Website Rebuilds, AI Tools, and UX in 2026

Speaker A: 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 a tool that has made me cancel my Figma subscription, how I update websites differently since AI, and how the role of UX designer is going to change in 2026. My name is Paul Bogg. Joining me as always is Marcus Lillington. Hello Marcus.

Speaker B: Hello Paul.

Speaker A: How are you? I am bloody amazing as well, you know, because I'm about to go on holiday. Well, you've just got back, haven't you?

Speaker B: Yeah, well that's when I'm, I'm like

Speaker A: m. Yeah, that's true. Did you have lots of snow? Was it fun?

Speaker B: It was excellent fun. We were near where the Olympics was, so that was kind of. Well it wasn't. I was gonna say that's double fun. We didn't really see any of the Olympics. It just kind of felt nice being just over there. But yeah, yes, we had a big dump of snow the week before we went and it was mostly nice weather when we were there. So yeah, it was fantastic. Perfect. Much fun was had by all. Uh, a few minor injuries is a bit strong, a few minor knocks, but nothing major. So hooray.

Speaker A: Good. I'm, I'm never, I've never skied. I just, I just don't feel I've got it in me. My, my coordination is just completely lacking and I think it would just end very badly.

Speaker B: It can end very badly, yes. Um, but that doesn't mean to say that you can't do it. I mean you see, so you know, you see guys and gals in their 80s and very out of shape but they probably, you know, they could probably ski before they could walk. So it's just like, you know, that's

Speaker A: the thing, old dog can't learn new tricks and all that.

Speaker B: Yeah.

Speaker A: So, so I'm, I'm sticking with my, my elderly person's cruise where I, I sit on my balcony and look at snow rather than, you know, interact with it.

Speaker B: Oh yeah.

Speaker A: So, so sad when you say it like that. Isn't is.

Speaker B: But I was about to agree with you that half, half, I don't know, random percentage, pull out of the air, uh, of a ski holidays appeal is the view.

Speaker A: Yes.

Speaker B: I mean swishing down slidy slidy. Yes, that's all good fun but uh, but just to just go, wow, get your mind blown every 10 minutes by the, by the view is very much

Speaker A: a big part of it.

Speaker B: Yeah, yeah, yeah.

Speaker A: And we are gonna. I mean, we are doing things. We're gonna do, um, husky dogs. We're gonna go on dog riding. That'll be fun. So I'm looking forward to experiencing that.

Speaker B: Sit at the back, Paul. Don't sit. Why?

Speaker A: Why is that?

Speaker B: See if you can work it out. So you set the back of the sled, not the front of the sled, but the front.

Speaker A: Sounds like a lot more fun.

Speaker B: Well, you've got a lot of dogs facing you in a certain direction, so.

Speaker A: So you're saying you don't want to look up a dog's ass the whole time?

Speaker B: Well, yeah. You don't. You don't. And they don't care when they do their business.

Speaker A: Oh, really? Yes, yes.

Speaker B: So. Oh, okay.

Speaker A: That is a top tip, then.

Speaker B: There you go. I would imagine that probably for the tourist version of it, that's not an issue. And I don't know, they'll somehow be covered up, but it's a thing, you know, on these kind of arctic trips.

Speaker A: Gotcha.

Speaker B: Don't want to be covered in shit, basically.

Speaker A: Yeah, yeah.

Speaker B: No, no, I doubt that very much for you. Oh, I'm very. I'm very envious. I would love to do that.

Speaker A: Uh, yeah, it sounds like it's going to be a good thing. We were hoping to do snowmobiles as well, but we couldn't quite fit it in, which was a shame. But that would have been fun. I've done them, so it's gonna be a good trip.

Speaker B: They're scary.

Speaker A: Have you?

Speaker B: Yes.

Speaker A: Are they?

Speaker B: Yes.

Speaker A: My wife loves anything scary dangerous that involves a motor, basically. She's a bit of a petrol head,

Speaker B: I was gonna say. But she quite likes skiing then. But that doesn't involve a motor. That's part of the beauty of skiing, as you're doing up to 40, 50 miles an hour without any engine.

Speaker A: Yeah, see, that just. That makes it sound even worse. As far as I'm concerned, you've totally put me off.

Speaker B: There was one point when I realized I was going too fast and went and, um, sort of put the brakes on. It took me about 50 meters to stop, but I didn't fall over. So I was like, oh, there you go. Yes, get in. Because I was just enjoying myself so much. And then suddenly it got really steep, and I'm still going as if it wasn't really steep. Then I'm like, but anyway, enough of that. Uh, sadly, there won't be any more skiing, uh, until next year because.

Speaker A: Oh.

Speaker B: Got other holidays to book. Have Booked another holiday in the summer? Yeah. Yes.

Speaker A: After our wedding. Oh, yes. What a pseudo honeymoon type thing?

Speaker B: It's a real honeymoon. Uh, because it's after a wedding, isn't it?

Speaker A: I don't know why I called it a pseudo honeymoon.

Speaker B: You're too old online, Paul. We go to Italy again. We've just been asking. And then we're going to get to Sorrento and Amalfi.

Speaker A: Oh, lovely. Anyway, I guess people don't listen to this podcast to hear us talk about our holidays.

Speaker B: I reckon they do.

Speaker A: Do you reckon they do?

Speaker B: Yeah.

Speaker A: I'm not sure. Or well, possibly just for the cathartic feeling of. Of. Of just hoping that we have horrible holidays and. And just getting bitter and twisted that we're going on about how great our. Know that it's really nice when you hear a couple of smug old men going on about how great their life is and you can judge them. That's what I.

Speaker B: There is that. But yeah, I mean, I really can't complain, but I'm about to.

