What Does a WebflowAgency Sell Now?

0:00Janne Parri: Okay, Stephen. So has Demandflow started offering other technical implementations besides Webflow?

0:08Stephen Wise: Yeah, so maybe over the last six months or so, we've started to offer a few different variations beyond Webflow. So previously, probably to be honest, for the last like three, four years, we were pretty much like ninety-eight percent Webflow solutions. But then over that six months or so it's been, you know, Astro with Sanity CMS, and then also a few kind of Next.js builds over that period.

0:29Janne Parri: That's interesting. Did that start from you or did the clients want or ask about other solutions?

0:36Stephen Wise: Yeah, so to be honest, I feel like with everything, I'm just really curious about different ways to build things, different ways to try, what is the new best way to do different pieces. So to be honest, the majority of this was a temperature in the market, seeing people looking at other solutions, people exploring different ideas, but then also just playing around with the agentic development and just kind of comparing that to the current workflow in traditional Webflow. Obviously, with the current like traditional Webflow, what we have today,

1:03Stephen Wise: you connect with the MCP, but one of the biggest issues that we found is that there isn't this full loop when you're working with the MCP where agents can go and check their own work. So what happens is it's kind of a bit more of a one-way connection, and obviously Webflow is an external place. Whereas if you're developing locally, it's the full loop there. So the agent can go and check its own work. So therefore the work it does is typically a lot better because it knows, hey, I've made mistakes or I haven't made mistakes. So what then happened is we start to play around with different ideas,

1:32Stephen Wise: play around with building, see what the pros, what the cons are. And then it was a, okay, now we're gonna start to suggest this to clients in certain scenarios where it makes sense for their projects. And it's really just an exploration because ultimately the solution that we sell at Demandflow is the marketing website as a system that the marketing team can operate on. Most of the time, we're still selling Webflow. It makes sense in most cases because there's usually a team involved. Maybe they're not super technical, but over this period of time it's definitely been more and more of a focus of how do we build that system and what is the new normal? What does it look like in 2026 going into 2027 to build a marketing site?

2:11Janne Parri: Yeah, yeah. Those are obviously like the things that everybody is thinking of, everyone is trying to solve. But right now what's your feeling or two cents about who should try out these more open environments like Astro and such for an actual production site?

2:29Stephen Wise: Yeah, so I think what we currently find is that handoff period is still the most difficult thing. It's like, okay, what does it look like when you hand a website, you hand, you know, a client the keys to a website, what happens then? Because we all know that can go wrong very quickly. And one of the things for me, like to get into the Webflow space maybe like seven years ago, was that I was building sites on like WordPress and just like some, you know, Squarespace or whatever, like just different tools. And one of the biggest things with Webflow is obviously, sure, at that time

2:58Stephen Wise: coding was a lot more difficult. So being able to do it visually was a huge sell. But it was also the handoff. Like, how do you work in there? Kind of designers, developers, and marketers can work in the same place. Right now, with this kind of, you know, going a custom stack using, you know, Astro, using Next.js, and building with Conductor, which is typically the tool we use to kind of connect Claude and Codex, still handing it off to non-technical marketers is a little bit difficult, right? There's no interface. And

3:27Stephen Wise: there's people trying to solve this problem right now of like, what is that interface? How does everyone work together? It works really well if you have technical marketers, right? If they understand how to, you know, connect their site to Claude Code, connect it to GitHub, and they feel comfortable making those changes, perfect.

3:43Stephen Wise: The tricky part is when you have non-technical marketers who are now having to work in production code and they're understanding that, hey, I can ask it to do something, and then it might break four things over here, but it did do the thing that I wanted it to do. Now you're in this weird situation where now you just have more and more work. People are pushing changes to production that they shouldn't have done, and like the guardrails are just not quite there right now. So what we find is it's very much a test of how technical are the marketers based on what solution we offer and

4:12Stephen Wise: recommend, because that is where the biggest gap is right now, is handing it off in this whole agentic workflow. Like I firmly believe this is the way we build moving forward. But there's definitely a gap. There's definitely like a lag in terms of being able to give it to non-technical people and then just be like, okay, here is how you do it. Here's the boundaries, go and, you know, launch campaigns, change text and do that kind of thing.

4:35Janne Parri: Yeah, yeah. I'm already getting excited because I feel like that gap, everyone wants desperately some sort of solution for that. And have you found any, like, possible solutions, apps or software, or are you waiting for maybe Source or something else to figure that out?

4:55Stephen Wise: Yeah, to be honest, right now, the main thing is just exploring all solutions and seeing what fits our workflow in particular. Something that I'm finding is that the more that we do this, the more that you understand that like you can customize technology now to what it is that you prefer, right? If you have certain workflows, you have certain things that you need, you're able to now build your own software on top of different platforms and kind of create these own experiences. At the minute, I know Timothy Ricks is building, and he's kind of built that interface over Astro, which is really

5:24Stephen Wise: interesting. Ship Studio is another one we've been playing around with, which I think is really interesting. You can also create visual experiences like with Astro and Sanity CMS as well. And then obviously there's Webflow Source, which is kind of like looming in the background as a potential solution for this problem that might come in in future. I think like right now, the biggest thing is I don't feel like there's anything that is at a point where I would be confident to put it in front of a client and say, hey, like for the next 24 months, this is your solution.

5:52Stephen Wise: Because when something is pretty small in terms of like the amount of people behind it, you obviously have that risk profile, especially if you're working with a slightly larger client. So at that point, it's now, you know, we're assessing what does it look like for us to build our own visual editor, what does it look like for us to use these different tools? And we're still very much just figuring out what that world looks like because we have this gap, right? The non-technical people, they're in traditional Webflow, makes perfect sense, it's a great place for them. We can build the system and everything. More technical, and if they

6:21Stephen Wise: want to go down that route, we can put them on GitHub, we can use Astro, we can use Sanity CMS, but the in-between part, like there's just not quite a perfect solution there.

6:30Janne Parri: Yeah, yeah, definitely. Yeah, I feel like many people feel that. Okay, before we go any deeper we need to do little intros before we forget. So you are now listening to Webflow and Friends podcast. My name is Janne Parri and my guest today is Stephen Wise,

6:47Janne Parri: founder of Demandflow, a design and Webflow agency that builds marketing sites for high-growth B2B companies. Stephen started as a freelance web designer. He launched Demandflow in 2022 and it has been a Webflow partner since that very year. Today Demandflow is already a premium enterprise partner, so the highest tier there is. Your team

7:16Janne Parri: has over ten people in it and you have a portfolio full of cool projects and clients. But I'm a bit curious, like, what's the backstory of Demandflow and what makes you different from all the other agencies?

7:31Stephen Wise: Yeah, so appreciate you having me on. Really, really love what you're doing. The Demandflow backstory is very much just an evolution of me being a freelancer. So when I was about 17, I was trying to figure out what do I want to do with my life, what career do I want. You're studying at college and you're thinking, okay, when I go to university, what do I want to do? For me, I never really found anything that I wanted to do besides run a business. And my dad was an entrepreneur as well, he ran a business, and I had no idea exactly

7:59Stephen Wise: what I wanted to do, but when I turned 18, which is the age in the UK where you have to stay in education until you're 18 before you can drop out, I decided I was going to drop out and figure it out. I'm going to be a freelancer. Started as a copywriter, tried freelance copywriting, didn't make much of a success there. Tried all these other things in terms of like e-commerce stores and just, you know, all the little side hustles, the traditional things. With this, every single venture you build a website. So what I decided I did was, hey, how about I can build websites for other businesses?

