Webflow, Astro and WhatFinsweet Is Doing Now

0:00Janne Parri: Okay Alex, so is Finsweet pivoting from a clear Webflow focus to a broader tool set? Or service offering?

0:09Alex Iglesias: Direct question. I love opening the podcast with this. Yes. I think that exactly every single agency nowadays is gonna tell you that at least they're looking into it, knowing how the market is changing. So the short answer is yes. Finsweet is not only looking into Webflow, which is still a big part of the agency, but also we are trying to adapt to this new market, if that's the is his way to put it.

0:34Janne Parri: Yeah, yeah. That's like on everyone's mind. But I know that Finsweet has all of those products and you have the agency business. So I'm curious what type of things are you looking at in this AI age?

0:48Alex Iglesias: Well, the first thing that obviously 'cause it's not only us looking in out to things, it's also receiving feedback from clients. 'Cause basically everything that we do in the agency is based on whatever the customers need from us. We can be creative and we can offer things, but the customers have the ultimate decision. So Recently we've got more customers being interested in either just being able to work more with AI, which Webflow is covering a lot of it, but also just not wanting to have a vendor lock-in now that they've got the full power of just building things using AI and not requiring a no-code tool like in the Webflow traditional sense. So we are trying to adapt to what our customers are asking and what we see that the trends are.

1:35Alex Iglesias: In the market. It just, there's a lot of tools. And obviously I think that everyone's talking about Astro. Astro is one of the biggest ones. But yeah, we are constantly exploring and assessing and offering things with customers and seeing what works and what doesn't... If that makes sense.

1:51Janne Parri: Yeah, yeah, totally. And I as a freelancer just would love to live in a world where I could I could use just the one tool for example, just Webflow. But as a CTO do you like wish that was the reality that you could just focus on one tool?

2:08Alex Iglesias: Well, I mean, having just one tool it's convenient, right? 'Cause you just and especially if you're good at it. 'Cause 'cause if there's only one tool and everyone uses it, that means that you just become mainstream and you have to become the best on it, or otherwise you're just gonna be left out. So having one tool it's very convenient, but also it means that you're probably not gonna be able to solve everything. There's no single tool that solves every single use case. There's multiple use cases, multiple different companies with requirements for their own projects. So having choices is great, in my opinion. Having less choices is easier to manage because, you know, you just have a few people that are dedicated to working on that specific tool. But having more choices gives us the freedom and the power to determine what's best for the customer. And I think that during the podcast we can talk about, you know, different use cases,

2:58Janne Parri: Yeah, yeah, totally. We're gonna dive deep and we're gonna talk about Astro and probably some other tools too. But let's do little intros before we get started. You are now listening to Webflow and Friends podcast. My name is Janne Parri, and my guest today is Alex Iglesias. Alex is the CTO of Finsweet. He's now located in Catalonia, Spain, and he trained as a mechanical engineer and then fell for web development. At Finsweet, he has been the technical lead behind Attributes, Finsweet Extension, the Cookie Consent, which is now called Consent Pro, and also a product lead for Wized. And lately you can hear and see him quite often on the Finsweet Spot show where he has been talking.

3:49Janne Parri: A lot about Astro builds. But before Astro and other stuff, Alex, from mechanical engineering to the CTO of Finsweet, how did that happen?

4:00Alex Iglesias: Well first of all I'm curious how do you how do you get that 'cause just to be clear, I didn't give you this information. So I reckon I you probably went over in LinkedIn or something just explaining my background. I'm curious about that.

4:13Janne Parri: Yeah, LinkedIn is one of the best sources but there are all sorts of articles and so on and I have to admit that I do use some AI to find the sources and so on.

4:25Alex Iglesias: Right. Yeah. It's easy. Just Claude give me everything about that guy. Yeah, okay, makes sense. Great. Well, yes. So I initially studied mechanical engineering and I only worked as a mechanical engineer during my practicum in uni. I actually like worked on a on a real position being paid as a mechanical engineer because I immediately already pivoted to web design. Yes. Cause so essentially what happened is that so I love maths, I love numbers, I love coding, I love all of that. But I made the wrong choice in career, I'm gonna be honest. I chose one of the careers that I thought I would love. Initially it wasn't mechanical engineering, it was electric engineering.

5:12Alex Iglesias: But during that time, I decided that I hated electricity. I l I decided that I didn't like working and designing el electronical components. So I switched to mechanical engineering because that was more fun to me. But during that time, I actually learned how to program. Because in mechanical engineering and electronic engineering, you learn how to program. I don't know how it's nowadays with AI. But back in my days, we learned to do it manually. And I realized that I actually loved programming. It was the most exciting thing that I did on my career. It was like, all right, I'm studying on how to build machines, how to design machines. But where I really loved was the typing of some logic in a program so I can execute things. During my time while I was studying, I actually started freelancing doing websites.

5:59Alex Iglesias: You know, I wanted to build things and I had some friends and different family members who needed websites. So I was like, all right, I'm gonna do it for you because I wanna do this. So that was the time when I started just learning how to program websites manually. It was a great time, just fun. I had to study for uni, I didn't like it. I just I started just programming on the site, it was fun. Anyways, so long story short. I actually started my own company while I was still finishing my degree. So the moment that I finished my degree, I already was, you know, working as a freelance. I had a few clients. And I think that it was around the time when I discovered Webflow actually. So I just felt that I liked the tech career path more. I just felt like it was more fulfilling.

6:45Alex Iglesias: That I would what I was studying and doing in the uni. I finished my career because I wanted to make my parents proud, but I never used it for anything other than just learning. And during that time then I was a freelancer, my way to get clients was doing open source. Cause I found that it was a great way to making me you know, to get people to know me. So I put my, you know, my learnings and skills to use and I started building things for the community. In that in that moment, because I started to get involved with Webflow, and I felt that my newish programming skills that I just got were kind of like a nice fit into the ecosystem because I could build things that people could not do. That circles back a little bit to if there's only one tool, you need to do something to stand out, right?

