AI Is Showing UI Designers the Door

Paul Boag: Hello and welcome to the Boag World show, the longest running web design podcast where we look at user experience, design, conversion, optimization and working in the web. On this month's show, we're going to discuss AI replacing UI designers, the crossover between CRO and ux, and how AI toolkits might become absolutely essential for your career. My name is Paul Boeger. Joining me, as always, is Marcus. Marcus Lillington. Hello, sir.

Marcus Lillington: Uh, the longest podcast in the world.

Paul Boag: Just like. Yeah, that's not really a selling point, is it? Yeah, yes. We just didn't know when to stop.

Marcus Lillington: We're on episode 10,425.

Paul Boag: I know. Yeah, yeah. It's no guarantee of quality whatsoever. It's just obviously not. No, no.

Marcus Lillington: We do ram the point to home quite a lot because we repeat ourselves so much. So, you know.

Paul Boag: Yeah, that's our excuse. When you've been doing it since. What was it, 2005? We started something like that.

Marcus Lillington: Yeah. Wasn't it?

Paul Boag: I think so, yeah. It was when. When you used to have to crank a handle on the side of your computer. Yeah.

Marcus Lillington: It made a funny noise, that's for sure. Anyway, Paul, I haven't seen you since you went away, uh, on a boat and got really sick.

Paul Boag: Yes.

Marcus Lillington: Like everyone does who goes away on a boat. I have never cliched of you, Paul.

Paul Boag: I didn't get. I didn't get sick. Well, I got sick afterwards.

Marcus Lillington: Oh, okay.

Paul Boag: So. So I, uh, mean, obviously if you go on a cruise ship, you're going into a petri dish, essentially.

Marcus Lillington: Indeed.

Paul Boag: Disease. Um, especially because I went on a cruise ship. Hate to say this, for over 50s, adults only, but senior adults as well. So everybody has got weakened immune systems.

Marcus Lillington: Well, you've just had a birthday as well, haven't you? Nearly 60m.

Paul Boag: Yeah, I'm pushing it. No, I'm 50.

Marcus Lillington: I can't get away with that one because I'm nearly 60.

Paul Boag: Exactly. Yeah. Yeah. Um, this is why I hang out with people older than me, because it makes me feel younger. So, yeah, I. I got flu afterwards when I came back, but it was my wife. Right. We've been cruising several times before, never had a problem. This time, sick as a dog. So bad. Such bad. Seasickness to the point where they were injecting her every eight hours and that was still not doing the job, really. So it was an absolute disaster.

Marcus Lillington: Oh, so it was seasickness rather than.

Paul Boag: Yeah.

Marcus Lillington: Poisoned by the food or somebody else.

Paul Boag: No, everything now it's really interesting. So we went away with Saga and I cannot, uh. Shut up. Marcus, don't grin at me like that. I could see your face. For those people who don't know sar, because it's a UK thing, saga is, is basically a, um, thing aimed at old people. Yeah. So 50 plus.

Marcus Lillington: And, um, I don't consider anyone 50 years old old anymore.

Paul Boag: No, I don't either say that.

Marcus Lillington: Yeah, they're youngsters.

Paul Boag: There were some people, there was one woman I was chatting with. We were, um, we were going up to the top of a mountain on one of these cable car um, things and um, she, she giggled like a little school girl as we, as we set off and I said to her, what was that about then? And she said, I haven't been on one of these when I, um, since I've stopped skiing. And I said, oh, you had to stop skiing, did you? Yeah, I was 81 and I hadn't had an accident yet, but I thought, it's only a matter of time. I probably should stop skiing. Bloody hell, these people were amazing. I met all kinds of people that just had incredible lives. I came away going, bloody hell, what have I done with my life? But anyway, that's beside the point. So, yeah, it was a disaster. But very interesting from a user experience point of view. Right.

Marcus Lillington: Is it really?

Paul Boag: It is. It is, it is. I swear. Right. This, this relates to.

Marcus Lillington: I'm sure Kath thinks the same thing.

Paul Boag: No, she, she was a disaster for her. But they have what really impressed me. It was almost a bit like Wiltshire Farm Foods, because do you remember when we used to work Farm Foods, for those of that you don't know, Wiltshire Farm Foods sold frozen ready meals to old people and they've been very, very, very clever in competing with the likes of Tesco's and that kind of thing where you can obviously buy frozen ready meals. So it's like, why. And all your other, you know, groceries. So why would you buy from Wiltshire Farm Foods? Well, it's because they really understood their audience. They understood the unique needs of their audience and they accommodated that. So they had things like, um, uh, people obviously worried about, um, a stranger coming to their house. And so all of the drivers were police checked and you got the same driver every time. They, they couldn't pick up the big trays of stuff to put it in their house. So the driver would come into the house because they trusted them and would put the stuff directly in the freezer for them. They could reorder on the door without having to faff around with the Internet and all of this kind of stuff. Really well thought through and saga was exactly the same. So it, like, it had things like, um, just outside, you know, where you got off the ship. It would tell you how far you had to walk in order to get to the, the, um, uh, you know, the bus or whatever that would take you wherever they, you didn't.