8:30Stephen Wise: Over the course of the next couple years, that started to get a little bit of traction, working on Upwork, working all over the place. From that point, I was working in WordPress, working in Squarespace, and having all the general kind of nightmares you have on those platforms, specifically WordPress, of just like everything

8:46Stephen Wise: breaking, always just spending all your time updating plugins and everything. Then started to find Webflow. I think I actually saw it, I remember seeing an ad for Webflow like pretty early on, because I was trying to work a lot with startups. So

8:57Stephen Wise: then I was, okay, well I'll give this a go. Seems interesting. One of my biggest issues was I started more on the design side, but wasn't so comfortable with the development side. So most of the time I'm spending it on all the forums trying to figure out how do I make this thing actually come to reality. Started to learn Webflow. It worked really well for what I wanted, like this kind of understanding of almost learning code at the same time, HTML, CSS, and a little bit of JavaScript through learning Webflow. So it was kind of this like two things

9:26Stephen Wise: at once. You're doing it visually but you're also learning code. Then because Webflow started to grow with startups, it really helped with the freelance business and I kind of grew with Webflow.

9:36Stephen Wise: Over the last four years or so, I started to build that out into a studio. There was a bit of time where I was like, do I want to build a team? Do I not want to build a team? But the main thing for me was this kind of like inflection point where I'm thinking, hey, like over those next few years, what do I want to learn? Do I want to continue to be in this space and focus just on, you know, design, development as the craft? Or do I want to learn a lot of skills around building a business? For me, I wanted to figure out, okay, what are the new challenges here? If you decide to build a business, if you decide to hire people, decide to

10:05Stephen Wise: work on larger projects. So Demandflow very much came out of that, this idea of how do we scale what I've always already thought that I was doing at a very good level and had a very kind of defined offer, and turn that into a larger team so that people can spend, you know, all their week on design, all the week on development, all the week on growth work, rather than me spending my time split between all of them and then also running the business. Took a few years to get kind of fully

10:33Stephen Wise: rolling. You have to make all the mistakes, learn all the lessons. But over, you know, this last kind of two and a half, three years, we've really started to build out the team. There's now, you know, close to 20 of us, and we're starting to really get into our stride, and some of the work we're putting out is really, really high quality and really interesting projects. In terms of Demandflow's differentiators, I always think that

10:55Stephen Wise: for us, we see design and development, like to operate at a world-class standard or to the highest of your ability, that is almost table stakes. I would assume that, like, hey, if you're a design and development agency, it should be a given that you're really, really good at what you do. For us, it's all about this idea of design and development is a means to an end. So to be really good at design and development and to do a really good job, you have to have

11:22Stephen Wise: some form of outcome, right? You have to help someone do something, because that's ultimately what business is. And for us, we always work with marketing teams. So if we can help the marketing team achieve their goals, make them look good to their boss, you know, make the company grow and increase revenue, we're doing our job. So as much as possible, we try to tie everything we do towards those outcomes, towards those objectives. If we do that well, it gives us more license, more freedom to do more interesting, more creative, more technical projects, and then from there it's a case of, hey, every project we take on, we should be learning something and should be growing from that.

11:58Janne Parri: Yeah, yeah, that's a beautiful story and good points with the differentiator. To kind of go, well, not go back, but kind of go back to the freelancing thing. Like I believe there's quite a few like freelancers, me included, listening to this conversation, and that's like the natural progression if you wanna grow, like go from a freelancer to

12:24Janne Parri: a studio or agency. But there is that inflection point, and it's good to actually realize that, okay, if I start to build like the headcount, the job will definitely like change. Like you will not do the same things that you used to do as a freelancer. But nowadays there's this big AI cloud making everybody scared in

12:49Janne Parri: the freelancing and in the agency side. So people might be even like scared taking big leaps right now. So what do you think of this whole AI situation? Should people be scared or should they be excited? Or how do you see the whole thing?

13:09Stephen Wise: Yeah, it's a really interesting thing right now to be like doing a form of work where it's impacted so much by what is going on here. You know, it's usually one of the first spaces to see the impact of any form of AI, is, hey, like design, well, development in particular, but also with design. And the way that I kind of, it took me a little while to try to figure out what my thoughts are, and I'm still very much like evolving as we go, because, you know, every day you're on Twitter, you're on LinkedIn, you see, look

13:37Stephen Wise: at all this new stuff it can do, and look how much better it's getting every single day at all of these different things, and there's a lot of hysteria about people replacing people with jobs, etc. And the way that I kind of view this, at least from my perspective, is this is, for someone of my age, so I'm 26, this is my version of the internet becoming a thing, you know. This is that next transition where everything kind of changes. In that period of time when the internet came in, the people who were the early adopters, who

14:04Stephen Wise: were on the frontier of learning and building different things, you know, when everything kind of came out in the wash, those people were in a very good position, right? Because they understood how to build on the internet and they were very well positioned to build companies in like kind of Web 2.0, like the Airbnb type era. So the way that I view this is, like from my perspective, this feels like that transition. You know, things that weren't possible previously are now possible. I always say this to

14:33Stephen Wise: the team, and there's one thing in particular in Webflow that I always remember. I was working on a client project. For that client project, they wanted to have a quiz before the actual lead form. So they wanted to kind of say, hey, like, you know, you might not know which service you need. So here's the little quiz you want to go through before that.

14:51Stephen Wise: That took me probably like a week and a half to create. Like it was an absolute pain. I had to go through the quiz, create it with the JavaScript logic. I had to go into all the forums to try to figure out how to do this, because I barely knew how to code. I had to use Airtable with the API, pass it to Airtable, do the calculation, pass it back, make sure it worked, and then make sure everything was fine.

15:15Stephen Wise: Right now, if we were to do that, it'd probably take maximum half a day, because you can work with AI, it can obviously build that custom, and you're able to now set these things up way, way quicker. So that just means that we're able to now do work at a level that we worked previously. Within the same budgets, we're able to now just deliver something to a much, much higher level. So the way that I kind of view this is like, hey,

15:36Stephen Wise: we never know what's gonna happen, right? Things are gonna shape and change as we continue to go. But this is like an exciting opportunity. Like for me, I'm more and more excited about work now than I have been in a very long time, because there is now this like huge amount of possibilities that just didn't exist previously. So who knows what's gonna happen? Like you just have no idea. But to build the business, like we're finding the demand is still there. It's not like the demand disappears. I do think that there's probably

16:04Stephen Wise: low end is somewhat disappearing to AI, in the sense of, hey, they might just build it in Lovable instead, or do whatever with vibe coding stuff. But at the same time, that's not necessarily where I'd place myself, and I don't think that's aspirational for a lot of people, to work with companies like that who are nickel and diming everything. For us, the companies still have the same problems. They, you know, might be able to vibe code a website now, but they don't understand the problem deep enough. So what we're being paid for is the

16:32Stephen Wise: strategy, the understanding of how to build a system and how to enable their marketing team, because they'd much rather pay us to do it really well than they would to fumble around themselves, and they're already super, super busy and overworked and being told to use AI for everything else, and like, I just have, you know, absolutely fried. That demand hasn't gone away, right? If anything, it's increased for us. I think high agency people, like people who are really just kind of like, hey, I'm gonna get stuck in and figure it out, it's a very, very good time to be someone like that.

17:01Janne Parri: I always say that AI was supposed to help us and like free us of all the work, but it feels like it puts just more pressure for everybody to do more, to know more, just to be able to like check if the AI has done like good work. But you also mentioned before the call

17:22Janne Parri: like how the value that agencies provide has shifted, in a way that back in the day it was quite easy to say that, okay, we are a Webflow agency, or I am a Webflow freelancer, and nowadays we kind of need to provide

17:39Janne Parri: actual results. Well, obviously that has been the thing in the past too, but it was somewhat easier to say those Webflow things. Do you think that this is a bigger trend, that the tools will fade to the background even more and everybody has to do like results-based stuff?

