[00:00:19] Nathan Wrigley: Welcome to the Jukebox Podcast from WP Tavern. My name is Nathan Wrigley.

Jukebox is a podcast which is dedicated to all things WordPress. The people, the events, the plugins, the blocks, the themes, and in this case Elementor’s decade of growth and its future plans with AI.

If you’d like to subscribe to the podcast, you can do that by searching for WP Tavern in your podcast player of choice, or by going to wptavern.com/feed/podcast, and you can copy that URL into most podcast players.

If you have a topic that you’d like us to feature on the podcast, I’m keen to hear from you and hopefully get you, or your idea, featured on the show. Head to wptavern.com/contact/jukebox, and use the form there.

So on the podcast today we have Miriam Schwab. Miriam has been deeply immersed in the WordPress ecosystem for around two decades. Starting out offering WordPress as a service, she went on to lead a custom WordPress agency serving major tech companies and nonprofits before founding the startup Strattic, pioneering static WordPress architecture. After Strattic’s acquisition by Elementor in 2022, Miriam took the role of Head of WordPress acting as the key liaison between Elementor and the wider WordPress community.

Elementor growth over the last decade has been prolific. Miriam says that it now powers over 13% of the entire web. She gives insights into the challenges and responsibilities that come with maintaining such a large user base, especially around major updates and backwards compatibility.

Much of our conversation centers around the rise of AI in WordPress, from built in AI tools for generating images and content to the standalone Angie plugin that introduces agentic AI capabilities across WordPress. Miriam outlines Elementor’s multi-pronged approach to innovation, talking about how their Site Planner tool uses conversational AI to guide beginners and professionals from an idea all the way to a wire framed website. And how the upcoming AI integrations promise even more granular design control.

Miriam also shares her perspectives on how the new Abilities API is set to change what’s possible inside WordPress, and what this means for developers, designers, and support teams navigating the complexities of AI driven workflows.

For those interested in how AI is shaping the future of WordPress, Elementor’s strategy, and the evolving role of creators within this ecosystem, this episode is for you.

If you’d like to find out more, you can find all of the links in the show notes by heading to wptavern.com/podcast, where you’ll find all the other episodes as well.

And so without further delay, I bring you Miriam Schwab.

I am joined on the podcast by Miriam Schwab. Hello Miriam.

[00:03:23] Miriam Schwab: Hi.

[00:03:24] Nathan Wrigley: Nice to have you with us. Now, dear listener, I have to be very grateful to Miriam because about three weeks ago we tried to capture this exact podcast. In fact, we did. We captured this exact podcast. Unfortunately, the tech failed and Miriam’s audio was entirely silent. So we had a nice long conversation, Miriam divulged her experience and wisdom, and I got to put it out on the podcast and there was nothing there. It was completely blank. So firstly, an apology for that. And secondly, enormous thanks for coming back and talking to me.

[00:03:54] Miriam Schwab: Well, I have to thank you for giving this another shot because I have a feeling that the technical difficulties were also on my side. So thank you for giving this another go.

[00:04:03] Nathan Wrigley: Yeah, well, I appreciate it. So Miriam has a long and storied history in WordPress. She’s working with Elementor at the moment, but the story goes back way further than that. Let’s just do the little potted bio at the beginning, Miriam. Just tell us a, just a short version of who you are in the WordPress space, and what you’ve achieved in the past.

[00:04:20] Miriam Schwab: Okay, so about 20 years ago, I discovered WordPress. Loved it, decided to work with it, started offering it as a service. Eventually that expanded into being an agency, and working on custom implementations of WordPress for tech companies and large nonprofits, which was a lot of fun and a great learning experience.

Did that for about 13 years. Sold the agency and founded a startup based on one of the many ideas I had had over the years. I had many ideas for products over the years, but I knew that I needed to make sure I wasn’t over spending my abilities, like between family and work and everything.