Speaker A: There you go.

Speaker B: So.

Speaker A: And that makes it even worse, doesn't it? Privileged, white, middle class, relatively well off. M moaning about shit there. Go on then.

Speaker B: Just. Just the. The usual moan that, um, you know, business is tough and I think it is. Well, maybe not everyone, but it seems to be.

Speaker A: I match. It's going quite well for me at the moment.

Speaker B: Oh, good for you. Well, yeah, so now I've joined the other crew, you know, the people listening.

Speaker A: Yeah, yeah, you're now. You're now with the group of people going a smug bastard.

Speaker B: Yeah, I mean, we're making some inroads, but it's just. It's just hard work. And it didn't used to be hard work. Used to be easy.

Speaker A: No, I know.

Speaker B: Yeah.

Speaker A: Yeah, it does. It goes through stages like that, Marcus. It just, uh. You can't odds it, can you? Sometimes it goes well, sometimes it doesn't. And I've tried for years to kind of have some kind of idea about why, but it just seems completely random, you know? I mean, why am I suddenly busy again? I mean, that's just. There's nothing has changed in the world really, you know, and yet suddenly it's picked up and it just makes no sense. I just give up.

Speaker B: Nothing has changed in the world is to not. I won't get too political, but I think one of the reasons why business has been hard over the last three years is because of the current sort of up in the air style. I'll be nice. The up in the air State of things and that. Nobody wants to. No one wants to take a risk. No, no. Everyone was waiting to see what's going to happen. So, yeah, decisions aren't made. Nice. And you're right. I mean, that hasn't changed in the last. I think it's just. I mean, as I've said many times, um, for Headscape, we're only a small agency. We don't need tons of work. We don't need hundreds of things come to the door. Um, and I think the same applies for you, Paul. So all it takes is for a handful of stuff to come in and suddenly it's like, well, we've got too much to do. So, yeah, yeah, we only need a little bit of luck. And that's probably still possible.

Speaker A: Uh, it will be. It will happen. You just got to be patient and you've got to try not to freak out every time it happens. And I. I have a real problem with that because I. I've got that kind of obsessive compuls it. As I'm sure you might. You might have noticed, Marcus, over the years, once or twice. I am terrible with having to meet, um, my target every month. Right. Even if I'd had the best month ever, the month before, it's like, oh, I haven't met Target this month. Right.

Speaker B: And it.

Speaker A: It's just an obsession and I don't know why I can't let the bloody thing go, but I make myself miserable. Just over pointless. Because it'll be great next week. It was great. No, great next month. It was great last month. Why am I worried that I'm having a slower month this month? It just makes no sense.

Speaker B: Don't have monthly targets, just have quarterly targets.

Speaker A: Yeah. I should probably just hide it from myself, but I can't help it. It's. Anyway, let's. Let's talk about more positive things.

Speaker B: Yes.

Speaker A: Which is. Is the fact that Figma is falling apart. That's not a positive thing. I guess so. It depends on your point of view. Yeah. Their stock price is crashing at time of recording. It's all very dramatic and it's making headlines and stuff. And um. And I can kind of understand why, because I am indeed going down to the free account of Figma. Um, and that's not just because I am no longer a designer. Um. Although that something to do with it.

Speaker B: But yeah.

Speaker A: Well, no, because I still do lots of wireframing.

Speaker B: Yes, of course.

Speaker A: Lots of prototyping, that kind of stuff. But I'm not a UI designer, I guess. And um. The truth is there is just this plethora of AI driven tools now that I don't need to sit pushing pixels around or linking up frames to create a prototype or any of that kind of stuff that used to be such a pain in the ass. I can just. Now it's almost the way I mentally envisage it is when you're using Figma M, you're like hunched over, do you know what I mean? You're kind of down on your keyboard. And now I'm like sitting back with my hands behind my head, you know, going, you know, okay, well maybe if we try this and then what about this? And, and it's this kind of conversational process. I'm absolutely loving it. And so I've looked uh, at um, I was running a workshop on um, on uh, uh, AI and ui. And so I had to do a load of research for the workshop and I went through dozens and dozens of the tools that are available from Figma make through to Lovable to. There's tons of them. And the one that I settled on that has really worked well for me is one called Figur so F I G R ah dot design. And um, it's allowing me to create all of my wireframes and all of my prototypes now using AI. And for me, I mean I'm not saying it's right for everybody, but it checked all of the boxes that um, a lot of the tools didn't because it's not just about how pretty a design it creates. Right? Um, because you know, okay, it can produce a reasonably refined design, but it's never going to be exceptional, it's never going to be great because it's just taking the standard, you know, from everything it's seeing online. We talked about this a bit in the last one, didn't we, in last podcast about how it's going to just kind of do the average um, of UI design rather than very specific to what you want to do. But what makes this tool in particular is it ticks a load of other boxes. So it allows you to add a load of additional context things like style guides, design systems, Personas, all that kind of stuff that enables it to be more specific. It also makes it so easy to do design iterations. So um, you can, so it's produced a design and you go through it and you say, I don't like this bit, can you change this? But you don't then have to wait for it to do that one before you do the next one. You can go, I don't like this, I don't like that, I don't like this, I don't like that. And it's builds up a queue of these things and then it can go bang through the whole lot in one go. So you're not constantly waiting for it to finish doing what it does. Um, it also, if you want to, it allows you to export to figma and also allows you making. It builds fully interactive, uh, and responsive design if you need it to as well. So accordions, opening and closing button states, all the rest of it all can be built super easily just through a conversational interface. And you end up with a prototype that frankly is better than the prototype you get out of figma because it's, you know, it allows things like, oh, you can type in a text field, right? Radical stuff. I can't believe FIGMA doesn't support that. But there you go. And so, so you can create these really good, really great, um, interactive prototypes. And for me it's been amazing. There are so many new ones coming out the whole time that are even better because the only thing that figure design doesn't do, um, is it doesn't allow me to manually edit things. Right? Um, so I have to keep going back to the conversational interface. I've discovered a couple of more products that I haven't reviewed yet and if they're good, I'll mention them on future shows that will allow you to do that as well. So you can have the conversation then go, oh, I'll just change the padding myself rather than asking the AI to do it again. But that's the only criticism I've got of this product so far. Um, and I, I'm loving it, right? It's, it's allowing me to create wireframes and prototypes super quickly and easily and it's just now been incorporated into my design workflow. So, you know, whenever I am designing something, um, all I do now is in, I create a document where I outline all of the content that's got to go on the page, right? You know, I need a header and it says this and I, I usually do it in markup, um, markdown even. Um, you know, these are all the elements. This is where I want this to be. Three columns. Copy and paste the whole thing into figure design and then it produces it. Bang. Done.