18:00Stephen Wise: Yeah. So there's this really interesting concept that I saw. I can't remember where it was from, but it's this idea that in the Mad Men era of advertising, so like kind of a decent chunk of time ago, to be really good at something, you had to be, you know, really good, let's say, at advertising, but then you had to have all the resources in place to actually make an impact, right? It was a very long process to actually make any form of impact, required a lot of people in that process. Then we kind of had like, let's say, Web 2.0, which is, hey, do you know how

18:27Stephen Wise: to use the tool? Because if you know how to use the tool, then you can get to the end result, right? So if we're talking about Webflow, for example, you know how to use Webflow, that was a competitive advantage, because other people didn't know how to use the tool. So even though the end goal that people pay someone for is a website that ideally, hopefully improves their business, they're really in a lot of cases just paying someone because they know how to make the thing they want happen in the tool. This is the same for a lot of tools, right? You know, technically like Figma is of

18:57Stephen Wise: this, you look at even like email marketing tools, things like that, is you know how to use the tool. People were just using the tool to get the outcome, though. Like the tool only exists because the outcome is something that people want.

19:07Stephen Wise: Now what seems to happen is we're almost closer to the outcome. So now the idea of actually knowing the tool in particular is less of a moat. It's less of a value add. It is more how do you operate at a really high level. So what I've noticed is, like, people who are maybe creative, when they really adopt this kind of new workflow and new abilities, they're able to create way more, way quicker. And they're not limited by the tools that they're using.

19:34Stephen Wise: This is the same, I think, in most spaces right now. It's not knowing Webflow, it's knowing how to build an effective website, that's really the value there.

19:42Janne Parri: Yeah, yeah. And I believe it was Punit who was on the show a few weeks ago, and he had a good point about all the design tools. Like nowadays, yeah, there are a bunch of new tools, but they all kind of look the same, and if you know how to use Figma, you can figure out a new tool in an hour or something like that. So just get started, basically.

20:09Stephen Wise: Yeah. One tool that I've been playing around with is called Paper. Like Paper is very similar to Figma, but it is much more like HTML, CSS, canvas. And like it's more agent native, like it's easier for an agent to work with in there because it understands that language better, which is quite similar to

20:25Stephen Wise: when an agent is building a website in code, it obviously understands that code better than it does understand how to use a tool like Webflow, for example. So what's really interesting there is fundamentally the tool is very similar. All right, you know, there isn't quite the full flexibility of Figma there yet, but now the ability to like mock up ideas is pretty quick. One like concept of this is you can pull something that's on your website back into Paper, it'll do it pretty much pixel perfect.

20:54Stephen Wise: You can then use Paper as you would use Figma. You can make any changes you want to, you can do all these types of things, and obviously that is using the tool, but then you can send it back into code. So like that workflow of just being able to kind of go between both of them and be able to, you know, add those polishing touches, add that kind of real detail and finesse, like that still exists. It's just the workflows are changing. All the concepts are pretty similar, though. It's not like the concept of building a site or the concept of designing or the concept of craft is changing

21:23Stephen Wise: that much. It's now just how quick you're able to get to the outcome, and if someone is very good at the thing that they do, how much closer they can get to just being able to like execute on that thing themselves.

21:34Janne Parri: Yeah, yeah. Those Papers and Pencils, I believe it's Pen nowadays, the other one, are interesting, but you said the magic word, workflow. I'm curious, have you like already implemented, let's say, Paper or other newer tools in the general workflow in Demandflow, or are you just testing those out every now and then?

21:59Janne Parri: And do you think that, let's say, Figma or some other major tools will be just a thing of the past and replaced with newer tools?

22:08Stephen Wise: Yeah, so in terms of the tools that we're using, like as I mentioned, Paper is something that we are using, but it's not like integral to our workflow right now. We're still designing in Figma, but Paper is like a good place to play around with different ideas, things like that. Some of the generative AI tools are quite interesting, like Flora AI, Midjourney type tools, you know, image generation, things like that. In terms of like our day-to-day, the one that I use the most, specifically for building, is Conductor, mainly because if you're working on like

22:36Stephen Wise: a site agentically, you can pull in like multiple threads and you can just work on multiple things. So for example you could work on a home page, an about page, and then, you know, whatever, like a careers page. You can work on those, use Wispr Flow just to speak to it, say what changes you want to make, and then like align each of them to work side by side, essentially. That's one thing that I've been using a bunch. I think the biggest thing with tools that become obsolete, at least for me and like my current workflow, is anything that doesn't work

23:06Stephen Wise: in an MCP or API fashion, if they're trying to just keep things super siloed.

23:11Stephen Wise: People want to go into other interfaces less and less. So for example, on a lot of our, well, pretty much every project where we focus on the growth side of things, so helping improve SEO, improving AEO, and kind of increasing conversion rate, we'll use PostHog pretty much every time. But I do not look in PostHog. I use the PostHog, you know, MCP to pull in everything and get all that data from there. And like I will look in PostHog when I need to go to like user sessions and watch how people are utilizing the site.

23:39Stephen Wise: But in general, I'm not using PostHog dashboards, because I want it to come into my tools, I want it to come into the context of where I'm doing my work. That is one of the biggest things, I think, is like if a company is very

23:51Stephen Wise: anti integrating with anything else, I find that very difficult to see how that fits in with future workflows, where quite a lot of at least my work right now is being able to think clearly what we want to achieve. Certain areas is, you know, it's jumping in design and doing different tweaks and jumping in the code and doing different tweaks and jumping into Webflow and doing those types of things. But a lot of it is how do you clearly understand what it is you want to achieve, what tools, and like how you can achieve that thing.

24:19Stephen Wise: And then how you get there through iteration, through taste, through refinement, which doesn't work so well if you're in a siloed tool, which maybe is like an Adobe type tool where it's not super integrated with anything.

24:33Janne Parri: Yeah, yeah. I know there's a lot of talk about the Adobe tools and the kind of clones that have been appearing, but I wanted to follow up with, was it Conductor? Because I believe that could solve like my personal problem. Like I'm just like talking to Claude and waiting ages

24:52Janne Parri: for it to do its work. So would that solve that kind of problem, and how does it work, where does it like live and so on?

25:00Stephen Wise: Yeah, so Conductor is like a coding tool. So you can just kind of bring your own key to Conductor. So if you have a Claude subscription, if you have a Codex subscription, like you can bring it into there and you can just select what model you want to use and you just clone a GitHub repo. So for example, you have a website, you can just clone that GitHub repo onto your machine. You can work in the repo, but you can just create separate threads. So inside of that, as I mentioned, maybe you want to work on the home page. Okay, great.

25:29Stephen Wise: You have one agent working on the homepage, you say what you need to say about the changes you need to make, it goes off and does that, and then you open up another, like another pull request, working this next thread on the about page, you go do those things, and then you just continue that as much context as you can hold. Personally, it's about three work streams at once if it's on the same thing. Anything more than that is kind of a little bit too much. But one really interesting tool that I found as well in this like new work,

25:57Stephen Wise: a cool tool called Agentation, which is this ability to like spin it up and comment directly on the canvas. So if you open up on your local machine the website, you can just click on there, add a comment to a specific area, and then feed that context back to the agent that's working on it. So you say, hey, like I want to change these words for this thing, or I want this to look a certain way, or you can even point it to different references. That has also become something that's kind of ingrained in the workflow for

26:26Stephen Wise: us, and a lot of the work we're doing is something that's happening more and more, specifically because we have quite a large scale project where we're migrating like hundreds of websites across from Wix into Astro websites currently. So like a lot of that stuff is pull it across initially and then refine these details, and obviously work with an agent on changing these little pieces.