7:35Alex Iglesias: So I decided to mix these two different skills that I was that just got and building stuff for the community. I was building you know, open source libraries for I remember I built something for multi step forms, it was very popular. But I also was building like demos and things that could be cloned, etc. And that catch the attention from Joe, the CEO of Finsweet. Joe is great at hiring. 'Cause he always looks for talent and he decided that I was an interesting hire, potential hire. So he reached out to me. He was like, Hey, you wanna drop everything you're doing and join Finsweet? And at the time, I was creating my own agency. I had already two people working for me. But yes, but Joe is very convincing. And actually I was already looking up for to Finsweet at that time because Finsweet was

8:27Alex Iglesias: You know, very well known in the Webflow community. They great did great things and I was a fanboy of them. So yeah, I just dropped everything and joined the company. Not as a CTO. I joined as a I could say developer technical lead for custom solutions. It was a mix of a little bit of everything. And then inside Finsweet, I grew in the I grew within the company until eventually it became the CTO, which was a great career path for me.

8:54Janne Parri: Yeah, yeah, that's interesting. So I gotta ask, like you mentioned that you started actually developing like static sites so you do not have a dark WordPress background like many of us do.

9:07Alex Iglesias: No, no, no, hold on. I do. I do. Because so I was I learned how to program and I learned how to build websites myself because that's what I actually love to do. I wanted to build things from scratch. But when I needed money and I needed clients to do things, that's when I actually got to work with the dreaded tools of Elementor for WordPress or Wix and Squarespace and all these all other dreaded platforms that were very limited. So yeah. I've done everything. I also have a dark past.

9:40Janne Parri: Yeah. As many of us do. But what about nowadays? What does the CTO of Finsweet do?

9:46Alex Iglesias: Well, so you I think that you did a great brief at the beginning. So I lead all the product development. We've got different products at Finsweet, but nowadays we're especially focusing on Consent Pro because it's one of the biggest products that we have. So I lead all the initiatives. There's a great development team working on those products. And other than that, I use some of my time to explore the new directions for Finsweet. One of these directions is obviously is Astro and things that we're cooking around Astro that maybe we can talk about. Maybe I can do a little a little hint. But yeah, exploring. Also lots of client calls whenever and you know there's a high profile client that requires some technical talks. I'm always happy to be involved within the company. And then everything related to leading the product's development.

10:34Janne Parri: Yeah, yeah, beautiful. And Consent Pro is the consent management solution. I'm user of it and I always say the cookie consent pop-ups, that's like the worst part of my job. So you guys do make it easier. So I'm so super grateful for that. But could you could you go a bit through what are all the products and services Finsweet offers and what are like the percentages, like how much is Consent Pro versus for example the agency business and so on? Just like ball ballpark numbers.

11:07Alex Iglesias: How much wood? How much effort?

11:09Janne Parri: How much of like the whole pie and whole focus efforts also?

11:14Alex Iglesias: Yeah, well I can tell you that around eighty percent of our resources today are going to Consent Pro because that's our biggest and fastest growing product and the one that we are heavily investing on because we've got we've got a great roadmap rate, we've got lots of plans for the product, not just for the Webflow ecosystem, but beyond the Webflow ecosystem, which is again another topic. So eighty percent of our resources are going into that. Twenty percent of the rest it's going into maintaining the rest of the products. The rest of the products right now they are as mentioned actively maintained. But those are products that were built in a pre-AI era that were

11:54Alex Iglesias: Solving problems that many people nowadays don't have anymore because of AI. We're still maintaining them. That doesn't mean that we've abandoned them, but we know what is replaceable today with AI and what is not. So we're putting our resources in what we are betting. It's something that we can still bring value to people, even if they have the full potential of AI on their hands.

12:18Janne Parri: Yeah, that's I wouldn't have guessed that like Consent Pro is such a big part. What about the agency business? Is that sub substantial at all?

12:29Alex Iglesias: Yeah, yeah, yeah. So actually, so products is one side, agency is the other side. When I said 80%, that only affects products. Because agency is still by far the biggest, the biggest business within Finsweet. And during years, for years, the products division has been subsidized by the agency side. So all the products that we've built, they're all self funded. So we don't have any external investors, everything is being funded. Internally. From the beginning it was money from the agency. It was it was a great synergy, I gotta say, because it was we were building products which were giving great marketing to agencies. So it's always been a great, a great business model. But yeah, the biggest revenue agency products is now starting to break even. Starting

13:17Alex Iglesias: You know, to become its own thing that could become independent of agency if it wanted to, if that makes sense.

13:23Janne Parri: Yeah, yeah, totally. That's interesting. And I get it. I get it.

13:28Alex Iglesias: Changes super fast as well. You know? You think one product is going to take off and then suddenly the market changes and you know it's a risky market. It's a it's a very unstable market.

13:38Janne Parri: Totally. Sometimes sometimes I'm like thinking that why am I in this technology business like there might be something that is a bit more stable. But yeah, I'm kind of a nerd so I like it. And I also like think that whenever I'm like stressed about let's say Webflow's future or how the market is changing. I'm always remembering that okay someone has like a huge agency build on the same foundation and if I'm

14:09Janne Parri: Like stressed out what about those agency owners? But yeah we can pivot to the possible solutions so as we have teased a of couple of times we're about to talk Astro And maybe the first question could be like why is everyone so excited about Astro?

14:28Alex Iglesias: Alright. I would love to do like a little breakdown. Because I mean the easiest thing to say is just because AI and people can build things and that's it, right? You know, but it goes I think that it goes beyond that, and especially for Webflow people. Cause i this can the experience can be different for people who comments from other platforms like WordPress, for example, but specifically for Webflow. We're talking about tons of agencies and freelancers that not only have spent years just visually drag and dropping things with this no code movement that was so well marketed, all these people is people who actually have learned how to use HTML and CSS. Because Webflow, you know, it's this own beast of a product that it never decided to just go abstract and

15:17Alex Iglesias: Keep things simple, where you know, in WordPress you could just install a theme and then just update a couple of things and maybe drag and drop a widget and just call it a day. In Webflow, since day one, they decided that they wanted to get closer to what's actually happening under the hood, which is HTML and CSS. And evolving within the years, it's not it not only did stick to just HTML and CSS. Webflow has been adopting a lot of concepts from traditional development frameworks. You know, you're talking about components, you're talking about code splitting across pages, you're talking about integrations, like just lots of things that are, I would bet, that are heavily inspired in how you work in a traditional code base with a framework like Astro, but it could be React or anything else.