Speaker B: Yeah, I mean now, yeah, we've had, we had some good experiences actually with Figma make. Almost too good. I think that's what we were talking about last time I saw these, I

Speaker A: saw these post on it yeah, and

Speaker B: it's like I haven't used it, but it sounds quite similar. It's just, you know, standard AI chat interface. Uh, and I've at least did a demo for all of us showing you, showing us what he'd done. And it's very impressive. Uh, I mean it doesn't produce anything, you know, you could ship, but for prototyping purposes, absolutely superb.

Speaker A: So I found found make a little bit fiddlier and, and I had to go back and forth with it a lot more than I do figure design. But they are essentially a lot of these tools are the same.

Speaker B: Yeah.

Speaker A: Um, but, uh, yeah, I mean, what's the cost going to, you know, Um, I guess it depends on your level of usage because the difference between these new AI tools is, is re. You get a token limit basically a number of times you can use it and if you're using it heavily, then the price can go up quite dramatically. While Figma's a fixed price at a fixed monthly fee is fourth. No, not make, but for Figma, the actual, you know, the main app. Um, so yeah, it's. But they're all pitching at roughly the same level, uh, in terms of token value as far as I can tell. There was, was one that I liked. Um, so the drawback of, of Figma, of figure design and all of these is that you're paying a monthly fee for a set number of tokens. Right. And if you like, that's fine. If you're someone like Lee or Ed that are constantly designing, they can just buy whatever token level they need and they can just kind of work through it consistently. If you're somebody like me who might have a month where I don't do any wireframing at all. Right, that's annoying then that you're. You're kind of paying a, uh, token level that you're not using. So there was one other tool I found which is pretty good. I'll put a link in the show notes AI. I'll have to remind me to do that. Um, which is, uh, a tool called Inspector and that runs locally on your machine. Um, and what it will do is you can connect it to the Claude API, which means you're paying as you go. Right. So instead of paying a flat monthly fee, you're paying as you go. I've had some problems getting it to work, which is a bit annoying, but that looked very good as well. I got it working for a while and then I had issues and, and so I've given up on it at the moment. But I think what, uh, I'm kind of encouraging people is, look, take some time to look at some of these tools and don't dismiss them out of hand because you don't get a good. Design them the first time you try. Because I, whether it's good or bad, I'm not going to get into that argument. But I do think this is the way things are going to go. Um, and so I think people need to be, you know, beginning to get comfortable with these tools because sooner or later you're going to be forced to use them, which is a bit depressing because it's like we just got ahead around Figma after, you know, what was it, Sketch before that and Photoshop before that. And now it's, oh God, new tools to learn.

Speaker B: I wonder if that's. That is. I mean, because that was kind of where I was going to go. I was going to say, you know, should we be using these tools, um, from an environmental perspective? Um, and you're right, if, if certain jobs will demand speed, um, and they will demand efficiency, etc. Etc. Um, but I guess the question here that I'm just, I'm spouting this is coming straight out of my head. Should we be offering two different levels of. You know, we all. I've just done a proposal where we're talking about the social value that we can bring and the environmental value we can bring. Well, it's like. Well, yeah, hang on then. Yeah, we, we can, we can develop our wireframes manually or, you know. Yeah, uh, it's just occurred to me as we were talking about this.

Speaker A: It is a. Yeah, I mean it is a valid point, I think for years of a couple client that is, is deeply concerned about the environmental impact of AI you can, you can say to them, look, all right, I won't use AI for this project. It will cost more, obviously, because it will take longer to do. Um, but, you know, fair enough that I'm, I'm completely open to that. I mean, I feel like if I'm honest that the, and I'm gonna get in trouble for this. But you know, I'm not, not claiming to be an expert. This is just what I feel and probably somebody needs to educate me differently. But from what I can gather in my limited understanding of it, the environmental impact is significant, but only when taken in isolation. When you compare it to the environmental impact of other things that we just take for granted. Right Then it's minuscule. So, for example, the energy consumption of, of, um, AI queries is tiny. Compared to you eating a burger. Right. Um, and we don't, you know, wring our hands every time we eat a person. We probably should be all cruise ships. Yeah, Right. Or flying to Italy twice in a year, you know.

Speaker B: Yeah.

Speaker A: And equally the water. Water is the other consideration because these data centers use a lot of water. But compared to America's addiction to, to, um, growing, uh, corn, because it's a highly subsidized things to support the farmers, corn uses way more water than AI will ever use. At least that's my understanding of it. So it's all a little bit relative, I think, is, is my point. That's not to excuse it. And if, if clients want to avoid AI, then I would entirely support them in that. That's their call.