26:46Janne Parri: Yeah, yeah. And I believe those are exciting things, but those are mainly for the more technical side, and I feel, probably other people feel it too, like coders, developers utilize maybe more AI than on the design side. But I'm curious, you mentioned Paper already, but do you use a lot of different AI tools on the design side?

27:11Stephen Wise: Yeah, so in terms of the kind of AI tools on the design side, I do think that like ultimately we're still in a better place where, like, designing the website in Figma is where you're gonna get the best result, right? Like you kind of need to establish that foundation, that system in the beginning. And a lot of that, we're kind of seeing it be more and more

27:30Stephen Wise: like web design and brand design work in one, because you're creating, okay, what does a visual application look like? What does this brand world look like? So everything is still inside of Figma, like that's the base. Then in terms of obviously generative tools, if we're looking at like trying to create really high quality images, things like that, we'll use like a Midjourney, but it's not every single project that does that. A huge amount of the detailed graphics, detailed pieces that we're doing for product illustrations, that's all done manually inside of Figma.

27:58Stephen Wise: Not like a generative thing, and it's not something that we're finding you can create, because when you first create it, you typically need to be the person who actually creates it. The AI is very good at just repeating something or going right into the middle of everything, which is why generative AI is so bad at writing, right? Because it just takes all writing, and it's like, I'll just give you right in the middle of everything ever, and it's just the worst thing you've ever seen. So that's definitely one thing. I think in Figma, the agents

28:24Stephen Wise: are pretty interesting. If you have to make large scale changes across loads of pages, that is really nice, not to have to go into each page and change some copy that the client has decided, instead of using this word, we now use this word, and you have a hundred pages. It's nice that it can do that itself. Some motion design concepts are pretty interesting to just do in the latest version of Claude, with Opus 5.5. And then in terms of the other side of things, it's really like, you know, using Paper,

28:51Stephen Wise: seeing what you can do, seeing what we can't do. But the biggest changer we do have is prototyping. So it's like, okay, can we take something and illustrate a concept in code? So for example, let's say a designer has an idea for an animation on the hero section, they can now like generate just a rough version of that by connecting the Figma MCP, you know, kind of dictating what they imagine the experience to be like. And now the development team understand the concept

29:19Stephen Wise: almost in production-ready code, which is a thing that just didn't exist before. So what we're seeing is design, they're getting slightly more technical, in the sense that like designers are able to prototype and work with Claude and other AI models to build things in code, and also kind of like improve their skill set, I guess, or the capabilities they have. But then on the dev side, on the development side, there's kind of two ways

29:47Stephen Wise: the developers seem to be going for us, which is, do you become more of a design engineer, someone who is moving more into the bridge between design and development? Or are you more like a front-end engineer who's becoming more systems structure, using Cloudflare, using these setups and really building things out from an infrastructure layer?

30:07Janne Parri: Yeah, yeah. And there was that Flux Academy video where Ran basically said that, like, the development part of the usual website lifecycle, if we go from design, development to maintenance, like the development part is shrinking, and maybe freelancers or developers should branch into more design strategy side or the marketing maintenance side.

30:32Janne Parri: And I feel that there is some truth to that. So do you think that maybe in the future every role will mix with everything? Will we have traditional developers and designers as we now do, or what will the roles be?

30:49Stephen Wise: Yeah, that's a really interesting question. I think it's definitely easier now to do more than you could previously. The idea of being able to be a designer who can now actually, just through sheer brute force and just telling Claude to do the right thing enough times, you can build something. You can build a website, right? You can design that and you can see it the whole way through in custom stack, right? You can do that. It's not ridiculously difficult. It does take a lot of learning, et cetera, but, you know, it's not like it previously had to be, where you had to, you know, get the textbook out, actually learn the coding language, and

31:20Stephen Wise: it would take forever. That is a really interesting workflow. I think while things are merging more, I don't know whether there'll ever really be a scenario where design and dev become the same thing, because ultimately specialization is important. People need to be good at certain skills, and being really good at something is better than being average at everything. So even though we could kind of like expand that

31:48Stephen Wise: breadth of what we can do, I don't necessarily see it fully going away. You mentioned the video from Ran. I agree that development cycles are shrinking. That's definitely the thing that we're seeing in terms of the speed side of things. But at the same time, there's still things that take time. Like polish on a website, even using an agentic workflow, it's not quick. It's not like you're instantly just getting there straight away. There's still a decent amount of work to be done. And I think that like

32:15Stephen Wise: this idea of where developers go, I do think that is the case of, do you move more into design strategy, directing what is built? Do you move more into potentially marketing? Like that's if you're a freelancer, assisting marketing teams and really being their enabler, the person who is more technical than them, who can understand their problems and actually build things for them. Or even do you move even further into being slightly more technical in this

32:44Stephen Wise: space for marketers, so that when people have very complex setups, they have very complex challenges they're trying to solve, you're able to do that because you understand how to utilize development technology together to be able to do more complex things. I do think that the middle of the road, just doing development inside Webflow, that kind of thing, and, you know, same for a lot of other platforms, I think it does need to evolve in some way.

33:08Janne Parri: Yeah, if someone was just a quote unquote Webflow developer, that was kind of a shallow niche or profession. And AI will not decrease the amount of knowledge or understanding that you should have. It will probably do a lot of work for you, but the knowledge and understanding will still be beneficial.

33:32Stephen Wise: Yeah, I think a lot of people overestimate that. People don't know what they don't know. So like if you're a developer, you know that, for example, there's accessibility standards for the web, or you know that, you know, markup schema exists and you can do all of these different things. And a marketer who has no idea about this stuff does not know what they don't know. So they don't know how to orchestrate AI to do this stuff. So even

34:00Stephen Wise: it can be really, really good. I do think there's this idea of being uniquely positioned to solve a very specific problem. That still exists, right? It's not as if it's just solving every problem. I'm sure everyone has had that experience with AI where as soon as it goes off, having like context, or maybe it misinterprets something that you have said, you have given it, and it just starts to make stuff up. And by the end of it, whenever it's working, it is so far away from what you actually wanted. You're just like, why did this even happen? Like this was just a complete waste of resources. It's just made stuff up and I can't

34:35Stephen Wise: use it. That's the exact same thing in all areas of work, right? It's like if it's not shaped in the right way, if you don't have the right experts in the right seats to do these different things, it doesn't matter, right? It's gonna just create a load of rubbish stuff. So I think that idea of actually having taste, actually being able to offer a strategic input and solve a very specific problem, like that's always gonna exist. It just continues to evolve.

34:58Janne Parri: Yeah, yeah, totally. I'm curious, since you have like the freelancer background, do you think that this whole situation, like all the new tools, all the AIs, new roles and everything that we have talked about this far, do you think that this benefits more the freelancers or the agencies or some type of agencies? Any ideas?