Marcus Lillington: So you can have a wee, Have a wee before you go. Yeah. Long walk.

Paul Boag: Well, no, because some of them, some of them were very dodgy on their feet. Right. Yeah. Okay. Although I like the wee joke. Yeah, that's very true, even for me at this point. Um, yeah, but they also did things like, obviously these people couldn't carry their own luggage. So a, uh, driver comes to your house, right? All you need to do is get the luggage to the front door. The driver then puts it in your car. They drive you to the destination, right, of the, the cruise ship to pick you up. You don't touch the luggage when you get out. Some, a porter unloads it and takes it through and it is all just done for you. They also cover every dietary requirement imaginable to man. It was really impressive from that point of view. Terrible from a seasickness point of view

Marcus Lillington: for my wife because it's different seas, different smaller boat, bigger boat.

Paul Boag: I really, I really don't know what it was. It was just. She was really bad for the entire time. Really weird. Anyway, it doesn't matter, does it? Yeah, such is life doesn't matter. No, well, I, I, my attitude is we have had so many amazing travel experiences over the years.

Marcus Lillington: They're not all going to be perfect.

Paul Boag: No, they're not. Things are going to go wrong. And we still got to dog sled and we still got to see the northern lights, and we still got to see natural beauty. Snow on and. Yeah, it's great kind of snow on.

Marcus Lillington: Shit's good. That's always good.

Paul Boag: Snow on. Yeah. Well, not literally, but, you know. Anyway, should we actually talk about something useful?

Marcus Lillington: Um, do we have to?

Paul Boag: Yeah, let's talk about it. Let's talk about the app of the week. Um, there's been a lot of buzz around the week. App of the Week. Yeah. It's not. Except we do it monthly, so I don't know why I call it App of the Week.

Marcus Lillington: Of the Month.

Paul Boag: Yeah. God, this is such a tired format. M at this point. Um, so, yeah, there's been a lot of noise around. Um, Google Stitch. Um, so Google Stitch, if you don't know, is a tool that was released by Google, um, a little while back, uh, and they recently updated it to be V2. And um, it's an AI tool basically. Um, and so what it will do is it will create user interfaces for you that um, you just give it a command, you say what you want, it builds the interface, job done and then you can even go in and manually edit text or change images and basic stuff like that. It's not like as fully powerful as figma is in regards of manual editing, but it's got some basic features and then obviously you can iterate the design that the um, AI is produced Now I still think that figure dot design which I recommended in the last show is better. Um, it doesn't have the manual editing but it's in my opinion produces much better designs. Um, and is just. It is more considered I think than, than Google Stitches. But that's not really what I want to talk about. The reason that I brought up Google Stitch is because it, it's now brought this. Oh well, UI designers are dead. Right. And.

Marcus Lillington: Well, yes they are.

Paul Boag: I'm beginning to. Everything in me wants to go, no, of course not. You know, um, when you say UI

Marcus Lillington: designers, what do you mean, Paul?

Paul Boag: I mean the people like, um. I won't start naming names because that's a bit on the nose.

Marcus Lillington: Um.

Paul Boag: Yeah, you know Dave, he's fucked. Um, no, I'm talking about people that are just doing UI design, um, not looking at the broader user journey, not um, uh, doing user uh, research. I'm talking about the people that sit down, open up figma design a webpage. Right. Um.

Marcus Lillington: Well, I am going to name someone. Go on. Our friend Andy. Andy, a UI designer?

Paul Boag: Um, yes, I think he is, yeah. Okay, so that's really unfair of poor old Andy.

Marcus Lillington: No, but he's the reason why I've said him because he's a bit of a special designer. He brings something to the party and so that.

Paul Boag: Yeah, that's where it gets interesting. Right. So my heart wants to say that we still need UI designers. Um, and certainly in, in many cases I think the answer is yes to that. So, um, Andy is a really good. Yeah, like you say, it's a really good, uh, case in point. He's been doing this for years. He's got a very strong style. Um, if you want that style, AI isn't going to be able to replicate that. Um, you know, he brings a creativity. He's also working on brand identity, all of those kinds of things.

Marcus Lillington: What I was kind of driving at is if somebody has set your, your online brand so not just your brand, not. They haven't done A brand style guide with a logo. They've done more than that. They've, uh. You might have a design system.

Paul Boag: Yeah.

Marcus Lillington: Then, but then maybe that's. Of course you can create a UI out of a design system. I guess I'm trying to say, where, where's the line?