Speaker B: You know, just an interesting point. I mean, another proposal I'd recently completed, uh, asked if I had used AI in the completion of the proposal, which is interesting, but I was why. I wanted to get why, but I didn't.

Speaker A: I hadn't used AI why are they asking it? Why would you use AI?

Speaker B: Why are they asking?

Speaker A: Right. That's interesting. I use it. I mean, I had a, I had a meeting this morning, so I, um, with the National Trust. Right. And I put together a proposal for them a year ago and we had a. Had a meeting. I'm going on holiday tomorrow. Did I m. Mention I'm going on holiday? I don't think I did. Um, so. So they wanted an updated proposal and it was like, well, I'm out of time. And so I just got AI to do it and it, it, I'd recorded the meeting, so it heard everything that was in the meeting. It had the previous proposal and it rewrote it in 10 minutes. In fact, I actually put off sending the email because I didn't want it to look like I'd put no effort into it.

Speaker B: Well, yeah, that rings a bell, uh, with Lee's story. Yeah, um, yeah, it's just an interest. It's an interesting ongoing subject.

Speaker A: Yeah, it is. And we're actually. It's a subject we're going to explore even further. Because the main topic that I wanted to talk about today is really how I've completely changed the way that I rebuild websites for clients. Um, and it's so different now to the way that I used to do it, and largely because of AI, not entirely, but. But, you know, a big part of it. And so I thought I'd just talk it through because I think it's so easy not to challenge the fundamentals of the way that we do our jobs. Um, and so yeah, I, I've kind of radically changed the way that I do build websites and I do still build websites. So people think I don't actually do that. Um, I, I don't, I don't do the UI design, I don't do the development, but I do run the projects essentially for the, for the clients because there's a lot of clients that just want somebody who's experienced in it just to run it all for them. So, so I tend to do that quite a lot and I've completely changed the way I approach it. Um, and to be honest, the way that we rebuild or read, I'm avoiding the word redesigning websites because it's not just about, it's not just about the design, it's about the technological platform, it's about the content, it's about the whole things. But m. Use whatever words you feel comfortable with. But when I rebuild websites, I, I think there's been a big need for years to change the way that we've done this. Right. So we often, um, redesign and rebuild really just by migrating all of the existing content across into a new pretty interface. And that's always been right. Um, uh, but it's been, it's been kind of to some degree excusable just because of the huge workload involved in, in recreating content. But whenever you migrate content across into a new website, you're never going to see that much of an improvement in performance because ultimately people go to the website for the content. They don't go. Unless your website is a complete train wreck from a usability point of view, the content is always going to be more powerful than the design or the technology on which it's built. Um, and if you migrate content across, you're bringing across all the rot, the redundant, the out of date, the trivial content. Normally most content on most websites is written by people that don't have a clue how to write for the web. And that's not a criticism of them. You know, they're experts in their own field. Um, they're not website copywriters. Right? Where are you going like that for? Marcus, you said most.

Speaker B: Maybe that's true. Um, I don't know. I feel that today our clients, and maybe they're, they're not normal, maybe they, they do fit under the, the opposite of most, but they are much better at writing content, um, and be. And content aware and how to do, you know, associate their content with social media and all that kind of thing. It's not okay. Quite. Yeah.

Speaker A: All right, I can accept that. Oh no, it's not as bad as it used to be. But even so, I feel like, um, that's. Yes, I would agree for some content. Um, so I would agree when you're talking about general marketing content that's being written by a digital marketing team that knows digital. That's fine. Yeah, you know, they're going to be great. But it's where you get content written by subject matter experts. Right. So maybe it's a research project to the university or maybe it's uh, you know, uh, like I'm working for a company that produces, um, nursery trees. You know, like, um, you know, trees they sell to people. Well, it's not actually.

Speaker B: Yeah, yeah, yeah.

Speaker A: Kind of, but that kind of thing.

Speaker B: Saplings.

Speaker A: Saplings, that's the word. Right. So obviously they don't know anything about writing for the way. So it kind of depends on the client. It does. Uh, yeah, um, so most. Yeah. So stakeholders don't tend to know how to write good web content. Content. And of course the other problem is that they don't have the time either. Right. So these people have got full time jobs. Um, and so where do projects always stall? They always stall on the content delivery. Right. You know, everybody asks you for these. When you write a proposal, you have to put in your timeline and you know damn well that it, that you won't meet those deadlines. And 99% of the time it's because the client isn't delivering something because they've got a full time job. And I don't think they always realize how big a job content creation is, at least in my experience with my clients.

Speaker B: I've just done, uh, a timeline. Again, it's the same, one of the same proposals that mentioned earlier. And the content production starts at the beginning of the project and ends at the end of it. It's a bar that's the length of the entire project. Yeah, yeah. And it's, and it's a color. The color is associated with the client. Not, not to degree with us. But yes, I'm agreeing with you and I'm disagreeing with you. I think it's where, where teams have, have people dedicated to writing and looking after content of which I think there are a lot more of those. I think that's where I was going with it. It's breath of fresh air compared to how it used to be. But of course I don't know whether the, you know, the team that we, we did this proposal for has anyone in you Know who is, who has the skills or it's even part of their job, in which case nightmare ensues. So Paul, tell me how to do it.