35:21Stephen Wise: Yeah, personally, I find that my personal ability to do things, specifically the things I didn't want to do previously, like paperwork type stuff, has increased a lot. Like I now have this ability to, you know, do things at a speed that I just wasn't able to do previously, and I may have procrastinated on for a very long time. That

35:41Stephen Wise: idea of, if I was a freelancer right now, I would really, really like that, because that was some of the stuff that, like, you know, just kinda sits there for a long time. And like, you know, it can even, if it gives you a first draft of an email based on your tone of voice, if you start to kind of train it so that the output is pretty good,

35:58Stephen Wise: then you don't have the blank screen problem of having to write the email from scratch. You can now edit that thing. That is something that would speed up your ability to do stuff in general. So personally, if I was a freelancer now, I would very much like this time, just because it's this ability to shift and evolve and you're so nimble. I do think that very big agencies, you know, like the traditional hundreds of people, thousands of people agencies, might find it quite difficult

36:27Stephen Wise: to shift. Even some of the companies that we've worked with, you see that AI adoption is at very different levels right now. Some people see things on Twitter every day and they're like, so far behind. But like you're using Claude, you have, you know, subscriptions to all these things and you're using it, and you still think you're really far behind. But then there's huge, huge companies who still are at the stage where they're using Microsoft Copilot and they're not allowed to use Claude and they're not allowed to use any good model, basically. So

36:54Stephen Wise: those types of companies, where institutionally it's very difficult to make a change, I feel like those type of companies, and if we say the agency space, obviously there are bigger traditional agencies that are like this, will struggle. Being smaller is probably in a better place, just because you have the ability to shift, evolve, and kind of change with the market itself.

37:17Janne Parri: Yeah, yeah, that's totally true. If you are flexible, nimble, you can try out and you can implement stuff faster. But then again, if you're just a lone freelancer and you're fully booked, obviously that's a good thing, but you might not have the time to try out everything. So it would be great to have like friends, coworkers

37:39Janne Parri: who tell their experiences and give you the best tools to try out first. Wouldn't you think so?

37:45Stephen Wise: Yeah, one hundred percent. I do think that there is a nice sweet spot of having that network of people who are also interested in all of this stuff, which is really interesting, to see the different people kind of coming at everything from different angles, using different tools and kind of keeping your finger on the pulse. Like for me, that is one of the more interesting things. Like, you know, over this weekend coming up, I'm probably gonna spend a decent amount of time playing around with some different tools, because that's when I get time to actually work on these things. And I think that

38:16Stephen Wise: to me the most important stuff is just being able to understand, okay, are there new ways to do stuff? Can I do things that previously took me a really long time quicker? Areas where we can actually just improve the quality of work that we output. So I do think that even just keeping up to date with what is happening, having people you talk to in kind of similar realms, I think is really useful, to understand their experience and what's changing for them, because no one knows what's happening, right? Everything's just evolving. No one has a full understanding of where this will all go.

38:47Janne Parri: If you count out the influencers on social media, they always know everything for sure, and you must do this and that and so on. So yeah, I feel like sometimes when I'm trying out a new tool, I'm just fact checking, like, is this actually possible the way someone did it on the internet? But yeah, since

39:08Janne Parri: you mentioned finger on the pulse, one current thing is obviously the already mentioned Webflow Source, or Source by Webflow. It has been a topic in almost every episode in the last month, so I kind of have to get your thoughts. And do I remember correctly that you had a quote on the announcement trailer, or the article?

39:33Stephen Wise: Yes, yeah. That was a bit of a surprise at Webflow Conf. I kinda forgot I actually submitted the quote, but yeah, it's a bit of a surprise. But Webflow Source, it's a really interesting problem they're going after, and I'm glad to see it, because, as we've already alluded to, a lot of this agentic development, while the MCP and Webflow works and it does a lot of things, it doesn't necessarily do the full loop where you own the code itself. I think the problem they're trying to solve, like

40:02Stephen Wise: that is very much a gap that exists right now. And it's a gap that, as someone building an agency and working with a lot of scale-up and enterprise marketing teams, that doesn't exist in its current state, right? This ability to hand off to those teams, there isn't something where I can reliably say, hey, go over here, use this tool, and you can edit things visually, and here's a nice marketer-friendly interface to do that thing.

40:26Stephen Wise: Obviously, there's a lot of things to be figured out. Obviously there's user adoption. How do you get people, if we're gonna use these tools, we need to sell it to our clients. So that means that things need to align with us, the builders, to make sure that we want to sell it to the clients and it works for everyone in a good way. But I think it's a really interesting problem now, is that there is this,

40:47Stephen Wise: it's almost like the other side of things. As we were talking about, a lot of things are going to the chat interface. There's still certain scenarios where it's good to use a UI. It's very good to just go in there and tweak things as you actually want to tweak them, because it's very difficult to explain exactly what a tweak is. So interested to see where it shakes out over the next kind of six to twelve months.

41:06Janne Parri: Yeah, that is the gap that we talked about earlier. Like that is kind of tough, especially if you work with big companies and you sell like premium websites or premium design. You kind of want to have some sort of editing experience that is polished. And right now in that gap there might not be

41:27Janne Parri: such a thing. I obviously haven't tried everything out. You mentioned Ship Studio. That is one thing that is already out there, but I'm not sure how easy it is for non-technical people to use. But there is definitely an opportunity. Yeah, go on.

41:41Stephen Wise: Yeah. All of these tools, right now in the current stage, they're not that difficult for non-technical people if you're technical. If you're like, it's not that difficult to set this up on your localhost and do all these things. There really isn't anything right now that I would say is matching what the experience of the current Webflow marketer slash build mode is. So that is

42:09Stephen Wise: why I think it's a really interesting challenge to try to solve, because we also don't know how people are gonna evolve in this next period of time, right? Everyone thinks that because now doing technical things is easier, everyone's gonna go do those things. But we don't know that's gonna happen. And it's really interesting to predict what this next stage is, and what people actually even want from this evolution of a marketing site. Like we're even seeing the way that we're building marketing

42:38Stephen Wise: sites is changing, right? Because now we have AI search, which is like, AI agents are crawling the website and bringing answers back to people. So even the structure of what a marketing site does is changing. And with all the companies changing, because a lot of them are integrating AI even into their products, like everything is just very transient. So I think it's really interesting to figure out, okay, what are the challenges that are gonna be, what becomes the new normal? What becomes, in 18 months' time, what is everyone doing?

43:08Janne Parri: Totally, totally. I'm having high hopes for Webflow, because obviously Webflow is a big company, they might be able to pull this off, and I'm imagining something like super easy. I have one particular client that has a marketing manager who has a lot on her plate, and she is always asking me about all sorts of minor

43:29Janne Parri: technical stuff, and I always imagine her when I'm thinking how easy this thing should be. I want it to be like super easy from the start. If we do not know where something is, we can just ask, or the user interface would be super intuitive. And if it's easier to do with

43:51Janne Parri: prompting, then there would be the prompting thing. If it's easier to deal with a UI, the UI would just pop in and be as good as Webflow's UI has been on Webflow classic, if you may. So yeah, that's a big challenge, obviously.

44:08Stephen Wise: Right. It's always that way, isn't it? I think everyone says about when you build a product, it's like death by a thousand cuts, because building a product is very, very difficult. There's so many things that you have to think about, so many second order consequences of making any form of decision. If they can execute on it to the level that traditional Webflow is executed on, that would be perfect. Right now, if that existed in the Source ecosystem, that is exactly what we would use. But it just doesn't exist right now. So there is that gap. There is that, like, what comes here? How do we make sure we solve this problem?

44:44Janne Parri: Yeah, yeah. There's so many questions on Source. It has been a topic for many episodes, but there are no clear things that anyone can talk about. I like this topic still and I'm kind of wanting to speculate a bit more. Do you think that it will become a tool that agencies use and love?

45:09Janne Parri: Or will it be a tool for everybody, from freelancers to clients themselves, like in-house teams and so on?