16:04Alex Iglesias: So all these people that, you know, two years ago they thought that they were just no coders, they were actually not just no coders. They were drag and droppers of code. That's how I like to call them, right? They were building things with code. And it was and it's still today, because I wanna make it clear, I don't think Webflow is dying, and I think that Webflow is still gonna keep growing. But I don't wanna you know branch out here. All these people who know what's happening on their hood the hood, suddenly they realize that without having to learn JavaScript and learn traditional development of frameworks, just with AI and their knowledge that they already have, they can spin up sites that look and feel exactly or better.

16:48Alex Iglesias: Than Webflow in some cases, but without the vendor lock-in of Webflow. Because in Webflow, obviously the platform is great, but you also pay a lot of money. And you also if you want to switch, you have to do a migration, which is rebuilding things again. So I think that's what it's getting people excited. Like, hey, up until now, I was going to the city center, paying for a taxi every day, 20 bucks. But now I can use, you know, the Tesla new autonomous taxi and just pay three bucks and still gets me to the place faster and cheaper, right? So it's this kind

17:24Alex Iglesias: Of excitement of just realizing that you can achieve the same with your knowledge, but with some benefits on the max.

17:31Janne Parri: Yeah, the openness is one of those benefits. I believe Astro sites in some cases probably perform better in like loading speeds and so on and it might be more native for the AIs to like edit. So are those like correct assumptions?

17:49Alex Iglesias: Yeah, you've got AI who's doing a great job at just coding for you. When you're working with Webflow, AI needs to deal with this abstraction that it's Webflow that sits in the middle. So although I think that they're doing a great job at improving the MCP, so AIs don't think in Webflow, they think in code. I'm pretty sure that's the direction of the MCP. There's always, you know, this middleman of Webflow. And when you have an Astro an Astro project, AI is just gonna work directly as Astro. Which is always gonna give you better results for sure. Performance is great. No vendor lock-in besides the framework, obviously. You can deploy Astro to many different platforms. You can deploy it statically, you can have your own node server, you can deploy it to Cloudflare, to Netlify, to AWS. Like you've got all these different adapters where you can just decide where to put it. And then there's it has a lot of built-in.

18:39Alex Iglesias: Functionalities, you know, for reusability, for delivery, images, assets, islands for better loading things. I don't know. It has all these different goodies that I think that gets people excited about being able to build better things.

18:53Janne Parri: Yeah, yeah. And if we stay in the like Webflow context like for people who are coming from Webflow, what about like the downsides? Obviously you get more control, but with control comes responsibility and you're now responsible for stitching all the all the things together. I know that you have been talking with Joe on Finsweet Spot about all the things and

19:17Alex Iglesias: We have, yes.

19:18Janne Parri: I'm right now going through all the things and I gotta say like when I'm going through those things I miss all the stuff that Webflow took care of for me. So if I try to formulate this to a question like who should start using Astro or try it out and who maybe is better off using the Webflow classic, if you may.

19:43Alex Iglesias: Webflow Classic, I love that. Are you differentiating between Webflow Classic Webflow Source? I guess. Okay, great. Well well, it's clear that you have more responsibilities, that's for sure, because now you're managing a codebase and if things break, they are on you. If you're working on Webflow and things break, it's on Webflow. And you've got this billion dollar company with a lot of engineers that has SLAs in place and they are gonna be working on fixing whatever is happening. But that being said, you know, th this was a big thing before AI, 'cause

20:14Alex Iglesias: You needed at least somebody who understood the code to be able to just dive into the code and find what was breaking, which you know it's resources that you need to invest. While nowadays, for I would say 99% of the potential issues that you're gonna run into, you're just gonna tell Claude, site is broken, please help me. And it's gonna do something, right? So you know it eliminates a little bit this factor of responsibility that you have. That being said, one thing is building something, and the other thing is building something great. There's a big difference, I think. And to build something great, it requires you owning and understanding everything that is happening and having a clear vision, which in Astro, because you manage everything, it means a lot more overhead, a lot more of responsibilities and things that you have to think about.

21:03Alex Iglesias: In my opinion, going back to your question, because otherwise I'll just keep talking. I think that everyone should try it. Because it's not until you tried that you know if it's gonna work for you or not. And especially you tried and you put it into an actual real life scenario. Like, all right, you've got this new client coming in. Usually maybe you would be suggesting to the client that you build in Webflow because it's great. Maybe that client doesn't need Webflow. Maybe it does. We can talk about because Webflow really offers a lot of things that you don't get with Astro. And people has to be aware of that. You know, you're putting aside a lot of things that Webflow gives you, like, you know, role-based

21:40Alex Iglesias: Access controls, the editor, the CMS, the backups, the like it has lots of things that you're not gonna get out of the box with Astro. But if you get a project that you see that potentially the customer would be okay with it, you go ahead and try it. And then you're gonna have a real life experience and you're gonna see how it's like when getting requests from the client. How does your workflow look like? If the client suddenly wants to be the one editing the site, are you gonna be feel comfortable with it? Are you gonna let the client just prompt Claude and destroy the website? Are

22:14Alex Iglesias: I know, it's something that you need to find out, you know? So I think that everyone has to try at least to get the experience and be able to compare.

22:22Janne Parri: Yeah, yeah, totally. I'm exactly at that point. So I'm doing my first Astro build and the thing that bothers me is I kind of lack the confidence that I have with Webflow. Like if there's some issues on the Webflow side, it those might be like my own fault. I have experience on those and I can go and fix them. Obviously AI is there to help me, but I haven't gone through many like issues and bugs. Maybe after a couple of projects I am confident that okay this is an actual like good package to offer to my clients. But yeah the experience part is lacking at that point.

23:05Alex Iglesias: Nice. I presume that you've been just building everything with AI in the for this Astro project? All right. And did you achieve exactly what you wanted or did you have to settle with whatever AI came up with and you just stopped fighting it?

23:22Janne Parri: Yeah, I'm achieving like almost the same level as I would with a Webflow project but it takes a lot of prompting, so it is not what you see on the on the internet like one prompting and awesome site yeah, yeah.

23:35Alex Iglesias: No, I hate those videos. Yeah, absolutely.