Paul Boag: Yeah. And uh, this is, this is. I've just found myself twice recently going to clients. I think we can skip getting a UI designer. And that really disturbed me. Um, and I didn't like saying it. But as a consultant, is my job to give people the best advice possible for their business rather than what I ideologically believe. And in both the situations it was like, there's limited budget, there's limited time, um, they've already got a well defined brand. Um, and that actually all they needed is me just to pump the thing through AI, look at the results, iterate it a little bit. Right now you could argue, well, I used to be a UI designer.

Marcus Lillington: So you know what you're talking about.

Paul Boag: So I know what I'm talking about.

Marcus Lillington: Could they have done it?

Paul Boag: They couldn't have done it, but you

Marcus Lillington: could m. Yeah, I kind of know. You know enough educated in UI design. Yes.

Paul Boag: Yeah. So you're not actually a UI designer, but I think you, you could do it. So it's about. It's turned now from, from really, it's experience. And I want to use the word taste, but I hate the word taste because it's so wishy washy and vague and pretentious.

Marcus Lillington: But it doesn't think it is.

Paul Boag: Paul.

Marcus Lillington: I disagree. I think taste is perfectly fine and reasonable.

Paul Boag: Okay.

Marcus Lillington: And knowing. Having taste in design. What am I trying to say here? I think is, is the difference. That's what we're talking about. It's not just it is experience of what works on, in a, in an interface, but it's also knowing whether something's good or not is a taste thing, surely.

Paul Boag: Yeah. I think my problem with the word taste is that a lot of people think they've got tastes that don't. Okay.

Marcus Lillington: I can't argue with that.

Paul Boag: Uh, but yeah, fundamentally it's, it's having that experience. Experience having that, that sense that of, of whether this is working or not is all you really need now in order to produce an interface. And I think that is a really. I feel really uncomfortable saying that and I feel, I know that I'm going to be making a lot of people that uh, are listening to this feel uncomfortable as well. And probably some people angry at me for saying it. And this by the way is just my opinion. Right. But we, we. I feel like we're now at the point where UI design or we're getting towards the point where UI design could become like vinyl. Right. In the sense that the vast majority of people don't need it. Right. Don't care about it. There will be some people that passionately still care about it and think it's incredibly important and want it. But the I, I think as a field it could reduce dramatically is my concern. Which is a bit depressing really, isn't it?

Marcus Lillington: Well, yeah, I mean sort of. I'm currently kind of railing against the AI music producing apps like Suno.

Paul Boag: Yeah.

Marcus Lillington: Um, and I kind of get it. If you have no musical skill but you're good at prompts, you can make music.

Paul Boag: Yeah.

Marcus Lillington: And I kind of get that. That's kind of quite, that must be fun.

Paul Boag: And, and also I think more than that, in fact I uh, think that is an incredibly good analogy for what we're seeing with UI design. Right. Is that if you're just throwing together a, ah, YouTube ad. Right. You might as well, you know, nobody actually listens to that background music. It's not particularly important. Is good enough. Right. To use some AI generated stuff is what will go through most people's minds. You'll probably rail against that and understandably so because that is the, you know, that's your area of expertise. But that doesn't mean that the need for real musicians is going to go away, but it does mean that the, the amount of opportunities for real musicians will be dramatically reduced. Right. And that is, yes, that's my concern I guess in this situation. And I'm um, I'm hoping it won't go that far because I think that the differentiation again is with the music analogy is you still need somebody with experience and taste to know whether what comes out the other end is any good or not. Right.

Marcus Lillington: Potentially, yes. I mean I think with music it's. The music is closer to art than maybe ui. UI design is. When we get into discussions about art then that's what I think makes humans special different. Um, and we shouldn't be doing stuff to make that less likely to happen. Yeah. Um, and where I was going with that thought was is taking UI design away from humans adding to that whole sort of. I, ah, know it's not. UI design isn't really art. Maybe it is, maybe it can be, but because it's something that involves a kind of artistic bent, do we need to be worrying about taking that away

Paul Boag: from I do things?

Marcus Lillington: Yeah.

Paul Boag: I mean, I, I would argue, going back to music, that not all music is art. Right?

Marcus Lillington: Of course not.

Paul Boag: You know.

Marcus Lillington: Well, well, that's, That's a hard one.

Paul Boag: Yeah. Well, all right, all right. In my, in my mind, art is something that is primarily created as, um, expression.

Marcus Lillington: Right, yeah.

Paul Boag: For itself. For itself. And then as soon as it's being done to a brief or something like that, in my world, that's the difference in art and design. Right. Design is done to a brief for a purpose. Okay, yes. Um, so that YouTube video I mentioned with the background music, I wouldn't call that art. I'd call that. Yeah, something else. I don't know what.