Speaker A: So, so, okay, so first of all, why things are different now? Why is the time when, when you can start doing things different now? Obviously AI is, is the big part of it. AI makes content creation better. And um, no, not better, that wasn't the right word. Easier than it's ever been before if it's done right. Okay, I'm putting that big caveat on it. Also, AI has made online research simpler because there's another problem here with content creation. When I say people don't know how to write good web content, I'm um, not just talking about they know how to keep the reading level down, they know how to make it scannable, all of those kinds of things. I'm also talking about not just lecturing people on what a great company you are, but actually addressing real user needs. Which even if you know intellectually that when you work in an organization you become institutionalized, it's an, an inevitable part of working somewhere. And so it becomes increasingly hard to think in that user centric way. So AI makes some of the online research simpler. That I'll come to in a minute. Um, there are better uh, tools out there for doing AI work that's content orientated these days. And we've also got to start adjusting our approach because of things like AI search engines which demand a more question orientated approach. So really AI has shifted the whole landscape. So how do I work? Right. First of all, disclaimer. Right. I'm going to say something in a minute where I'm going to say you should start from scratch. And I don't literally mean that.

Speaker B: Right.

Speaker A: Um, the disclaimer is you can of course you can migrate some stuff, right. Old articles, right. That, that just need to be compliance information that just has to be on the site. Uh, you know, old, you know, events, maybe an events database, all of those kinds of template based kind of boring content.

Speaker B: Right.

Speaker A: That's just got to be there. Sure, go ahead and migrate it. Although I would say probably even with things like some of your blog posts and stuff, they could probably do with being rewritten to be better and more friendly from a search engine point of view. But you're not going to want to go back until to 1994 and you know, rewrite all of those. You'll pick and choose what you'll write. But basically most of the time now, even if somebody has got a relatively large Existing website. I will say, okay, let's imagine, uh, and start from scratch. Okay, we may end up migrating stuff, but let's pretend we're not doing that. And so the first thing I do is I immediately go online and I just do online research. Now if I could do a proper, you know, research phase with user interviews and all of that kind of stuff, I'll absolutely do that. But as a minimum, I'm going online with something like Perplexity, and I'm saying one of taking one of two approaches depending on how big the client is. If they're a household name like the National Trust, I will ask specifically about the National Trust. If they're not a household name, I'm going to get the research around their sector. So the heritage sector in that example. And I'm going to tell Perplexity, go and do online research into either the company or the sector, looking for, um, what people are saying about that sector in terms of what questions are they asking, what tasks are they trying to complete, what objections do they have, um, to working with them, what's their goals of working with this kind of organization, what are their pain points, etc. And it will go off and create a massive report of all kinds of stuff. It's found across social media, across maybe research that's been done across websites, all the rest of it. Then what I do is I take those, um, and I turn them into Personas. So I use AI to say, okay, begin to find, um, different segments and different groups of people and different commonalities across the research and let's create a set of Personas around those. Right. So I've got a set of Personas. Now some online research. Then I, I will write up a overview of the company's offerings. So just kind of short page. Normally I just point AI at the meeting that I have with the client when they explain what the hell they do. Right. So it now writes me a nice summary of what the company does, what their products are, what their services are. Great. So that's my core information. Now if time allows, I'll run a top task analysis. Okay. So this will allow me to collect all of the questions that a user might, uh, have, um, about the products or services, um, and prioritize them based on what, um, is most important to the users. If I can't do that, I get AI to brainstorm those questions for me based on the user research, based on the Personas, that kind of stuff. Okay, yep. Once I've got that, and if time allows, I will do A card sorting exercise, right, Where I'm now creating my information architecture. I tend to use a tool called UX metrics for this uxmetrics.com um, and I will go through all of that process right, using those top tasks that I've identified. So I've now got an information architecture based on my top tasks. If I can't don't have time or budget to do card sorting, then again I will get AI to um, just create me an information architecture based on those top tasks based on the company profile based on the Personas if I can throw in, oh, and here's the results of some card sorting, that's a bonus, right? Um, but ideally that's what I want to do. So we're now got an information architecture load of questions that people want answered. Personas, company profile. Next stage then is I build out the information architecture. I tend to do it in notion because I'm a big notion geek and all of the rest of this information is going to be in notion anyway and I use notion AI. But you can use whatever tool you want in. Probably the best tool is actually the CMS you're going to build the website on. So even though you've got no styling or anything like that, you can still create the information architecture. You could give it basic navigation and then we end up with the information architecture. Then I assign ah, each of our tasks from our uh, top task analysis to various pages in that information architecture. Okay, so we've got an information architecture that's navigable with all of the questions that need answering on each of those pages. At this point I then go to our subject matter experts, right? This is where the client comes in all their internal subject matter experts and I say to them, I'd like you to review each of the pages in this information architecture. Um, and I would like you to do two things. One, just bullet point answers to the questions on those pages, okay. And two, make a note, bullet points again is fine. Make a note of any other talking points that you would like to cover on that page. Just bullets. Absolutely fine. Right. So they've done that. Now I sit down with AI and I get serious. Okay? So the first job I do is I get AI to write a web copywriting best practice guide, right? So this is where I say it needs to be scannable, it needs to be seventh grade reading level or below. It needs to use bullet points, it needs to use content blocks, headings, blah blah blah, blah blah blah. Any of the generic inside. And in fact that applies to every project so I'm basically already pasting it. Yeah, I've already got that. Secondly, I, uh, then get it to write a content style guide for that company. Right. So this is tone of voice. Are they friendly, are they whatever else? Now what I would traditionally, uh, typically use for that is any existing material that they've given me, maybe they've got a brand guidelines, maybe I might use their existing website if they're relatively happy with the tone of voice, just to kind of get an overall sense of how they speak. So for example, I was working with a guy, um, who's got a huge social media following, who does. You know, it's got quarter of a million people on YouTube. So I just took a load of his videos, took the transcripts and looked at how he spoke. Right. To create this, this style guide. So then I, um, I also, with that start that, that tone of voice style guide, I point the AI at the Personas that we created and at the company profile so that it can take those things into account when it writes that, that style guide and then that goes off to the client to make. They're happy with it. Are you happy with this tone of voice? Right.