23:38Janne Parri: That's my personal site that I'm redoing but I'm thinking that I could start a bit more manually next time I start trying out something and with that project I didn't use any like boilerplate or ready-made framework but I was thinking that for example Timothy Ricks has the Lumos for Astro I could give that a go and so on. Actually one big challenge. I know that you guys at Finsweet are doing the whole like run through with all the all the like technical stuff, so we do not need to get into a lot of details and every aspect that one needs to take care of when they are building a Astro project.

24:23Janne Parri: But I'm I am curious about like the client side editing. If you wanna give some sort of at least simple interface to edit the site, what are the options for that?

24:35Alex Iglesias: That's great. Well, the easiest one is just giving it a CMS, right? There are great options for Astro. Yeah, recently was the V1 version of EmDash, which is a CMS specifically designed for Astro by Cloudflare. As this has been shipped and I think that for example that's a already great solution. Or you get all their CMSs, headless CMSs like Sanity or Contentful or Prismic or all of those. But obviously you're just gonna give them controls to one part of the project. It's mostly around the content collections. You can still give them more control. Like for example, Prismic, I that they've got the slices concept, which kind of you map components within the your code base, so then

25:16Alex Iglesias: Within Prismic, the clients can be drag and dropping different components to compose pages. Like there are solutions like that. I believe that Sanity has something similar as well, or I've at least I would imagine that they do have. In EmDash, you can achieve something similar as well if you set up a content collection that in the code base maps to the components. So then when the they use a specific value, you can inject the components. Like there are ways of doing it. But at the end of the day, what mostly works is the CMS. If you need your clients to have full control, like in Webflow, where you know they get this designer and they can just explore everything that exists, drag and drop sections and compose things differently, et cetera. I believe that there's a big gap in the market right now. And I and there's something, you know, that's why we're actively exploring, because we're trying to

26:08Alex Iglesias: To find and fill this gap if possible.

26:10Janne Parri: Have you found any anything that you can mention? And maybe could the Source by Webflow be that filler?

26:18Alex Iglesias: Could be. Could be. I know that Source is very ambitious and I know that it's very AI first focused. So I believe that their main focus is gonna be mostly around building agents that help you iterate your site with AI. Although I believe they also have like this visual editing and stuff like that. And I know that there are great community projects is also popping up. I know that the what's that project by Julian?

26:42Janne Parri: Ship Studio.

26:43Alex Iglesias: Ship Studio, it's a great product as well. Timothy Ricks also explored building the again, I forget the names all the time. The Stacki as well.

26:51Janne Parri: It's pr was it Stacki or something like that? Yeah.

26:55Alex Iglesias: The there's lots of things popping up. And same as it happened ten years ago where there was this gold rush race towards becoming the best framework. You know, React and Vue and Svelte, all these components were just competing to become the best framework. I believe that now there's also race for finding the best way of managing sites that are code-based sites, not a closed platform like Webflow, giving controls to the non-technical people. So no no definitive solution here. I would be a that I could, you know, tell you. But I all I can tell you is that we are also doing our own.

27:30Alex Iglesias: Exploring of building our own internal tools for our clients. Yes, we are.

27:34Janne Parri: That's interesting. That's interesting, yeah. I have used many Finsweet tools, so I'm happy to see another one that could fill my use case. Yeah, go on.

27:45Alex Iglesias: No, I just wanted to say 'cause what we're trying to do right now is answer this same question for our clients. Because all the websites that we're starting to offer as Astro sites for our clients, if possible, we want them to get the same experience that they would get in Webflow. With the editor, with all the capabilities. And as mentioned, there's lots of things that are starting to pop up, but there's nothing that has being defined as the go-to tool, as the best tool for doing it. So what we're doing is building for now just kind of like our internal interpretation of what's the best experience, but not as a product that we want to sell, but more like an in-house exploration. So we can offer to the clients, see how they react to it, how they work with it, and then we'll decide how to where we go from there.

28:34Alex Iglesias: Hopefully in the future I can I can I can show you more about that depending on if the experiment goes well or not.

28:41Janne Parri: But you mentioned the clients and I asked who should like try out Astro but if we switch that question to the client side, like what kind of company what kind of client should like move to Astro and the code base type of style.

28:58Alex Iglesias: Well nowadays, and I think that this circles you know connects very well with what we just talked about. I think that nowadays the best customer is the customer who's going to be relying on you to fully manage the website. So if they want new pages, they will ask you to do it. If they want to update the content, they will ask you to do it. Or at least they don't expect a lot of control around the website. Because they still get benefits. They will get cheaper hosting for sure. They will get better creative freedom when building the site. And they will get great website experience with all the performance of the upgrades, etc. But what's true is that the from the customer side, the experience that Webflow offers is super hard to beat. Super hard.

29:42Alex Iglesias: So I think that the ideal customer is the one that is not expecting to get everything that you currently have in Webflow. The editor, the CMS, the previews, the password protection for things, being able to share, like role-based extra controls, like to all those sorts of things are hard to beat.

29:59Janne Parri: Yeah, yeah. I'm just I do not have like proper experience with this, but I'm thinking like two kind of use cases. Either the client wants almost like no access to the site, like it's super static, like maybe a small client that you can whip up a quick website and host it almost freely. And then some bigger client that might actually have in house developers who want to have a open code base. So what do you think

30:33Janne Parri: Of that?

30:34Alex Iglesias: Yeah, absolutely. The good thing of this is because previously it was very distinguished between going to an agency for a no code build and going to an agency for a full code build. It was kind of like two different agencies. You wouldn't go to the same agency, right? You go to the Webflow Pros or you would get these traditional web development or development companies that they have all sorts of skill sets for all these different languages, etcetera. Now things are starting to converge a little bit and I think that everyone is starting to be capable of doing everything.

31:06Janne Parri: Yeah, yeah. But I feel like at the same time, like if we wanna hand our clients the access to edit the site, the challenge is almost the same. Like even with the Webflow builds, you could like and you still can like build a beautiful component system and teach the client to use that but it's like another question is the client actually going to do that or are they gonna

31:28Alex Iglesias: They destroyed.

31:30Janne Parri: Just ask you to do it? So I'm a bit puzzled about this challenge.