Marcus Lillington: I mean, it kind of could be if you took it away from the. And it became a thing on its own. But. Yeah, but, yeah, I agree with you.

Paul Boag: Context is everything. With art as well, that's the other thing, isn't it? So, so, but that's, that's arguing semantics. I do worry about this, but I think it's inevitable and, and perhaps I'm defeatist. I don't know. But I do think it is the way that it's going to go for right or wrong. Um, and, and you, uh, know, I only. I have to. As a consultant working with clients, I have to say what's best for my clients in their specific circumstances. But that doesn't mean I'm not aware of the broader impact of this. But I'm being paid to give good advice to my clients. So it puts me in a.

Marcus Lillington: We all use AI to maybe cut corners isn't the right thing, but to, to help us be more productive.

Paul Boag: Yeah.

Marcus Lillington: So, yeah, why did we do that?

Paul Boag: So if you're a UI designer, where does that leave you? I guess. And I think it leaves you in having to do. I think there are two areas where, um, you can still add a lot of value. I think one is, um, to look beyond just the pure interface that you're working on, but look at the larger business problems, look at the user needs, you know, to become more of a UX designer and, um, almost a digital kind, um, of consultant as part of the UI design job. So you're adding value on top of, you know, what the AI can do for you. I think that's one area and then the other area, um, is the kind of gap between interface screens. So at the moment AI, and I think this will change, but at the moment AI thinks very much. I'm working on this one screen that's front of me now. I'm working on this screen that's front of me, and it's not really considering the journey or the user flow. So you're still absolutely providing value there now, but that matters less in some contexts than others. So, for example, if you're working on an app, that is absolutely, absolutely essential. If you're working on a website. Well, now you're talking more about information architecture work, which maybe is a different discipline and maybe a discipline that you need to roll in to what you're doing if you're not already. Um, because that is something the AI certainly helps with, but doesn't take over in the same kind of way. It's a weird time. Let's move on from that. Talk about more AI shit. I guess I find it. I'm finding it very hard to talk about anything else at the moment because it is absolutely dominating everything. Um, so this is an interesting one that I wanted to talk about, which is the idea of AI or you having an AI toolkit and that we're all going to eventually need, really, whatever our role, whether we're designer, a developer, a, uh, project manager, a salesperson, a product owner, uh, whatever your job is. I think increasingly one of the most valuable assets that we're going to end up having is an AI toolkit. And that's what I want to talk about today.

Marcus Lillington: And what do you mean, Paul?

Paul Boag: Yeah, so there are platforms, uh, most of these AI platforms, things like Notion, things like Claude, are beginning to push users towards developing what they call AI skills. Okay. An, um, AI skill is a reusable set of instructions that you write once and then use repeatedly. So we're not talking about having a prompt library where you kind of copy and paste your prompt across the whole time and, you know, uh, and it's got a bit of information in. These are like documented processes, okay? The kind of like an sop, a standard operating procedure for how to do a specific thing. The kind of thing that you would hand across to a junior colleague and they would just follow it. Step one, step two, step three, step four. Right. And that distinction between a prompt library and a skill, uh, really matters. Right? Think of a prompt library as like notes pinned to your desk, okay, that you could just kind of pull off and reference, um, as you need them. A, uh, skills library is a set of repeatable processes that produce reliable, consistent outcomes. Right? So AI skills have become particularly popular because of context management. Right? So what do I mean by context management? It is perfectly possible to overwhelm an AI with too much information and you begin to see shit results. So what we're moving towards is a world where we can say, okay, I want you to achieve this goal, right? Look at these specific skills that will help you achieve that. So instead of just dumping everything on the AI all at once, you're saying, oh, maybe you need to look at this skill first, then go and look at this one and do these steps so that it's kind of better management. You're keeping the skills, keep the instructions focused, instead of overwhelming the AI with everything it needs to know at once. So why do I say that we all need these? Um, I'm. I've built up a library of about 60 skills at the moment, and it's growing literally every day. Um, and I've come to see them as possibly outside of my Boag World website, the single most important asset that I own as a business. Right. And I think that we will get to a point where companies will have their own library of AI skills, right. For delivering consistency. So, for example, Headscape would have a library of AI skills. One that would make sense for you is proposals, A skill for producing proposals. Okay. That always covered the same points, that made sure that the same thinking went into it, the right questions were asked, and all the rest of it. So instead of you being the only one that can write proposals, because you've documented all that Emma could write a proposal. Right. Dan could write a proposal. People who aren't experts. Now, that's not to say you wouldn't need to still check them. Right. Um, but essentially, that is Headscape's way of writing proposals. Okay. And you could.

Marcus Lillington: To do that in the past.

Paul Boag: Right? Okay.