But I got to the stage where I was like, okay, I can do this. And I actually really loved the idea. So I founded a startup that was called Strattic. It published WordPress websites in a static architecture while retaining dynamic capabilities. And by doing that, it solved pretty much all the issues related to speed security and scalability. Raised venture capital funding for it, and it was acquired by Elementor in June, 2022.

And after joining Elementor, continued to lead the Strattic team for about six months, and then they offered me this new position that hadn’t existed before. Initially, we called it the Head of WordPress Relations, and then it evolved to just being called Head of WordPress, where I act as a liaison between Elementor and the broader WordPress community on many levels. Practically, strategically, community, like you name it and I’m probably doing it. So that’s a bit about me.

[00:05:43] Nathan Wrigley: Do you have like a job description? Because when you say sort of Head of WordPress, there really seems to be almost no bounds in terms of the WordPress and Elementor connection. Do you have constraints on what is out of bounds? Or is it literally anything that’s connected with WordPress comes under your purview?

[00:05:57] Miriam Schwab: It is literally anything, pretty much. It suits me because I was the CEO of two companies, right? I led them. So I have an entrepreneurial character, I guess you could say. And I’ve always been involved in a lot of aspects of the businesses that I’ve been running, and I like that. I actually like to be involved in marketing and sales and even the financial side of things, and internal and external and all that kind of stuff. So it allows me to continue to embrace that side of me, even though I’m now like corporate. And I really appreciate that Elementor allows me that flexibility. It keeps the work that I do very interesting.

[00:06:33] Nathan Wrigley: Yeah, I’ll bet. And you’ve always me as an extremely curious person, who likes to be busy, let’s put it that way.

Does Elementor continue to grow? I mean, I guess Elementor for me feels like it’s coming up to a decade old or something along those lines.

[00:06:48] Miriam Schwab: Yes, exactly. June is going to be 10 years since it was founded. We’re in our 10th year now, like heading up to it.

[00:06:53] Nathan Wrigley: Yeah, and when it began, the plugin offered an incredibly valuable free version. And this was a real good way of kind of promoting growth because page builders were becoming all the rage at that point, and the market was already becoming saturated. But Elementor managed to put some daylight between themselves and the competitors because of the fully rounded free version. And then obviously, you know, built the pro version on top of that.

And then when you chart the numbers of WordPress’s growth over the last decade, I don’t know what the exact numbers are, perhaps you can fill us in on that, but it feels like quite a bit of the WordPress growth, so WordPress generally, that 43% figure that we constantly like to refer to, it feels like quite a bit of that belongs, well, not belongs, but you know what I mean, is because of the popularity of Elementor. So the growth was meteoric. And I don’t know how that’s carried on, whether the line continues that trajectory, or whether it’s kind of flat lined, but still growing. Can we just get into that a bit?

[00:07:50] Miriam Schwab: After Elementor was launched, it really quickly reached a million active installs, I think within a year or something because it was the right, very much the right product at the right time, bringing a lot of value to users. And like you mentioned, like with the free version, for sure, that definitely helped power its growth and adoption. But also because it was a valuable, and is a valuable product for web builders to use.

In terms of the growth, so amazingly, Elementor’s adoption continues to grow. So you know W3Techs? Every year they publish like top stuff. And the various categories is based on the absolute number of sites that that particular technology accrued over the course of 2025. For the third year running, Elementor was given the title of Top Content Management System by W3Techs, beating out Wix and Shopify and WordPress. WordPress was the winner for many years.

And I’m not privy to the exact numbers or their exact calculations, but based on what they say, if Elementor is the top CMS, it’s because over the course of 2025, it’s usage base grew by more websites than those other platforms grew, which is wild. And like there, they even states, I think Elementor started off 2025 with 11.7% of the internet, and ended off 2025 with 13.1% of the internet.

[00:09:12] Nathan Wrigley: Wait. Do that again.

[00:09:13] Miriam Schwab: Of the entire web, yes, of the web. Not WordPress.