Speaker B: Yeah.

Speaker A: And often, you know what, like at the beginning of projects, normally you've asked them for keywords that represent their brand anyway, haven't you? So those can get built into it because they say we're friendly or professional or trustworthy or whatever else. So A.I. now, yeah, we're either. Don't get that a lot, surprisingly. It's surprising. Yeah, yeah, yes. Um, that's the. One of the things I started doing actually with those brand was this bit of a tangent. But if they say something where the opposite isn't valid. So if they say we're friendly. Right, well, hang on a minute. That's a bit. Yeah, you don't get to have that because unfriendly is not a viable alternative. Right. Anyway, that's beside the point.

Speaker B: Yeah, you're not allowed. We used to do the, um, the famous person thing. You're not allowed to have Obama and you're not allowed to have Steven Fry. It's like.

Speaker A: Yeah, it's too easy.

Speaker B: Come on, everyone's gonna think a bit.

Speaker A: Yes, exactly. So now this all gets packaged up together. So the web copywriting guides, the company style guide, the Personas and the company overview. Right. All get packaged up into a set of instructions for your AI tool of choice.

Speaker B: Right.

Speaker A: So on all of these platforms you can create either projects. They're normally called projects, I think chat. Uh, no, Gemini calls them gems because, you know, let's get the branding right. Um, and, and you can just dump all of these in, give it instructions about, you know, um, to refer to all of this information. And then I, by the way, my top tip is, is use. If it's writing you're doing, seriously consider using Claude. I think Claude is by far the best writer out of all of them. But anyway, that's beside the point. So now all I'm doing is I. For each page, I'm grabbing all of those questions and bullet points and talking points, popping them into AI and say, write me the content. And it's going to use all the information that it's gathered about the company style guide, the Personas, the company overview, the, the how to write for the web. And it will write the first draft of content for you, right? You can churn through pages like nobody's business that way in a really, really short length of time. And then once those. You got that copy, you just paste it over the top on the website. So instead of it having those bullet points, now it has draft copy. And then you ask the stakeholder to go back through, right? All of that copy, just make comments, right? Comment on whatever you want. This is why I like doing it in notion. Um, come, you know, anything you don't like, or anything that sounds wrong or anything you don't like, just put comment in. And then I literally go back to AI and say, check out all the comments and rewrite based on them. And then that's the second draft and then it's passed back to the stakeholders one last time for the tick. Yes, that's ready to go. You can burn through so many pages that way. And basically I do pretty much that entire thing for, for the client, rather than asking the client to write the content and waiting on them. Um, and it's a really, it's a really valuable service that most clients are happy to pay for because it's not the huge cost it would have been for us, you know, in the past for you to write all the content for them. And it saves them a load of time and it improves the overall quality of the content. So that's kind of how I'm approaching content these days when I went with websites.

Speaker B: Yeah, I think that's really cool. Um, maybe offline we could talk about how long it takes.

Speaker A: I'm happy to talk about it.

Speaker B: Depends on how many pages you're looking at, isn't it?

Speaker A: Yeah, I reckon so. So setting aside all the research and the upfront stuff, so doing do it.

Speaker B: Right.

Speaker A: Let's go through it. We can do it quickly. So online research with perplexity. Ten minutes, Right. That's all it takes. Creating the Personas, couple of hours probably to, to faff around with them and that kind of stuff. Creating the overview of the company offering. Well, if you've had a meeting with them and you've talked about it. Ten minutes again. Right. It can create that. Running to our skip running top task analysis and card sorting because that's all, you know, relatively complicated. Uh, creating your information architecture, final information architecture with AI. Call it a couple of hours again. Right. Um, setting up your CMS or notion tool. Probably another couple of hours. Right? Yeah, I know I'm saying a couple of hours a lot, but I'm being vague. Then assigning all of the tasks to the pages. If you're doing it in notion, AI can do it automatically for you. If you're having to copy and paste it into your um, cms, it's going to take a little while to do that. Call it a couple of hours. Um, then of course you've got a delay while the stakeholder is going through and putting their bullet point answers in and their talking points. So that's not you that does that.

Speaker B: Yeah.

Speaker A: Then you've got to create the web copyright best practice guide. Well, once you've got one, you can use it on every project you ever work on, right?

Speaker B: Yeah.

Speaker A: So that only produces. It takes one go. It's a couple of hours again, you know, to, to do something like that, to get it right, to go backwards and forwards and you know, debate about it. Although it does a pretty good web web first draft, so probably I'm overestimating. I suspect it's an hour or less, to be honest. Company style guide, well, that takes a little bit more work because you've got to uh, find your different sources of information. You've got to go back and forth with the client a little bit over that. You've got to match it with the Personas. Three or four hours, you know.

Speaker B: Yeah.

Speaker A: Um, the company overview, that's fine. We've already done that. Then it comes to the big part, which is the, the getting AI to go through each of those pages. Now in truth, it doesn't take long. It depends on how much fiddling you want to do with it. Right. So a lot of the. I'm hired a lot of the time to do conversion rate optimization work. So I'm very fussy over the wording, I'm very fussy over the flow uh, of the page. So I can be spending probably 20 minutes to half an hour on an individual page. Right. Um, but if you were just rebuilding a website, five minutes a page, just to run the command copy and paste it across into the cms, it is that simple to do. Now, when they come back with the comments, right, um, and they've got a load of changes that need to be made to the page if those comments. Again, this is where Notion works really well for this because Notion AI can read the copy and see the comments. Right?