Marcus Lillington: Interesting. Not particularly. Not with AI in mind. I mean, Long time.

Paul Boag: Yeah, yeah.

Marcus Lillington: It's been. It's been on my list of things to do for probably 15 years. Yeah. And it's hard.

Paul Boag: It is hard. It is hard that. You know, And. But what I now gives you, um, which we'll come on to later, but hey, let's jump around, is that you can write out your. Your process, right? You're not entirely happy, but let's see what AI does with it, right? Run the process. Oh, no, that wasn't quite right. I'll fiddle with it, Gabe. Run it again. So you could quickly iterate upon it, while before you'd. You know, you'd have to get it perfect before you handed it off to Emma. Do you know what I mean? Yeah. Um, and so it's a much more manageable process that you can quickly iterate on. So companies will end up with libraries of AI skills. So Skills like how to run a, uh, top task analysis, how to do a survey. So I'm m naming the ones that I've got now. Yeah, how to create a Persona. All of these kinds of library of skills that could be used across the organization. So companies are probably going to have fairly generic ones in most cases. So an agency like Headscape would have more specialist ones. But if you're a bigger company, it might be um, uh, how to claim expenses, how to book holiday, things like that so that the AI can go off and do those things for you. Then you've got the UX teams within the organization now they'll develop AI skills that can be used by their colleagues, right, to do some basic UX work themselves. So the ones I just mentioned, how to create a Persona, how to do um, journey mapping, how to do um, surveys, etc. Etc. Right. So what that means is you can now start um, democratizing ux. So instead of the UX team being this bottleneck for everything, a developer could quite easily go, um, for example, going back to our earlier point, um, you use the skill. I need you to write me, um, a brief to brief an AI on how to create a user interface, right? And it would ask you a load of questions and then it would write out that brief. You could copy that brief across into Google Stitch or into figure design and bang, there's the interface, right? So product owners could now prototype their own sites because they've done their own user research, because they've got used all the best practice about how to create a survey, uh, built into the skill. They know how to do it, they're not going to screw up. Content writers use the web content writing skill so that suddenly everything's broken into chunks with scannability in mind, with accessibility in mind, with a reading level of seventh grade and below. And all of that kind of stuff is just built into them. And then so there's these three levels, there's the company level, the UX team level. Now in your case in Headscape you just have the two, you'd have the company level and then the final one is individual levels. And this is where I think it's got a real career value, right? So these are skills that are unique to your specific role and your specific way of doing things, right? So there might be skills that only you do it, you know, jobs you only you do Marcus, and you never want anyone else to do them. So you would build out a library of skills that, that do those specific jobs. But more than that, you can build out a library of skills that do a, uh, specific job in your way. Right. So there will be libraries online of skills. In fact, there already are. You can go and find a skill that will teach you to do pretty much anything, will teach the AI to do pretty much anything, but they won't necessarily do it your way. And the example I use is Top task analysis. Jerry McGovern, um, created top Task Analysis. I've completely bastardized his, his Top Task Analysis. Um, he would be horrified if he saw the way that I did it because it's completely different from the way that he did it. And I probably shouldn't even call it the same thing, but I have my way of doing it. So I would create a skill that was my way of doing it. So can you see how you can then start building up this kind of library of skills that you, um, can prompt AI to start using and to start going away and doing different tasks already? But what's really interesting about it is not, not just the value in that, but the value in the process of actually creating the skill in the first place. You said how you found it really hard creating that proposal one. Yeah, right. Because it forced you to articulate what you do subconsciously and naturally. And that's quite a challenging process to go through. Um, but the act of creating an AI skill kind of helps you think through the process and it surfaces hidden assumptions that we make. All that intuition of, well, what good looks like that you've never actually articulated before. So I've ended up with like, this library at the moment of about 60 skills that cover everything from how I write, you know, when I write a report, a business report, through to how I brief a designer, uh, through to how I do top task analysis or write unbiased, um, survey questions. I know how to do all of those things. But the advantage of me creating a library is that it means nothing gets missed. Right. Um, you know, because you end up on projects, juggling so many different things that things get missed, things get dropped out of it. So everything, um, nothing ever gets missed. And also I get the benefits that AI does a lot of the work for me now. Right. Um, so, I mean, I've always been a big, a big, ah, proponent of playbooks. Haven't I've talked about playbooks before on the, on the podcast. And this is the next evolution of that. So I've come to believe this is my biggest asset, really.

Marcus Lillington: Certainly is. If you're. Yeah. I mean, with you, Paul, I mean, it's your biggest asset from the from the. It's kind of. Because it kind of reminds you of all of the things that you offer and as you say, you won't miss anything. Uh, when you're going through various processes and you're busy on this and you're busy on that, um, the pro. What would be very valuable for me, particularly because I'm, as we've already discussed at the start of this show, getting, getting a wee bit older and won't be around forever. Passing on.