45:18Stephen Wise: Yeah, that's a really interesting question, because obviously you've been in the Webflow space for a while. You understand the sentiment towards, like, freelancers in general feeling a little bit standoffish of Webflow as a company, right? Like when I first got into this space, nearly seven years ago, it was all about freelancers, all about the initial creators and all of those things. And, you know, you had your videos from McGuire, who was a cult legend, basically, created a huge amount of that kind of culture. You basically gave designers the ability to do things

45:47Stephen Wise: they couldn't do previously. And then over time, obviously, the bigger the company gets, the more they focus on who their main customer is, which is a marketing team. I probably don't see that changing, to an extent. You know, the marketing teams will be the customer, because ultimately those are the people who need the website to do all of these things moving forward. In terms of the adoption from, let's say, more creative freelancers who are really into the idea of creating website experiences, and maybe some of the people who in this previous

46:17Stephen Wise: era went to Framer, I'm not sure where they go. I really have no idea, in this next iteration. Does, you know, Webflow Source cater to these people? Does it not cater to these people? Is it another platform? Does Framer evolve into something that fits this kind of scenario? It is a really open question. I think if your work aligns with what a marketing team needs, like if you do work with, let's say, even just marketing teams of two people or more,

46:45Stephen Wise: this is probably going to be a fit, right? Because that is who they know they're trying to cater towards, marketers. So I think that side of things definitely fits. I'm really intrigued to see how pricing shakes out. That is obviously gonna be a big thing, right? I think everyone's holding their breath on what comes out there,

47:02Stephen Wise: because, you know, there's been some interesting pricing in previous iterations of Webflow, and it's pretty confusing, right? So I think that will be a really interesting one, because even just, like, how do you build, who gets billed

47:16Stephen Wise: for the build, like actual building of the website, in terms of if it's AI credit usage, all of these different things. This is all gonna be really interesting to figure out, because again, this is very similar to what we were talking about before, but no one knows what this next iteration of work looks like. So does a huge company really want to pay for AI credits from another company, or do they just want to use their API key that they have for Claude and their amount, and kind of utilize their own usage? Because

47:44Stephen Wise: we're seeing this happen like loads. Every company is figuring out their AI policy on the fly. So tooling that fits with these things kind of has to play with whatever's happening at the bigger companies. But yeah, I think

47:58Stephen Wise: when you're at Webflow Conf and everything's about marketers, I'm like, okay, I understand it. It would be great if there was more of a track for us builders, to be like, hey, it is for marketers, but if the builders don't exist, the marketers don't get the thing that they want built. And, you know, even in the current iteration of Webflow, right? We've inherited a lot of Webflow sites that are not built very well, and no one's getting any value from it, because it is built so poorly that you may as well be on

48:27Stephen Wise: whatever old CMS slash website platform, because, you know, you're not getting any of the value of what Webflow is. So everything, from my perspective, is the only way a tool works

48:41Stephen Wise: well, and you've seen this happen more recently with Framer, obviously, getting a lot of designers there, is the builders have to come first, because the builders are the people who build the solution for the people who pay the money to do the thing. So it's like, if you can sell an amazing solution to a marketer, but if they can't set it up themselves 100% to be optimized how they need it to be, you need someone else to do that. And if you don't have that, then it's gonna be harder for them to find the value in the platform, in the product.

49:11Janne Parri: Totally, totally. There was so much there, but with the Webflow Conf builder thing, like this year there was the builder keynote, and that was kind of like a weird thing. Okay, builders get their own thing, so they are appreciated, but is that kind of like a bait and switch, or some sort of sleight of hand, that okay, the

49:34Janne Parri: big thing are now marketers? Obviously it has been for a couple of years. I'm thinking, what if the order would have been just different, like the builder keynote would have been the first and then the marketing stuff? So that might have fixed a lot of sentiments. But yeah, maybe that's

49:56Janne Parri: a thing of the past. And we have lived through the community golden days, and we're kind of reminiscent of those, obviously.

50:06Stephen Wise: Yeah, there's definitely, I think everyone who's been in this space for a while has some form of story, or some form of, that they reminisce a bit on, like, here's McGuire's Webflow videos, and they're like a great experience to learn this tool. This tool gave you the ability to do things you couldn't do previously, and it's like everything in life, right? You're trying to recreate good experiences, and

50:30Stephen Wise: times shift as you go on, and every company has to evolve, and this is happening everywhere, right? And I think it is a really interesting dynamic in the Webflow space right now. Personally, I thought the developer keynote had some really good things they dropped in there, and I'm like, you should have used this material in the main thing, because I'd have been more excited in the main one if I heard about even just components in the CMS and everything. But yeah, it's an interesting, it's difficult, isn't it? Because it's almost like they have a similar dynamic to a marketplace, where it's like, hey, if you have a marketplace, you have two sides of your audience. So how do

51:03Stephen Wise: you cater to both sides of your audience? And usually a marketplace is going to cater towards the ones who pay the money. Like Airbnb, to an extent, is going to put more external marketing towards the people who pay for it to stay at an Airbnb than they are towards the hosts of the Airbnb, because they need the people to go stay at the Airbnb so the thing exists. And in a way they have a similar dynamic, right? The marketers have to be there. They're the ones who ultimately control the budgets that pay the bills, but without the other side of the marketplace, the whole economy, the whole thing doesn't exist.

51:37Janne Parri: Yeah, yeah, totally. And I just had Kyle Lambert here. He works at Webflow and has been designing Source, and we talked about the same topic, and he had a good point, that there's a limited time they have to tell the story, and obviously the focus has been shifting to bigger clients and marketers. But especially with Source and

52:00Janne Parri: the AI tools, like developers, builders have been utilizing those for quite some time, and many feel like the marketers have not had that much resources or stuff, so it was kind of focused because of that too. So yeah, and

52:18Janne Parri: there was something I wanted to circle back to. Yeah, all the Source pricing, the APIs or MCPs or like bring your own agent stuff. I believe that Webflow has not decided on the pricing yet. That's the thing that you mentioned. Like, are they gonna just be token resellers, or will it be seat-based? Can you bring your own agent? I kind of feel that people will get disappointed if there is not a possibility to bring your own agent, because this is a step towards more open type of building with all the code bases and such, and then if you lock the agents,

52:57Janne Parri: you'll be forced to use their own agent. So that feels kind of a turn off.

53:05Stephen Wise: I think the way I view this, and this would be my ideal scenario, I have no idea what is gonna happen in terms of pricing, but I understand the idea of obviously a platform providing tokens and that kind of thing. It does make sense when you're thinking about agents running on an existing website and marketers interacting with it. From my perspective, though, if you're a dev working on a client's website,

53:26Stephen Wise: when you're building new things, you wanna use a frontier model. You wanna use the best model possible to build larger things. Frontier models on API pricing is very expensive. It is not that expensive on bring your own key, because they obviously subsidize them so much right now. It's probably another problem when Claude and ChatGPT are trying to make more money and not subsidizing these subscriptions so much, but maybe at that point the compute is lower and these costs may go down. From my perspective, the idea of being able to,

53:58Stephen Wise: let's say we go in and do a new build for a client. If the entire build process has to be on API pricing, like credit pricing, I'm gonna find that very difficult to justify doing that. This is from my perspective, like that's a very difficult thing, as, you know, well, anyone, freelancers can find it very difficult to do that because it's gonna be so expensive. Agencies again gonna see the same thing. So the way that I would ideally like something to be set up is this idea of, like, how can we develop

54:26Stephen Wise: locally, using Source or whatever the solution is, to be able to use our own keys for certain things, you know, making bigger changes, and then have, let's say, a Source type environment be the maintenance type thing. But if you're making big changes, you don't use the credits. Because the credit, that's the thing for me, and I've used a lot of tools. There was an AI personal assistant tool called Vellum a little while back, similar to like Grokbot and some of these others, and it was a credit-based system. You get set up and you pay, you know, $50 or whatever.