Speaker B: Yeah.

Speaker A: And so two minutes, right, Just for it to go through and rewrite the stuff. If, however, the comments are separate to the content and you have to bring it together, you're probably talking about 10, 15 minutes, I would have thought, per page, just to give it. Here's the content of the page, here's all the comments that I've got. Now run it. Yeah. So it's not a massive job depending on how many pages you're producing. Yeah.

Speaker B: I mean, you probably turn it all around in a week, easy. Yeah, yeah, okay. Yeah, Interesting.

Speaker A: So I think it's quite a good way of working, you know. And, you know, if you're a massive university with 200,000 pages. Well, first of all, I would argue you've got bigger problems and actually need to seriously reduce the size of your website.

Speaker B: Well, you do not view it As a, as 30 different websites and you attack it. Yeah, that's a different way. Yeah.

Speaker A: But that basic working methodology, it works so well, Marcus. I've used it on half a dozen projects at the moment and I finally feel like I'm not just slapping lipstick on a pig, do you know what I mean? Because if you don't address the content, I don't feel like you're really doing a rebuild, you know? But anyway, that's just me.

Speaker B: No, no, no. Yeah, I think that's absolutely right. And in the past, there has been too much onus on people who, whose job it isn't. I'm just repeating what you're saying here and using different words, um, to do this huge part of the project that half the time they didn't realize what they, what they were, you know, what the, what that role meant. I mean, it's, you can, you can look at it in different ways. I think that a lot of websites do include a lot of content that you are just going to pull over. News is a classic example. Uh, you know, we work with a lot of law firms and their news and blogs and thought leadership is all really important. Uh, and it's not going to change. It just has to come over and be reformed.

Speaker A: Fair enough.

Speaker B: Bios are the same thing. I mean often they will rewrite bios, but there's kind of a bunch of important intro pages that probably should be

Speaker A: the ones that talk about the products and services that you're actually selling, you know.

Speaker B: Yeah. And the leads lead into all that, the detail. They're the ones where this is ah, a really useful approach. And even if, uh, even if the client sort of maybe feels that there's a little bit too much, it's a bit too hands off, it can still be the starting point.

Speaker A: That's the point. I mean the other thing that I do, I mean with some clients, like, um, I've got a client at the moment and it's like I've done probably 20 or 30 pages for them using this methodology of the main, different template pages. And now I'm just going to hand the package, the AI instructions and all of that across to them and they're going to run the rest. Right. And, but they're still not having to spend hours poring over this content. But neither are they just creating AI slop, which is, it seems to go in one way or the other. Either it's a huge job or it's a load of generic AI slop with EM dashes everywhere and you know, and, and statements like and here's the trick or this will blow your mind or all of that crap that it puts in by default, you know. So the last thing that I want to talk about in the show today is an article that I came across, uh, which is another Nielsen Norman article. I do tend to talk a lot about the Nielsen Norman Group's article because they tend to do some fairly deep thinking into different issues. And I, I liked this one. Um, it was about um, the state of UX in 2026. And um, it's titled Design Deeper. Ah, to differentiate. And I really agreed with the vast majority of this article because of um,

Speaker B: the things you've been saying for the last six months to a year.

Speaker A: Paul, I love it when people agree with me and then it's okay for me to agree with them. It just proves I'm always right. Is, is fundamentally what we're saying here. So anyway, good news to begin with. Um, it, it, the article says that the UX has very much stabilized since the big, the crash that we saw over 23 and 24. Um, and I would say, generally speaking we're beginning to see that. I think it depends on where you work, um, and how far behind there is. But I think we're coming to the end of that period. Um, but what we're seeing out the other end, which I think is a good thing, even though I know it's been really tough for many teams, is much leaner teams, um, uh, which means that UX's have to cover a lot more ground and are under more pressure to prove business impact. And I think actually that's a good thing because I think too many UX

Speaker B: teams were less fiddling with, with their software.

Speaker A: Yeah, less fiddling with individual projects and you know, getting sucked into um, the details of individual projects and a lot more, more strategic thinking, which is, I think what we really needed within ux. UX was always supposed to look at the entire user experience from end to end.

Speaker B: Right.

Speaker A: Not just one interface somewhere. You know. Um, then the other thing they talked about was AI has moved us, moved um, past the hype, into the AI fatigue, um, era, which I would agree with as well. Um, uh, and so I think there's, there's beginning to be a move beyond the oh, designers and developers are going to be replaced with AI. I don't think anybody believes that anymore. M and, and I think as well there is a reaction amongst users towards what I'll call what they call, um, lazy AI features, you know, where it just adds one of those sparkle buttons to the interface for no apparent reason, just to please shareholders, you know. Um, so we're getting, we're making progress there as well. And so now it goes on to argue that trust is now the main central design problem for people working on AI tools. Because whenever you add AI functionality into anything these days, I think people are very, very cynical. Um, and so there is all kinds of concerns around privacy, around what you're going to do with your data, around the quality of what's being outputted. So there's a lot of work that we need to do from a trust perspective. Um, and then it goes on to talk about how UI is really almost become commoditized so that the UI work of actually just building the user interface really is uh, m becoming so much easier as we were talking about at the beginning of this podcast. Um, and so equating UX and UI is really dangerous now. Um, the ux, you kind of got to redefine yourself more from the UX point of view that you're looking at the bigger picture, you're thinking at a more um, strategic level and focus your skills in on those hard to automate Skills, things like taste, contextual understanding, critical thinking, judgment. Right. These are the things that we still provide value over. Um, and, and really if you want to Thrive As a UXer, you need to position yourself as a strategic problem solver, uh, with a broad toolkit of research, stakeholder M management, AI fluency. Not just a delivery factory. Right. That produces what they call design theater, um, which I quite like that phrase. Really good article. Totally agree with all of those points. No doubt Marcus will disagree with something just on principle because I picked the article.