Paul Boag: It makes it sound like you're gonna die. You mean retire?

Marcus Lillington: Yes, I might make it to next summer. Um, so anyway. But the idea of documenting what I know is something that's been front of mind for, well, certainly the last two years and not really done a lot of it. So.

Paul Boag: Yeah, yeah. And that's the great thing as well about these AI skills. They're basically just markdown documents, right? So that means that they're entirely readable by the AI, but they're readable by people as well. Um, so they're equally beneficial in both scenarios. So, yeah, that's a great example you almost need to get. Because Chris is down to a few days a week, isn't he, at this point? Yeah, he needs to be doing this, you know, producing these documents. Yeah.

Marcus Lillington: Um, and it's not even necessarily for AI, it's for. It's future headscape people. Yeah, yeah. So, yeah, but, but doing it in a format that could help those people use AI to do the things. Makes a lot of sense.

Paul Boag: Yeah. Because obviously the, the big thing is that, you know, let's take say for example, writing a business report. All right? So I've done a load of a great example recently. I was, um, uh, asked to review a website. Was basically just a UX audit, right. So I went through the website in the most haphazard manner I'd ever done it, right. I literally went through the website and recorded audio, you know, text to speech or speech to text. Should I say press the button and said the head is shit because of this, this, this and this. And they should do this instead and then go to the next, oh, this. And I just a brain dump of shit. And then I just ran my AI skill against that and I had a finished report. I had to read through it and check it was all right, two hours. I mean, it used to take me three days to write up a report like that, you know, so the time savings are immense once you've got this library. So how do you build it? Let's, let's talk about that briefly. So this isn't just about downloading skills from the Internet. I strongly believe that these skills should be unique to you and how you work and represent how you think. Otherwise you're not really adding any value, right. You know, and everybody will just be end up doing things the same way. Your competitive advantage is to create these things that are unique to you with those little things that you've learned over the years. Now, I'm not saying you can't start. If you're early in your career and you don't really know very much, then by all means, start with a library, um, you know, of things that you've downloaded. But the key there is constant iteration, right? You produce these things once and it produces something and you go, that's not quite right. So let's go back to that report example. I, I produced the report, I read through it. Generally speaking, I was happy, but I made a few changes, right? And then right at the end, I said to the AI, can you look at the changes that I've added, work out why those changes, you know, why I've done those changes, or ask me, right? And then update the skill to represent that. So next time you do a better job at it. Right. Um, and I just do that all the time, basically. Neither, I think, should you just create these things out of your head because you're going to miss assumptions that aren't obvious, that are obvious to you, but aren't going to be obvious to another person or an AI So what do I do to create these things? Well, first of all, I bring together everything I have related to that skill. So if I've got old Personas for, let's say I'm m creating a Persona 1, I get loads of old Personas and feed that into the AI and then I, um, write out a rough first, uh, draft of what I think that would, you know, that, um, skill should be, right. Not worrying too much. I give it the examples as well. And then I ask AI to review everything and ask me questions, right? What have I missed? You know, what assumptions am I making? You know, all of that kind of stuff. Challenge me over doing it, um, and then get AI to redraft it based on, you know, on the answers to those questions. Um, I then review it. I might iterate it a little bit. Most of the time, to be honest with you, I just accept whatever it's done, give it a go on a real project, then go, oh, uh, that wasn't quite right. And, you know, let it know why it wasn't quite Right. And iterate like that. So when I say I've got 60 skills, I probably don't have that many that are battle tested at the moment because I'm, I'm just doing this, but, you know, I might have 10 that I can reliably, you know, know for certain work. And the rest is, you know, a bit. Yeah, we'll see how they work out. Um, so when you create your skills, just bear three things in mind. You need to have a clear description of what it does. Right. And I'll tell you why in a minute. Clearly explain what the output should look like and clearly identify when it should be triggered. Right. So why do I say that? Well, this is where we get into the even more exciting bit that we are now moving into a world of AI agents. Okay. And an AI agent is capable of working independently on more complex projects without your input. Right. And, um, what it can start doing, you could set it a goal. Like, I want you to create a Persona, set of Personas. Right. And it will. If you've clearly described what your skill is, if you've clearly described when it should be triggered, then the AI is capable of going, okay, so I need to create a Persona. First of all, I need to do, um, some online research to understand the audience, right? So I'm going to trigger the online research skill. I run that. Okay, Now I've done that. Next thing I need to do is decide on my segments. So I'm going to run my segmentation skill. Okay? So now I've got my segments. Now I can run the Persona skill on each of those segments to create the final Personas. So suddenly you're in a world now where it can chain multiple things together to get from zero to a final deliverable. A great example for you, Marcus, is, um, I record all of my meetings, right? And so if I do a sales meeting now, um, when I, you know, a first meeting, a kind of kickoff meeting with the client, and after that it was decided that we're going to create a proposal. So when I get off of that meeting, there's a complete transcript there. I can say to AI, add this as a lead to my, um, database and write me a proposal. And it will literally, it's got the skill to go and know how to add the sales lead to my database. Um, it even adds a task, because part of the skill is add a task to chase them after a week, after I've sent the proposal, and then it writes the complete proposal for me. And all I need to do is read through it and that's one of the 10 that I've really dialed in. Um, and even as part of that it's, it's runs another skill before it writes the proposal which is to go and research the company right online and find out a load of background information. So it's chained all of those things together and I end up with a finished proposal I can just send to the client. In fact I've even been playing now with this is getting increasingly lazy but um, with it writing the email that it then sends to the client. Right. Um, yeah.