31:35Alex Iglesias: That's why I say that 'cause I don't have a an answer here. I think that we are all learning all during this process. Because everything is evolving so fast that this is a matter of six months. Like six months ago, everyone was just, you know, doing great things at Webflow and not even thinking about coding hands manually or with AI. To within six months, people still hasn't been able to close some projects or to deliver a project and on board the client and get the feedback and understand things. So yeah, I think that we're all learning and

32:09Alex Iglesias: With experience we'll we're gonna be able to tell better when a customer is a great fit for a code-based site or just a Webflow. 'Ca it's still a great experience in my opinion.

32:21Janne Parri: Yeah, yeah, totally. I'm thinking a couple of like detailed, a bit nerdy questions. We have been talking about Astro, about that the other thing that people are using is like Next.js sites. Like I do not know almost anything about those, so how do those compare?

32:40Alex Iglesias: Yeah, okay. So Astro was originally, because nowadays you can do lots of different things, Astro was originally thought around static sites, which is also what Webflow has always been good at. You if you wanted to build a web application with Webflow, you probably still had a big reason to have static static parts of the website and that's where you know tools like Wized that we built came in. But the moment that you wanna go beyond and build a full web app experience that it's very dynamic, Astro can fall short because you can do still many things, but these meta frameworks that are at the Next.js, Nuxt, SvelteKit, TanStack, like all these meta frameworks are built specifically around dynamic.

33:28Alex Iglesias: Highly interactive websites, you know, you would build Webflow with Next.js. You wouldn't build Webflow with Astro, if that makes sense. Actually I think that I

33:37Janne Parri: Yeah, yeah. Totally. So

33:39Alex Iglesias: Believe that Webflow is built with Next.js.

33:41Janne Parri: So be basically Astro is meant for the static sites and with Next.js and all the others you can you can do all sorts of applications. And do those like Next.js sites then need to be rendered in HTML and CSS or am I getting that like wrong

34:00Alex Iglesias: W well at the end of the day, everything is going to a browser, right? So anything that loads on a browser is either using, well, it's using HTML, CSS, and JavaScript. Could be using a little bit that it's not like that with WebAssembly, but we're getting into a rabbit hole in there. So when you build a website with Next.js, the you're still shipping HTML, CSS, and JavaScript. There's nothing different about that. It's mostly about the ergonomics and what the main focus of the result is. In Astro, the main focus is pre-building things, generating static files that load very fast, which is great for a marketing site, and deferring the interactive bits, which is usually known as the islands. But then the

34:47Alex Iglesias: Frameworks, meta frameworks like Next.js and all that stuff, they're mostly focused about hydration and injecting new things on the page without leaving and loading all sorts of data dynamically.

34:57Janne Parri: No, yeah.

34:58Alex Iglesias: So the ergonomics of them are like completely different focuses.

35:04Janne Parri: Yeah, yeah, interesting. And those like options are strongly on my mind just because Source by Webflow in the demo had those options and I believe they are trying to ship those options. And I'm curious, we mentioned Source before, but what do you what do you think of the Source? And I also in the last episode talked with Kyle Lambert who has been designing the actual Source app and he mentioned that the Webflow sites that you import they will be migrated to some sort of code base. So what do you think of

35:42Janne Parri: All the all the stuff with Source?

35:45Alex Iglesias: Interesting. I didn't know about that. Cause we actually haven't gotten our hands on Source yet. We are wishing to become part of the beta team. Webflow, if you're watching these, please let us try it. Give us access, please. So yeah, very interesting. It makes sense though, because as mentioned before what agents are good at, it's not about it's not at calling tools. They're good at them, but they're better at just generating code. So if you need to tell an agent to update a Webflow site, like let's imagine that it needs to create a hero section, right? When it's code base, the agent is just gonna go to the file that represents that page.

36:23Alex Iglesias: And it's just gonna see the HTML code and it's gonna be like, all right, yeah, I just need to update this. I'm generating text, right? I'm just changing the code. But in Webflow, you the agent first needs to call an API that tells Webflow to return the page structure. Then it needs to call another API that tells Webflow to return the CSS styles of that structure. Then it needs to determine how to change things and then call another API to tell Webflow to update some of the CSS classes that exist in Webflow. Another API to update the HTML of you know there's

36:55Alex Iglesias: There's lots of steps involved into getting the same result, but the AI agent needs to think way less when there's code in the middle. Right? So to me it makes total sense that they want to translate this Webflow layer into just a code base that gives you the same result.

37:11Janne Parri: Yeah, yeah, totally. When they announced that I was kind of expecting some sort of Webflow two point but instead there was there was Source. I guess in hindsight that is like the logical way to do like if you wanna stay true to the code as the source of truth, yeah, you do that, but I'm like you waiting to get my hands on on the Source and seeing what it actually can do.

37:38Alex Iglesias: What were you expecting of these Webflow two point curious.

37:41Janne Parri: I don't know I don't know. Somehow like some sort of Webflow styled visual editing that would in a way in a way like edit straight out like HTML CSS and be somehow more native to use with LLMs than the like Webflow Classic as I'm calling it. Some other folks have been calling it Webflow Classic too. Yeah, I'm not totally sure what I was ex expecting, but yeah, we got the Source and I'm like thinking it is a cool idea and I'm hoping that they can pull it off and do some great things with it.

38:21Alex Iglesias: I've got a lot of questions, but what I need to do is go and ha watch that episode 'cause otherwise I would be asking things that I wanna know. Nice.

38:29Janne Parri: There are not like two specific things. I like asked Kyle about the importing thing and he was he was quite vague he said that the old sites can be imported. That means that the new version will be created in some sort of code base and that's all he said he said that they it's an important part and they are working on it right now and they wanna make it make it like good because it's super important obviously, but there needs to be some sort of way to start using Source. And I tried to ask about pricing timelines, but no no luck with those.

39:07Alex Iglesias: Everyone wants to know that we don't know if it's gonna be five bucks a month or five hundred bucks a month. It's gonna be so different.

39:16Janne Parri: I do have well, they are probably doing the decisions as we speak and trying to figure out that, but my speculation is that there will be definitely some sort of like token based pricing, maybe a base price and then if you wanna spend more you can spend more and so on. What do you think? What does your crystal ball say?

39:38Alex Iglesias: Well, I think that there's gonna be a base price just to access the tool and it's gonna be per seat. Absolutely. So if you've got a team of ten people, you're gonna pay for ten seats, that's for sure. What I'm not sure about, and it's actually something that I would love to see, is if they expect to become token resellers with this tool, or if they just become the tool that lets you use your own subscriptions for LLMs. Did can you mention anything about that? I need to know it. No? All right.