54:58Stephen Wise: And I spent the first fifty dollars just setting up the system. So by the time the system was set up, I never used it, because I was like, I don't want to pay again. And I could see the same situation happening in so many AI products. And this would be no different. Is that, like,

55:13Stephen Wise: if you use all your credits straight away, before you've ever tasted the experience of how good the product can be, I'm never gonna keep using the product. So that is a really interesting dynamic now, with API pricing, using credits and all of this stuff, because I think there's a fatigue that would happen very quickly, where if you just start burning money very quickly, it's not a great situation.

55:39Janne Parri: Totally, totally. And obviously people could just develop the project locally or in some other environment and then import that to Source, so they would kind of bypass the building phase and then start maintaining the site. So Webflow probably doesn't want that. They wanna provide also the build from scratch experience too.

56:05Stephen Wise: Especially with this idea that you may have a big site already there, but you might need to make five new pages for a new launch. That is something I'm really intrigued to see, about how does that work? What is the workflow there? What is the dynamics of that?

56:19Janne Parri: So many open questions still. I'm waiting to get my hands on the research preview. But before we stop, I still wanna kind of visit the agency topic. Like we have been talking about the tools and AI, but there's all sorts of things about the maintenance of an

56:38Janne Parri: agency or studio. So has AI changed anything else about day-to-day work, like maybe hiring or management or something like that, or your day-to-day work life?

56:52Stephen Wise: Yeah, I feel like it's one of those things, isn't it, where it feels like everyone has some form of AI fatigue, because everything work related now has some AI built into it and you're getting marketed by it. But it is really one of those things where everything we do now has some form of knock-on effect, because it has changed the way the work is done. So, as you mentioned, with hiring, it's like, okay, well, now we're looking for a skill set that is different to what the previous skill set was. If previously

57:21Stephen Wise: we were very much looking for, let's say, Webflow developers who knew Webflow, now that has evolved for the new kind of setup here. We want it to be slightly more people who understand how to utilize these AI tools to be able to build, you know, better solutions, build quicker and do some of these other things. So that is one element that has definitely changed. But the other thing is also just building tooling. So we have an operations lead, and I was talking to him the other day, and we were talking about this idea that previously

57:51Stephen Wise: we'd maybe have to use a tool like ClickUp, and we do use ClickUp internally for task management, but for any workflow we want to create, we have to mold the tool around the ideal workflow we want. But then the tool is never perfect, right? You always find a workaround and you always have to hack something together to make it work. But now we have been building our own tooling, like our own operations platform and a few other things. And what we can do now is we can say, hey, for the work that we do, for what it is that we're building,

58:20Stephen Wise: how do we build the best solution for that thing? What's the best workflow that works for Demandflow? Then we go from first principles, we figure that out, and then we build it. We just build it in code, we build a tool for the thing, and we're able to now build solutions to this. One thing that I see happen more and more, with this idea of, hey, everything is shifting, what is a website as a system? Is how do you actually build your own website almost operating system, when, you know, someone's gonna connect an agent to a website and they're gonna start saying, hey,

58:49Stephen Wise: I want to build a new landing page for this thing. There's a million ways it can approach that. But if you already have all your tokens and you have all your components and visual styles and everything set up in there, and you have a set of rules for how the agent should interact with this stuff, what it's allowed to do, what it's not allowed to do, when the marketer connects to that thing, you now are able to build a lot more here. You're able to essentially use that and give them guardrails and build things that are on brand.

59:15Stephen Wise: This is really, I feel like it's infiltrating everything that we do. Now we can just almost think of everything from first principles of, okay, what's the ideal workflow? What's the ideal system? And how do we think about what we do to cater towards that thing? And that kind of fits everywhere, right? Like mini design tools to help us, mini tools to help our project leads, tools to help the developers do certain things, tools to help us manage the company. All of these things, they wouldn't have existed previously. We would have had to be in a much bigger agency to be able to commission those types of projects. Whereas now it's a much easier thing to do.

59:53Janne Parri: Yeah, yeah, that's super fascinating. I'm thinking, one thing I have been pondering about has been design systems, and especially design systems that LLMs can read easily, as you mentioned, those tokens and components and so on. And I'm wondering, will those be a thing in the future, or maybe they already are, but should those be built in GitHub or wherever, just to be able to easily

1:00:26Janne Parri: switch out the agent that is reading those and so on? So are you talking about that sort of design systems and specialized little tools that utilize those, or something different?

1:00:38Stephen Wise: Yeah, I think there's kinda multiple levels to this. And really, I think when you do any form of work like this, where you have design and you also have development, right, they're two separate disciplines that kind of come together, and they overlap in a lot of ways. One of the things that I think is really important, especially when you're growing an agency, starting to hit a bit more scale, you're getting more people here. But also it's important in freelancing, because you want consistency between the projects that you deliver, you know, want to make sure you're hitting the same standards and things, is that the system that you create in dev

1:01:07Stephen Wise: to an extent should ideally map what you've created in Figma in design. So that kind of continuity between the two of them, I think, is really important. Of like, okay, if you design something in a certain way, how does that actually enhance your development process, so there is a relationship between them and you can continue to work together with that? So if your design system inside of Figma matches what you did in dev, when in future you work further with that client, you just continue using the design system in Figma, design the thing, and it just naturally evolves in dev, because you've already got the system set up. So that's definitely one thing, is like,

1:01:41Stephen Wise: what is it like to create your own mini world when someone works with you? What is a website that you create? What are the things that you lean into? For us, we kind of call it like a Demandflow OS. It's like, what is the operating system, why you're choosing to work with us over choosing to work with other people, because we give you X, Y, and Z? Like these are the things, we build the world of the system that enables them, and sell to those people. That is very much a thing that we're trying to lean into as much as possible. Something that's

1:02:09Stephen Wise: super early for us is even just getting the entire team comfortable using GitHub, even for some knowledge base work. Just like, okay, what is it like to explore even having our docs inside of GitHub, or connected to GitHub? Because a lot of people would maybe pull it from an LLM and maybe can ask a question to it. And that's a nice and easy way to keep a versioned update, rather than keeping it in a really heavy tool like ClickUp's own docs, which are a little bit meh. So it's really just this evolution of

1:02:38Stephen Wise: system thinking, of every system that you have for what it is that you offer, almost like another level. One thing that I find really interesting from a service provider perspective, and I think if I was a freelancer right now I'd think a lot about this, is

1:02:51Stephen Wise: previously what we offered was a service. It wasn't so much software, but now you can offer your own software to achieve certain things. So if you're doing things consistently, you'll see people build design tools, dev tools, different tools to help with the process. Something that I saw someone create, really interesting, was this brand design system. So they're a brand designer, and they ask the questionnaire to the client when they get started. Previously, it was just, here is a bunch of questions, please fill it in in this doc.

1:03:19Stephen Wise: But instead, what they did is they turned it into a mini web app. The web app was an interactive experience, and the client went through the journey to give them the information. And they had it looped in so that if the client didn't give them enough context in something, it would ask follow-up questions, like the next round. So it's almost like a Typeform type thing. And that is the type of thing where I'm like, okay, well, if I work with you for a brand experience now, like, you know, work with that person, they're giving me something that other people just do not have, and they're tailoring this, they're getting all the information,

1:03:49Stephen Wise: because they have decided, hey, working with me is a different experience than working with everyone else. And while previously it was maybe more difficult to understand differentiation, if you're able to build your own offer and really uniquely tailor what your skill set is like that, that then just makes the thing that you're offering way, way more valuable, and a much more pleasant experience to work together.