Marcus Lillington: You know what you were saying earlier about UI designers, Paul?

Paul Boag: Yeah.

Marcus Lillington: Um, you're not going to be required quite soon by the sound of it. Yeah. Just so it'll just be AI doing Paul Boeg's yeah. Work.

Paul Boag: But that's kind of. But that's kind of where. That's why I say it is one of my biggest assets because I could get into a position where I could you know, um, effectively set off AI agents doing a huge amount of my work. And if I'm still charging for the same value as I'm m charging now, I'm going from um, you know, I don't work a huge amount as it is. I do like four or five hour days. I could be doing two hour days, you know, uh, and still have a self sustaining business. And my value comes from the intellectual property in those skills that those skills are set up for the way I think and the way that I work and the way I behave. Right. So it's replicating my expertise.

Marcus Lillington: Here's an interesting point on, on that front, will AIs when they train be using your personal skills?

Paul Boag: Um, that's very, very good question. They um, shouldn't. Right. Um. So according to all the terms and conditions, according to everything they promise, they are not training on, on the data you put in. You always have to. I, I'm cynical and I presume they will.

Marcus Lillington: But so uh, therefore you've only got what, a couple of years, Paul and then anyone can be Paul Berg.

Paul Boag: Yeah, but that's always been the case. Right. You know, because especially with, with the way that I operate. I've always been very open doing things like this podcast. Yeah. I'm doing it right at this very minute. M. Yeah. You don't stand still in this industry. Right. Or you shouldn't stand still in this industry. In two years time I'm going to be so beyond now. Right.

Marcus Lillington: Yeah, I was being.

Paul Boag: But no you're not. You're making a good.

Marcus Lillington: No, no, no. What I'm trying to. What I'm trying to say there is. People don't necessarily hire you or certainly not only hire you for your. Because you know how to do user research. I mean, obviously it.

Paul Boag: Yeah.

Marcus Lillington: Do need to know that, but they hire you because of your confidence and they. And they want help. Yeah. And that's never going to change.

Paul Boag: And that. Yeah, absolutely. And people always want to work with people and. Yeah. There's a lot of other reasons why people hire me. You're absolutely right. Um, so, I mean, really, I see this. You can see how that has huge potential for you as an individual, but you've also see how it's got huge potential institutionally as well. Right. One of the biggest problems, um, is, you know, um, kind of knowledge drain from an organization. Okay. That someone very experienced, everybody's relying on leaves, and now the organization's stuffed. Right. But if they've, if all of that has been expressed as, uh, skills and standard operating procedures and that kind of stuff, then the damage is less. It doesn't go away, obviously, but it maintains more of that institutional knowledge. So really, I'm kind of encouraging people to start creating skills. Right. Whenever you, uh, whenever you sit down and go, you know, do I do this thing regularly? Okay. Whether it be a proposal, whether it be a brief, whether it be a, uh, you know, whatever it is. Okay, can I turn that into a skill? Now, there is, there is a barrier that will stop many people. Okay. And this is where I think there is a huge opportunity for those of you who push through this barrier. And the barrier is, oh, I'm just so busy, I haven't got time. I'll just get it out. Right, right. Um, and they don't sit down to create the skill because creating the skill is going to take time to begin with. It's going to slow you down. But then once you've got those skills in place, it exponentially speeds you up. Right. Literally. You know, when a report goes from three days down to two hours. Right. That's a huge win. When proposals. I hated writing proposals. Right. The amount of thinking and effort that has to go into a proposal. Again, that would take me. What? I mean, I don't know how long it takes you, but it used to take me like a day, uh, to write a proposal. Yeah.

Marcus Lillington: Depends what it is. But it's usually days. Not solidly, but, you know.

Paul Boag: Yeah, yeah. On and off. And yeah, again, that goes down to like an hour. So it's absolutely worth doing if you can get past that. Oh, I'm just too busy. I haven't got time to do it. Right. And you make yourself slow down to speed up later. So, Marcus, you're gonna go create skills, or are you just gonna ignore me and go, that's a very good idea, Paul, but I haven't got time. I can't be bothered.