40:05Janne Parri: Yeah, yeah, that has Kyle didn't Kyle didn't, but I have heard about some something that probably some tier needs to have the integration to like bring your own agent basically. But that might be an open question, how does that work?

40:24Alex Iglesias: Yeah. Cause to me I think that it's gonna be a very big deal breaker for a lot of people. Cause you know, with like for example, if we go back to the Ship Studio tool that Julian's working on, it's a great tool, it looks great. And essentially what it's doing is just working with the local file system, right? You're just editing files like you would do in VS Code or whatever. So when you work with agents, it's your own agent, it's your own tool that's just doing the things. It's asked you're asking it to do things. It's your own subscription that you're using for everything. But if they suddenly just clock this requirement just that you can only use AI through Webflow. So you have to pay the tokens to them. That means that a lot of people who are just exploring this new freedom that they're getting with AI

41:11Alex Iglesias: Are suddenly just gonna go like nope, because you know. I can prompt my way out of this and I'm not gonna pay for a tool that locks me into paying tokens to you instead of what already I already pay. So I think that it's a big that'll deal breaker. I don't know what's gonna happen. Hopefully they'll be able to manage this.

41:29Janne Parri: Yeah, yeah. I'm hoping that the bringing your own agents is not a closed feature for the bigger tiers or for the enterprise. And I'm with you. I'm totally with you that it would be a like turn off or buzzkill if that was like closed. I'm looking forward to looking at these speculations after we have received some actual information from Webflow. Okay, there was one nerdy question that shows how little I know about stuff, but that was again about Astro because

42:06Janne Parri: Back in the day with Webflow frameworks were a big talking point. And nowadays not so much. I know that Timothy Ricks has the Lumos for Astro but in general is frameworks or I don't know boilerplates or anything like that like a big thing in Astro world.

42:28Alex Iglesias: I'm going to assume that you're talking more about styling frameworks. Is that correct? Cause right, because so you've got Astro, which is the framework for building the website. And now what you care about is having like unified stylings and a st a style system that can be scalable and maintainable. So you don't get again a lot of duplicated code and a lot of inconsistent things, correct?

42:55Janne Parri: Yeah, yeah. Totally that and make maybe on top of that some ready made like building blocks components or something like that.

43:02Alex Iglesias: Absolutely. Well this topic in traditional development it's older than the Webflow discussions or th the Webflow style guide fights between all these different popular different frameworks, Client-First being one of the one of the ones that a lot of people are using from Finsweet. In traditional development frameworks or just let's not call it traditional development. Let's just call it, you know, custom code frameworks. There's lots of solutions. Lots of solutions. You've got, you know, just plain CSS files. You've got Sass, which was very popular back in the days. Within the CSS, you've got different naming conventions. You got the BEM convention and that were many others as well that I just forgot.

43:48Alex Iglesias: But there are also ways or solutions to try to improve how the styling is used, like Tailwind, that you just go with utility classes. You've got CSS and JS solutions where you inject the CSS just using JS. You've got solutions like StyleX, which has become very popular later, because AI agents are very good at using it, that let AI agents write CSS styles using TypeScript, which makes it type safe and they know how to reuse styles. It's actually very it's great. I've been trying it out and it's nice. So there's lots of solutions, lots of different ways of writing styles for projects. And I wish I had like the perfect solution to tell you.

44:34Alex Iglesias: But I think that it's a matter of preference. 'Cause AI agents are gonna do a great job if you if you give them the right explanation of what you wanna get. It could be just plain CSS and they would do a great job. As long as you structure and organize the code base properly. And then you can go with the these new popular solutions like StyleX. I've been getting great results with it as well. So it's a matter of preference, in my opinion.

44:56Janne Parri: Yeah, yeah. It just happens that Timothy has done that one. I don't know what we should call that like a starting point for Astro. And I'm thinking you mentioned that LLMs will do a great job of building kind of your own system. So in the future let's expect that AI is getting going to new levels. Will all the frameworks and all the all the systems that we have been building will those just get obsolete and everybody does custom for every project.

45:30Alex Iglesias: I do have an opposite thinking around that. Yes.

45:34Janne Parri: Ooh. Please, please explain.

45:37Alex Iglesias: Cause I think that the more frameworks and opinionated decisions we have, the better results you're always going to get. So so because let's put into perspective. It's true that you could right now just prompt AI and build a site without any framework at all. You know, you really don't need Astro. You can just tell AI, hey, build this site and it's going to use HTML and CSS, right? But there's gonna be some point where you're probably gonna want to reuse the specific parts of the website into different places. So now suddenly you now have to come up with something that we've already invented, which is the concept of components. Right? So now you're reinventing the wheel and implementing components as a this new this new logic or whatever we call it in your project.

46:25Alex Iglesias: And then you keep building the project, and then you realize that you want to inject a React component within your website. So now you find yourself wanting to implement the islands approach from Astro. So you can prompt AI and it's gonna build it. It's gonna just create this huge code base with all the logic required for it, right? But you're gonna keep building and then you're suddenly gonna realize that you need to interact with the server when you submit a form. So now instead of using Astro Actions, which it's just out of the box, now you got your AI agent just building this framework from scratch when this wheel was already invented. Right? So it's easier for an AI agent to use a framework that exists because that

47:10Alex Iglesias: Framework is already solving problems that don't need to be reinvented again. So the AI agent it can focus on Grabbing these building blocks and putting them together faster and better for you instead of reinventing the blocks themselves. Right? And I think that this same concept applies to the styling frameworks, the conventions, etc. For sure you can tell the agent to come up with the framework himself. But if there are already studied things that have been tested, battle proven and they work, then the agent is gonna do a way better job at using the that convention instead of coming up with a new one. Is this making sense?

47:48Janne Parri: Yeah, totally. Like the or totally. My like brain is almost out of like capacity of thinking these abstract things and trying to predict the future about topics that I'm not totally an expert on. But you mentioned obviously the frameworks and all open source earlier. What is this thing that I have heard that open source projects have some challenges because of AI. So is th is that a thing? Do you know what I'm talking about? Is this related to

48:21Janne Parri: This prediction that we are talking about?

