hope is, especially in large organizations, that we figure out how to use it for experimentation Right now in at least the finance industry.
Speaker 3:To get to experiment, it has to be completely coded and in pilot and often our pilots are 20 people but it has to be completely coded and approved and ready to go, and I don't know how realistic it is to say could we replace that with a that's 20 person pilot? Couldn't it have been a vibe coded experience? And we're trying it with 12 different flavors of the same thing, and that's a leap because we don't know how to do it well right now. So there'd be a leap, but it would be a also. It's a great place for us to solve, like it's a problem that we could solve with vibe coding. Versus trying to figure out how to get better pilots, we could actually create vibe coded solutions.
Speaker 3:Yeah.
Speaker 2:I mean, I think I hope to that it democratizes analytics a little bit more. It democratizes analytics a little bit more. And you know, I think there's more data on how people are using our products and heat maps than people have time to make sense of necessarily or like really truly query and go in and understand. So I'm hoping it gives us the ability to kind of really understand where are people getting stuck, where are people spending time, like you know? How does that vary? What could it improve in our products? So I think that's a gap. And Does that vary? What could it improve in our products? So I think that's a gap.
Speaker 2:And I think, too, in financial services we talked about everyone has a bank account, right I think there's still so much white space when I think about, you know, commercial banking, wealth management technology you know you've worked obviously that there's still so much like UX, white space and experiences that have the ability to be disrupted. And so in commercial banking, treasury, those have been areas in the past that haven't had the design team scope that retail banking has had. I think about banker tooling, like banker teller experiences and like call center experiences. So I do think that there's a lot of areas that have been under designed in the past, that could be really optimized, that are going to have a really positive impact on customer experience, even if you're helping someone behind the scenes a little bit. So I'm hoping that, like some of those areas, get a little injection of design and design thinking and experimentation that they haven't. So any other closing thoughts on AI, design, tooling and what's going to you know, a year from now.
Speaker 3:Well, the biggest thinking I have is like, as designers move forward, like what are they going to do? It's like I think we're in a time of experimentation and I like to say fail forwards, like I wouldn't be shy about trying stuff out.
Speaker 2:Yeah.
Speaker 3:I think it's an opportunity to do more strategic work or to be a better cross person. Whichever path you go, but do it with the AI rather than avoiding it.
Speaker 2:Yeah that makes sense, awesome. Well, thank you for coming on. I think I have some new stuff to go try, so hopefully people listening at home have some new stuff to go try and can start failing forward. So, jason, I have a lot to think about too. Yeah.
Speaker 1:Thank you for joining us on Build what's Next Digital Product Perspectives. If you would like to know more about how Method can partner with you and your organization, you can find more information at methodcom. Also, don't forget to follow us on social and be sure to check out our monthly tech talks. You can find those on our website. And finally, make sure to subscribe to the podcast so you don't miss out on any future episodes. We'll see you next time.