1:04:11Janne Parri: Totally, totally. That's exciting. I might try to build something like that for myself. And one other example that I just saw, it was maybe a more granular, smaller thing, but a designer did this image generator for the client, a client that had these abstract images, you can imagine, like gradients,

1:04:35Janne Parri: but it was a client-specific generator that always generated the images in the style, so marketers could use that, and everybody else. So that might be the next standard, that just a PDF brand guide doesn't cut it anymore. You have to do something more.

1:04:56Stephen Wise: Yeah, 100%. That's something that we're exploring more and more, of, okay, how can we add these other things to the work that we're doing that helps them achieve the outcome that they want to achieve? Because I'm sure every designer has seen some form of reality of this, of like, you pour your heart and soul into a project, and you're like, this is the best brand design, the website looks so good, and then you give it to them and they're making social media posts with Canva, and you're like, that does not look like it should look. And you're just like, it's quite painful to experience.

1:05:25Stephen Wise: Even this ability now, like you mentioned, if they're creating backgrounds and different brand assets and things, just giving people the tooling to be able to always effortlessly continue on the brand is a really valuable thing. And I think a lot of the design work and a lot of the work that we're doing kind of goes into that, right? That brand design tool only exists because the person created something unique in the beginning. They created a language, a way of doing that, like a visual style, then they

1:05:53Stephen Wise: create a tool around it, trained it on that visual style, and then gave that one-to-one to the client so they could reproduce that thing. If you were just like, hey Claude, from the beginning, just create me a brand and then just build this tool here, and you just gave it no other input, it would look like every other Claude-generated thing, with the warm background tone and basically Claude's own colours, and everything just looks very kind of vibe cody. It's the exact same thing. It's just a shift of work. It's a shift of where the value is that we actually add to people. And I think that's a really interesting place right now.

1:06:25Janne Parri: Totally. Okay, we kind of have to land this plane at some point. But before that, what about Demandflow? What will the future bring for Demandflow? Do you have anything new coming up, or any predictions, any focuses for the future? Anything you can tell?

1:06:44Stephen Wise: Yeah, so I think really, as the company grows, one of the things that we're trying to do is really deepen our expertise in different areas. So it's this idea that we want to own the website experience when we work with our clients, but we're trying to offer an end-to-end service, right? So what we're trying to do here is understand how do we control brand design, web design, all the way through, ideally, copywriting, sitemap structure, user journeys, high fidelity design, into development, building out the system, and then hopefully growing the website with the client.

1:07:14Stephen Wise: Our whole thing here is trying to focus on how do we work with scale-ups, enterprise marketing teams, and help them essentially accelerate their work, make their lives a lot easier and hopefully more fun. And as we try to grow the company, the main thing right now is we're scaling, you know, we're hiring as a company, we're currently building out our own website, which is actually on Astro. We were talking about this a little bit before. But it's like we're building out our own new version of the Demandflow site and expanding that quite a lot as we

1:07:42Stephen Wise: start to evolve into this next phase of an agency. But the whole idea is, how does the bar continue to get higher? Like when you add someone to the team, they should challenge other people on the team, and they should add skill sets that other people don't have. And that's really the biggest focus here, is how do we just continue to be at the forefront of wherever these things go, play around with these tools as they're there, and evolve as the market evolves and as everything here just kind of develops. So I think the biggest thing for us is continue to grow, but

1:08:11Stephen Wise: really just hyper specific on the thing that we do, and build a world around that, as we mentioned, tooling and different tools for clients and things like that. Just evolving with what is now part service, part software that you offer when you work with clients in this world.

1:08:28Janne Parri: Yeah, yeah. And do you think there will be smoother sailing at some point, or will it be just as crazy as the two or three years since ChatGPT came?

1:08:39Stephen Wise: That's a great question. I honestly have absolutely no idea. And I think that's the fun part, right? Is that every day is a new challenge. You know, everything that we all do, freelancers, running a business, even in design and development, it's never boring, right? You're never solving the exact same problem every single time. The way I kind of see it is just like, hey, everything we're doing is just a really good learning experience to build this skill set that makes you more and more valuable as you move forward. You know, if this is always a thing,

1:09:09Stephen Wise: I would say is, if, you know, Webflow disappeared tomorrow, it doesn't mean that what we do isn't valuable still, but it has to shift, right? It just has to shift into a different thing. We learnt so much about what works, what doesn't work, how do you kind of develop taste, how do you build these things. So I think really it's just trying to be more and more resilient as things seem to get kind of wilder and wilder, and just understanding where it goes and then evolving with it.

1:09:35Janne Parri: Yeah, totally. Beautiful, beautiful, Stephen. Okay, there's one weird little thing at the end. But before that, if people wanna check you out, follow you, check out Demandflow, where should they go?

1:09:47Stephen Wise: Yes, so Demandflow is demandflow.co. We can't quite get the .com, someone's holding that to ransom right now. And then, yeah, social media, the exact same thing. LinkedIn, Stephen Wise, Stephen with a P H, W I S E. Don't really use Twitter that much, but you can follow me if you want. We also have a YouTube channel. Created a few different YouTube videos, but nothing particularly recently. So if you want to go look at some really cringy videos of me talking about Webflow, just type in Demandflow.

1:10:17Janne Parri: Yeah, yeah. And I will obviously put the links in the description. Okay, we're at the deep end right now. I always have to ask, do you happen to have any idea what the last section is about? Beautiful, beautiful. I prefer it this way, because then it will be a surprise. It's a thing that I do in every episode, right at the end. And it is this little section called the Finnish Line. And what follows is that I will give you a line of Finnish, Finnish as the language of Finland, and I will translate the line. It will be a metaphor or a saying or something like that. I will translate that

1:10:59Janne Parri: word for word to English, and your job is to tell me what it actually means. Is that part clear?

1:11:09Stephen Wise: That's clear.

1:11:10Janne Parri: Yeah, beautiful. I will give you your line, or saying. Sometimes these have some sort of logic, sometimes no logic at all, as sayings sometimes don't have. But your saying, actually, in Finnish this is just one word, and in Finnish it is jänishousu, and in English word for word it is rabbit pants. So what does it mean if I call you a rabbit pants?

1:11:44Stephen Wise: I'm gonna assume it means I'm really quick. I'm really fast. I'm a good sprinter.

1:11:50Janne Parri: Yeah, that's actually a good guess, because there's a clear logic, like rabbits are,

1:11:55Stephen Wise: You said good guess, which makes me think it's nowhere near what it actually is.

1:12:00Janne Parri: Yeah, it's nowhere near. Do you wanna second guess, or do I put you out of your misery?

1:12:05Stephen Wise: Yeah, put me out of my misery. I have no idea what that could be.

1:12:08Janne Parri: Okay, so basically rabbit pants means a person who is a coward, or afraid of doing something crazy.

1:12:16Stephen Wise: I guess, yeah, you could run away from things. It's kinda like, I could've probably drawn that conclusion.

1:12:21Janne Parri: Yeah, yeah. But I don't know, it's just weird. The first thing that comes to my mind is that someone has a rabbit in their pants, so how does that make you a coward? Yeah. Okay, this is just a funny little thing I do for every episode, just to root for our little country up north. But Stephen, this was great. I would have loved to talk to you another hour, and maybe a third after that. This was great. Thanks for being a guest, and for all the listeners and watchers, we'll see you in the next one. Bye.