48:23Alex Iglesias: Well kind of it's it is related. So the thing is that nowadays when you work with an AI agent, the AI agent the more effort you want the AI agent to put into something, the more it's going to be questioning itself about what it he is doing. I'm gonna say he, although it's an it is doing, right? So if you're trying to build something with a an open source tool and the agent decides that what he's trying to achieve requires an update to that tool, the agent

48:54Alex Iglesias: Nowadays are autonomous enough to just go into GitHub and fill in a GitHub issue to that open source package, saying, Hey, I'm trying to achieve that, but this issue. Now I suggest that you do this and that and that, right? And because this is happening Automatically, like you you now start prompting Claude to do something. You know jack shit about that open source tool. You don't know anything, right? But your AI agent could decide that he wants to fill that issue because that's his goal, right? So what is happening is essentially that these open source pack maintainers that are suddenly just being flooded with tons of different AI agents that want to do things that perhaps don't align with the roadmap of

49:36Alex Iglesias: The package. Or maybe just the maintainer doesn't have the resources to just attend everything that is happening, right? So the issue comes in more from this side, which is it's a good thing, it's a bad thing, you know, 'cause you you you get the AI agents that are reporting things and helping you iterate the product, but at the same time you also get a lot of noise, which it can be it can be overwhelming.

49:58Janne Parri: Yeah, yeah. I can relate in a sense that I always hate when I get like clearly AI written email. Why do I have to like go through this and so on? And I'm just imagining if I was maintaining something and I got like a bunch of requests clearly done with AI. So it doesn't feel right for some reason. Yeah. That is that is fascinating.

50:23Alex Iglesias: I mean, 'cause before AI, you know, you would get real humans who would spend time of their own lives to decide that they want to contribute to that open source package and go and read the copies, understand the issue and kind of sync with the maintainers. But now all of this comia humanity has been put aside and it's just a factory churning and going, you know, without proper directions just because i every AI agent takes its own decisions, so there's no unified direction for anything. So i it can be very chaotic, it can be very overwhelming.

50:59Janne Parri: Yeah, yeah, totally. I'm loving this future reflecting abstract stuff. And I wanna ask about like general predictions about AI, but you also mentioned that you wanted to say something about Webflow's future. So what was what was that

51:16Alex Iglesias: Mm-hmm. Well, first of all, I think that Source by Webflow is the natural step that they had to take. Nobody knows if it's the right step. Nobody knows if you know it's

51:25Janne Parri: D yeah.

51:26Alex Iglesias: If it's gonna succeed. But for sure we know that based on the direction of the market, Webflow needed to get closer to the code. And Source is their answer to getting closer. The code. So I think that Webflow is it's not going away, that's for sure. Especially for all these use cases where you've got a customer who needs to manage the website. Webflow is still great at it. You know you're giving your customer a full platform with all that they need.

51:57Alex Iglesias: They've got the editing experience in there. You've got the CMS experience in there. They've got even analytics and A B testing. Like they've got this whole platform that potentially the just the customer can just be there and work with, right? The only thing that was missing was a very nice AI experience with it. But to achieve that very nice AI experience with it, that's what requires Webflow to get closer to the code. That's what we've been talking about before. So in my opinion, I think that Webflow is just going to be slowly converging into this code first approach that AI agents love and are very good at, while still offering what they're doing great at, which is offering marketing tools and tools for non technical people to manage the sites.

52:41Alex Iglesias: So to me, this is a hot take because I don't know, maybe in two months from now AGI comes in and the whole website is market is just destroyed. But to me, it is clear that customers have needs. And those needs, many of those needs, Webflow is still great at solving them for a marketing site. So yeah, that's my hot take. I know we're gonna get closer to the code and maybe it's gonna be Source by Webflow, maybe it's gonna be something else. Lots of people migrating away to code bases, but big customers who require an experience where they can manage things non technically, it's still a big thing. You have to

53:19Alex Iglesias: Isolate the hype of using AI with what real people need, if that makes sense.

53:25Janne Parri: Yeah, yeah. And I'm like thinking if Webflow can make the MCP better, they have already done some great updates and if the speed of building comes close, even like eighty percent of what we are doing with the more native AI tools and static sites and maybe also the site performance like if they could fine tune that a little bit, like I would be happy at the moment staying still with the Webflow builds, like those are the only issues.

53:57Janne Parri: And I haven't tried the latest version of MCP. They keep on doing updates and I believe if a client comes to me and says that they wanna edit the site with AI, that kind of already answers that. It is just a matter of how much and how fast and what kind of edits you wanna do. Would you agree?

54:19Alex Iglesias: Yeah. And I think that there's a it's actually a great path of exploration that probably Webflow is doing, especially with Source by Webflow. Cause today you actually just have two choices. You let the client edit the site with AI, or you don't let the client edit the site with AI. But if they can edit the site with AI, they can do everything. Like everything, right? So I think that we're also going to be seeing how they want to give controls to creators so customers can use AI, but only for specific things. So the environments are controlled. Because that's very important, especially for big corporation sites where they really care about what's happening on the website. So absolutely I'm sure if it if it hasn't happened yet, I'm sure that it's gonna happen.

55:04Janne Parri: Yeah, yeah, totally. I'm always getting excited about Webflow's future when I talk to people in the in the community and then I then I go to LinkedIn or Reddit and then there's some post that people are leaving Webflow and that gets me down.

55:20Alex Iglesias: You know, cause because you yeah, but you're only see gonna see those. You're not gonna see a post from somebody saying, Hey, today I awarded a client to custom a customer to Webflow. Like, that's not fun. That's not engaging. You know, the engaging thing is saying, Hey, Webflow is dead! I just rebuilt everything with Claude 30 minutes and I'm not paying Webflow anymore. Like, all right, it's a cool headline and maybe it works for you, but there's a whole world outside of needs, you know, companies that have needs that you're not addressing with this. So it's not just blacks and whites. There's lots of grays in the middle that need to be analyzed.

56:01Janne Parri: Okay, I still wanna get your prediction what is going to happen in general with all the AI are we gonna get what was it like AGI like big all knowing computers

56:11Alex Iglesias: Huh. Yeah.