[00:09:17] Nathan Wrigley: We use the 43% figure to be WordPress. So that same kind of sentence, but substitute Elementor for the word WordPress in there. 13%?

[00:09:27] Miriam Schwab: Yeah, over 13%. So over the course of 2025, Elementor continued to grow at a very significant pace. So Elementor is still growing.

[00:09:36] Nathan Wrigley: Yeah, that’s remarkable. Do you know if that’s because of new adoption or is that that the current users of the platform, you know, who have a license for multiple sites and what have you, are just using it more and more, so that number kind of creeps up? So I guess the question there is, do you have a growing user base or is it just more broad use by your current user base?

[00:09:54] Miriam Schwab: Our current user base definitely contributes to growth, but by our estimation, we grew by something like three and a half million sites over the course of 2025. So a significant number of those sites are new. New users, or new sites, or whatever, that kind of thing.

[00:10:12] Nathan Wrigley: What a phenomenal story that, I mean it is the story really inside of the WordPress space over the last decade. I presume there’s nothing that can touch those kind of stats. That’s really remarkable. And I suppose it’s a blessing and also a curse. And what I mean by that is, you’ve got this giant platform, enormous user base, but I suppose it also means that anything that you do with the Elementor platform, there’s a lot of care that needs to be taken on every single update. So if you’re going to update to the 4.0 version, which you recently did, big overhaul, but a big user base to get annoyed if things go wrong. So I guess a lot of care and attention required because of your popularity.

[00:10:52] Miriam Schwab: Yeah. Backwards compatibility, you know, it’s a very important aspect of WordPress development in general. So at Elementor it’s the same. It’s super critical in terms of every update that we push out, determining how far back we’re going to support things, et cetera. So every version that goes out, it has to be super duper QA’d.

And even so, you know what WordPress is like, it’s like the wild west. Every site is like a snowflake. It has its own combination of themes, plugins, server configuration, PHP version, you know what I mean? Like, good luck with that. But the team does an amazing job of really managing to make sure that pretty much every version doesn’t cause issues.

And with regards to version four, that is an overhaul. And the team, while moving ahead with creating version four, because it’ll bring a lot of value to our users and it pulls Elementor even more towards the future, it’s being done so carefully. Like every step is considered.

And in terms of how it’s being developed and also how it’s going to work in conjunction with the previous approach to Elementor. Because the assumption is, not everyone’s going to just jump over to version four. Migrating an existing site to version four may be complicated. There’s discussions around how to do that in the best way, but it may be complicated.

So even an existing site, it might have pages still built on version three and prior, while new pages will be built with version four to gain the benefits of version four. So they have to live side by side for some amount of time. It’s incredibly complicated, but it’s a really important, exciting project because, like you said, Elementor is 10 years old, which means a 10-year-old code base. That means it’s time to give it an overhaul, even with all the risk and complication involved.

[00:12:32] Nathan Wrigley: What’s the staff count that you’ve got over there now?

[00:12:34] Miriam Schwab: It’s something like 350 now, I think.

[00:12:37] Nathan Wrigley: So enormous. But again, I presume that speaks to the things that you’ve got to do. You know, if you’re going to ship a big update, you need a lot of bodies doing the coding for that. But also, you know, checking the backwards compatibility and things.

Okay, so you mentioned the future. In the year 2025, the words, well, the letters AI, I think probably were said by more people than just about anything else in the English language. Apart from maybe the word the. Everything else seems to have gone AI. If you look in the WordPress space, all of the media, all of the excitement, all of the interest, all of the everything, the oxygen is being sucked out of the room by AI, let’s put it that way.

And so Elementor, I presume, has to keep up with those current trends. What is happening over there? In terms of the page builder, and also I know there’s other ancillary products and plugins that, whilst built by you, aren’t necessarily part of that page building experience. What’s going on?

[00:13:26] Miriam Schwab: So we have like a number of approaches to AI and each one is getting a lot of attention. And while users are enthusiastic about them, we’re trying to figure out where to give them the best experience and how with AI. I think just like the whole world, we’re also working on figuring out AI and how to implement the best way possible.