Marcus Lillington: Um, probably neither of those.

Paul Boag: You take it away from this.

Marcus Lillington: No, it'll be a different excuse as to why I don't do it.

Paul Boag: Oh, I see. Right, Fair enough. So who knows? Screw your staff when you retire.

Marcus Lillington: Yeah, that can make do. That's the thing that's making me going okay. I mean, even if it's not even as an AI skill, documentation needs to happen, and it may as well be.

Paul Boag: Yeah, what you've been describing, there's absolutely zero additional effort in making it an AI skill. You know, it's just documenting stuff. Um, and those people that have actually created a digital playbook may well have, you know, be further down the line of this than they think they have. So it's interesting. Anyway, I wanted to wrap up with, um, just an interesting read of the month. Um, that caught my attention. Uh, it's. It's just a random post that, that really encouraged me, um, which is how CRO and UX work together to increase website conversion. And this is something that I've been struggling with for a while in my own business that, uh, I felt like there were multiple things that I did that seemed unrelated to one another. Um, and one of them was ux. One was conversion optimization and one was design leadership. Um, but I've come to realize that those things actually all work together. Right. So, you know, um, UX and conversion rate optimization, as this article lays out. There'll be a link in the show notes to the article, which will be if you can get to boag world, uh, 33.

Marcus Lillington: 3.

Paul Boag: 33, 33.

Marcus Lillington: Yeah.

Paul Boag: Um, that will get you to, you know, all of the links and things that we mentioned in today's show, but in this article, it basically really well draws the parallel between those two, you know, and how they work together. And the design leadership really is addressing the larger organizational problems that are also undermining your conversion and your effectiveness and the user experience. So actually do fit very well together. So this article basically is arguing that CRO and UX are two sides of the same coin. You know, CRO, it asks, did they convert ux, asks was it easy and enjoyable? And that we need both. I mean, I would go further and say UX also, um, encourages retention, um, as well as acquisition. So, uh, and a lot of the metrics around them are shared. The macro micro conversions funnel, drop off satisfaction scores. They all apply both over, you know, conversion optimization and ux. And uh, really they both rely on data analytics, event tracking, heat maps, session recorders. Both show you why conversion is falling off, but also shows you where people are struggling, where the highest impact fixes will be. Um, and this is really interesting because often you find in organizations that these things are managed in two separate teams. You've got conversion teams and you've got UX teams or conversion kind of goes in with acquisition and marketing and stuff like that. And I really believe they need to be managed together. Um, the article doesn't pull that out specifically, but that was the big takeaway for me is look, we need to be unifying these two things. It also drove home this idea that really to be effective with CRO and ux, you need to be thinking about the entire journey and not just one page at a time. Which goes back to what I was saying about UI designers right at the beginning. Um, and that both UX and CRO require, you know, continuous experimentation. Um, and how post conversion ux, in other words improving the experience once people have converted should be treated as a ux, uh, CRO lever too because so often CRO teams just focus on that initial conversion, not considering the drop off that then quickly happens if the experience is. So just a really interesting article that, that I thought I'd share with people that they might want to check out and I found it encouraging because it made me feel less of a, uh, kind of random jack of all trades and master of none.

Marcus Lillington: Well, nothing wrong with that, Paul.

Paul Boag: No, not especially these days. Again with AI, you know, um, the generalist is on the rise finally. I don't feel, feel ashamed to be a generalist anymore. Anyway, Marcus, you got a joke to finish us off with?

Marcus Lillington: I have. Here we go. I just purchased an original Van Gogh coffee table. I know it's original because there's a bit of veneer missing. Oh.

Paul Boag: Oh, that's a, uh, that's quite a good one. I like that one. It just took me a fraction of a second to work it out. That little bit of effort made it, made it all the more satisfying. Well done, Marcus. You get a gold star for that one.

Marcus Lillington: Yay.

Paul Boag: Right?

Marcus Lillington: Lucky me.

Paul Boag: I think that's it for this month.

Marcus Lillington: Lovely.

Paul Boag: So we will be back again in another month and Marcus is going to do something amazing on the next show. Before we started today, he committed himself to do the whole next show single handedly. That was what you said, wasn't it Marcus?

Marcus Lillington: Something like that. Ah, maybe not quite that, but you

Paul Boag: know, I'm basically backing you into a corner where you have to do something on the next show so that he's miming editing and he's not to edit this bit out. I will check. No, I won't. Will you?

Marcus Lillington: No, you won't.

Paul Boag: Alright, until next time. Thanks for listening and goodbye.

Marcus Lillington: Goodbye.

Paul Boag: Sa. Sam m.