56:12Janne Parri: And so on and are the big players OpenAIs and such going to fall down because of the economics and so on so give me give me something.

56:23Alex Iglesias: Well, I don't think that the companies are gonna go down. That's the same as saying that initially because there was the dot com bubble, the web died. That's not true. Like one thing it's the economic side, maybe there are companies that are overvalued, you know, and the market can crash, but the technology is here and it's not going away. That's for sure. Cause people is using it and people use it. The technology is gonna stay, right? About how things are gonna evolve. I think that I'm nobody to answer this, honestly. I don't think I do

56:54Alex Iglesias: I have this big high profile to answer this. But I can give you my take though. Because you know,

56:59Janne Parri: Please.

57:00Alex Iglesias: As a as a developer who've seen the whole evolution from manually coding things to starting to get AI and being very skeptical at it. And just being like, this is shit. I write much better code than it. And then just going through all the grief stages of just suddenly realizing that I can do some more things than I thought with AI. Actually, what I'm saying, it's very well reflected with one big talk that it's become very, very viral last week. I think that it was by DHH, who is the creator of

57:35Alex Iglesias: Ruby on Rails. Cause that guy's been in the industry for like 20 years, 25 years, always coding things. And now he's just like, fuck it, let's just vibe code everything because AI is doing all the job for you. I'm not at that point yet, but I do agree that AI is great at doing things. And AI is already doing many things that previously was my job. You know, I was coming up with an idea and then I was architecting the solution and then I was writing the code myself. And AI is doing a part of that job. You know, AI is taking my job in that sense. But at the same time, while I'm doing this evolution, I'm noticing that it just changing the way that I work, but it's not taking away my work. It's just making my work different. You know, now instead of thinking

58:20Alex Iglesias: About every single word that I need to type into code, I do have. What I think I wanna do in my mind, but I can explore with AI and I can decide I can determine together with AI what's gonna be the architecture and then I come up with a plan and then I can be doing a lot of tests and then ask AI to implement things so then later I can review it and test it. So it's just changing how I'm working. And I think that the better AI becomes, it's just gonna keep doing that. Everyone is just gonna be evolving with AI. I don't think there's gonna be a point where just suddenly AI is doing everyone's job. And now, you know, I'm here just twiddling my thumbs, not knowing what to do. I just feel that however AI evolves, we are going to evolve the way that we work with it to solve whatever is being demanded in the market.

59:11Janne Parri: Yeah. Yeah. Addive.

59:12Alex Iglesias: And I'm trying to be optimistic because cause I could be pessimistic, which I've all also been, you know, I've been through all the stages. And I could tell you that no, we're gonna lose our jobs in two years from now. But I'm being optimistic, I do think that's what's gonna happen. It's just going to be evolving how the market behaves.

59:29Janne Parri: Yeah, yeah, totally. And there might be a scenario that we are not gonna lose our jobs, we just need to do like more because we have all these new tools like in the past if there is something that has made the work more efficient, you usually end up just doing more.

59:47Alex Iglesias: Right, correct. Yeah. I mean, before Webflow, you would build a website manually and it was a website, but it was a fucking shitty website. It was it was way worse, you know? And with Webflow, maybe you'd build better websites and you were able to maintain them better, right? And now with AI it's just the same thing. You you are able now to explore using Astro. And because you can explore using Astro, maybe you're gonna build a better experience because there are some things that Astro offers that you couldn't use before. So it's just this evolving of just delivering something that has more value because you have better tools for it.

1:00:24Janne Parri: Yeah, yeah, totally. Okay, this is a beautiful place to land the plane and there is

1:00:30Alex Iglesias: Just optimistic. Everyone, you're not gonna lose your job. Yes,

1:00:34Janne Parri: Just gonna do more of that. There is one little weird section at the end, but before we go there, if people wanna check you out, follow you, or maybe even contact you, where should they go?

1:00:48Alex Iglesias: W you can find me in x.com. I'm gonna be honest, I'm not active at all in social media. I used to be, but I've I'm not at all. But if you write me, I'm gonna see it. You can also find me in LinkedIn if you look for Alex Iglesias, or you can also drop me an email. My email is alex.iglesias at finsweet.com.

1:01:08Janne Parri: Beautiful. And I will obviously put the links in the description. Okay. The last section, so do you have any idea what it could be about?

1:01:20Alex Iglesias: Maybe asking me for striptease, I don't know. What it is.

1:01:25Janne Parri: That would be a pivot. But yeah, but no, not a not a strip tease. This will be a little Finnish section at the end. Like it is called Finnish Line because it is at the end and I will give you a line of Finnish like an actual saying of Finnish and I will obviously translate that to English and your job is to figure out or just guess like what does it actually mean. Like I will translate that word for word. It

1:01:58Janne Parri: It is usually something a bit weird and I will give you an example and then you try to guess what that means. So is that clear?

1:02:06Alex Iglesias: Yes, I initially thought you were saying finish and it I thought you were saying like for finishing the podcast.

1:02:12Janne Parri: A little like word wordplay. I actually

1:02:15Alex Iglesias: Love it.

1:02:16Janne Parri: Didn't come up with that myself. It was Nathan Huening from Webflow. Again credits for him. But your saying, firstly, in Finnish is vetää nenästä. And in English it is to pull someone by the nose. So what does it mean? If I say that guy pulled me by the nose.

1:02:39Alex Iglesias: He tried to trick you.

1:02:41Janne Parri: That is exactly right. Like yeah yeah. I'm like wondering is there anything like similar in any language? But yeah that's totally right and you could you could see yeah yeah you could

1:02:53Alex Iglesias: Yes, it is. Nice.

1:02:55Janne Parri: See like my surprised face because not too many guests get this right so you're you're probably in like in a group of five or something like that. Yeah. To trick someone or tell lies, like either one of those.

1:03:13Alex Iglesias: Okay. So can you can you repeat it? In Finnish?

1:03:14Janne Parri: That's yeah, that yeah. Vetää nenästä. I could say that

1:03:18Alex Iglesias: Vetää nenästä.

1:03:20Janne Parri: Yeah, beautiful. You're already talking Finnish. That's great. Okay Alex, this has been great. I got a lot of answers to my weird little questions. It was great to nerd out with you. Thanks for being a guest. And for all the watchers and listeners, we'll see you in the next one. Bye!