Speaker B: Yeah, I, I don't know. It must be hard being a UX person in a medium to large uh, organization because I can't imagine, I can't really imagine that I do UX work every now and then. Um, but it's just sort of part of loads of other stuff I do. So to be somebody who's very, who trained in it and went and got a job and then lost that job job, that must be hard.

Speaker A: They, I mean and, and they have cut back too far. I was talking to one company recently, a big national brand and in the entire organization they had one UX person. I mean that's ridiculous, you know, so I think it will balance itself out.

Speaker B: I mean they just become a box ticking exercise then you know.

Speaker A: Yeah.

Speaker B: Send it past the UX person. So yeah, finding the balance or obviously um, higher headscape. That's the other way to deal with it.

Speaker A: Well that's a far better way I think, I mean. Well there is a degree, it's an interesting debate that one because you know I've, I'm involved in building out UX teams within organizations and there is, there is a lot of complexity with doing it properly. Okay. One is you need far more people than you think you do. Right. And that's a big headcount demand. Um, and if you have two.

Speaker B: Cost a lot of money. Yeah, lots of people cost lots of money and they do.

Speaker A: Yeah. And, or they end up with too small a team and they kind of kid themselves into thinking that's enough when it really isn't. You're almost better going. Well if we can't afford to do it properly, you've got two alternatives in my opinion. Opinion which are uh, to be honest, interlinked. They kind of go together. One is if you're going to have internal staff, you create it as a center of excellence. So their job is not to build user interfaces or deliver products. Their job is to ensure standards, to provide resources, to um, you know, provide training, that kind of stuff and then use Outside freelancers and agencies to do the doing is. Is increasingly the way that I'm positioning things because it's just unrealistic otherwise.

Speaker B: Yeah, I mean I have conversations with some of our clients along the lines of, with us, you get all of the kind of web digital development, M. Apart, uh, from content, a little bit of content. But you know, we, you know, design development, UX research, all of those things. Strategic thinking, you get all of that which you hire us for. Whereas for the same price that you might get one person for just, just a developer, uh, or just a designer. So. Yeah, so it makes sense. Kids, folks,

Speaker A: that is just such a blatant pitch. I thought you were better than that, Marcus. I know things are a little slow

Speaker B: at the moment, actually. No, I can edit this guy. I'll put it at the start.

Speaker A: Yeah, just put it at the start and again in the middle.

Speaker B: And again.

Speaker A: Yeah, yeah. Repeat it multiple times through the podcast. High headscape. You could do it, you know your music at the beginning, you should just add into it. High headscape. High headscape.

Speaker B: I could do it really quietly all the way through. Subliminal, that noise. What's that?

Speaker A: Yeah, it's a great idea. Okay, do it. I'll do it for the next. For the next show or whether. Do it for this show because you can still edit this show. Okay, go on. I dare you to put it into the titles sequence at the beginning. Right, Marcus? Yeah. You're not sure about it.

Speaker B: You talk a good talk but it's just laziness. Can't be asked to do it.

Speaker A: Yeah, I don't think it will actually lead to more work if anything. I think it probably would just alienate.

Speaker B: Yeah, it'll be the exact opposite, won't it?

Speaker A: Yeah, yeah, it will be. On principle, people won't ever talk to you again. Right, Marcus, what's your joke?

Speaker B: Right. I stole a neck brace from the hospital. I feel kind of bad, but at least I can hold my head up high.

Speaker A: I must be, I must be in a holiday happy mood because I'm laughing at your jokes. So yeah, that is quite a good one. All right, thank you so much for listening guys. We will be back again next month and ah, in the meantime, if you've got any questions or if you want to chat over any of the stuff that we've talked about, both me and Marcus are absolutely up for a chat. Um, either hit us up on LinkedIn or just drop us emails. Um, you're marcus.lillington headscape.co.uk Correct. Is that right?

Speaker B: Thank you. Yes.

Speaker A: And I'm Paul@boagworld.com and we. We do just like chatting with people. So if you want to, you know, hire. Marcus is only interested in talking to you if you want to hire him. He's not really a friendly person.

Speaker B: That's not true.

Speaker A: Uh, it's not, actually. You're far more friendly than I am.

Speaker B: Yeah, I'm happy to. I'm happy to. Just, I, uh, was having a chat with a. Somebody from a Californian, um, university the other day where we basically had the beginning of the chat was, um. We aren't really. Can't really work with each other because it's too far apart. But let's just have a chat about this, that and the other, which we did. So it's fine.

Speaker A: I like stuff like that. I've started taking. Sorry, we really ought to wrap up this show. I started taking contacting people on LinkedIn now and saying, do you want a virtual coffee? I've no agenda whatsoever. It's just that since I stopped going to conferences, I don't talk to anyone anymore. Right. You know, I just don't network. And I've had some amazing. I was talking to a guy from Ernst, uh, and Young, like, a couple of days ago, and it's like, yeah, Ernst, and now you're never gonna hire me, are they? You know, they do what I do, but at a much bigger scale. But, um, it was just so great talking to other people and he had loads of great ideas and had great perspective on stuff. So, yeah, I love all that shit. So, seriously, if you are. If you are listening to this and you fancy a chat, drop either of us an email. I mean, obviously, if you're scum, um, we won't talk to you and reply. But if you're a normal, nice person, you might consider responding.

Speaker B: Excellent.

Speaker A: That bombshell. Yeah, we'll finish. Thanks for listening. Talk to you next time. And goodbye, Sam.