But we strongly believe that WordPress must have strong AI capabilities in order to secure its future. And we want to see WordPress have a strong future for many years to come. But not only that, AI in WordPress will actually give it a big leg up and advantage over the more proprietary platforms in our opinion. So there’s like intense effort being made on our side to create amazing AI tooling in order to try to secure WordPress’s feature.

So what do we have? We have the kind of basic obvious stuff, which is, you’re in Elementor editor, you want to generate an image, here, you can generate an image with AI. You want to create copy, text, add title, all that kind of stuff, the content release stuff, there you go. HTML, CSS, okay. Kind of like check the box, pretty expected stuff.

Then we have Angie which is its own standalone plugin, which applies to all of WordPress, not just Elementor. And it gives agentic capabilities to WordPress in general. So whatever you would think an AI assistant would be able to do for you in your WordPress environment, it can pretty much do that. As we know, AI is non-deterministic. You tell it to do something and it will do it one way, and it tells you to do something else and it’ll do it the other way. So there’s a lot of guardrails being developed by our team to direct Angie to give the results that the user probably wants.

But you can do things like around managing your site, managing users, creating content, changing categories, WooCommerce management, product management, things like that. So it streamlines a lot of stuff. And the team is working on some like really exciting capabilities, which hopefully will be released soon. And then we’ll be able to talk about it. But in the meantime, it’s free to use. Anyone can go into the repo, install it, start using it, and of course we want to hear feedback about that. So that’s Angie.

And then the third approach with AI is our Site Planner. So Site Planner is a very cool tool. You chat to it, and you chat your way to a very robust website. It’s not like a one line prompt, build me a flower shop website, and then it just kind of guesses what you might want. It asks you very specific and useful questions so that it can get the fullest picture of what you want to build. And it guides you along the way by asking you the right questions, and then it generates a site for you in Elementor, and then you can export that site into your own hosting environment, or into our hosting environment, or you can download it as a zip file or whatever.

So the Site Planner can be the first stage in your AI website building journey. It takes you, let’s say from zero to 80, I would say. And then the last 20% of the site you can do with Angie, you can do manually, you can do with Elementor, you could do not with Elementor. You could do however you want, and launch your site.

So that’s pretty much what’s going on with AI. We have some more stuff coming out soon, but we’ll talk about it when we can, you know?

[00:16:39] Nathan Wrigley: Yeah, so let me just reprise that and make sure I’ve understood. So the first piece of that jigsaw puzzle was the things with inside of Elementor, the page builder itself. And if we rewound the clock just five years, even though you are now saying, oh, the expected stuff, that stuff was the stuff of Star Trek. You know, oh, ask it for an image and it’ll make an image. Yeah, okay, that seems like a distant future that we’ll never reach. And yet now that’s kind of de rigueur. Everybody expects that stuff. So that’s the stuff inside of the page builder as you’d expect. Text, images and what have you.

But then you’ve got this Angie product, which if I’m parsing that correctly, enables you to leverage the abilities of WordPress. So I would like you to create me a post and assign these categories to it and publish it on this particular date. Do you know, does that leverage the sort of Abilities API, or have you done the foundational work yourselves as opposed to the Core Abilities API, which the teams have been working on more recently?

[00:17:32] Miriam Schwab: That’s a very good question. So the Angie team started building Angie at least a year ago, which was before the Abilities API was, I don’t know, even a twinkle in someones eye. So the team, in order to make Angie work, built their own tools, like it’s called tools, exposing WordPress’ capabilities, so actually invested a ton in exposing something like 200 tools in WordPress to Angie.

And not only in Angie, the team actually also did some work to expose tools in WooCommerce and ACF as like a starting point for, because they’re very popular plugins. So when the Abilities API came around, the team already had a lot that was done and also more than what the Abilities API has. The goal is to sync up with the Abilities API and leverage it. At the moment there’s still some issues, at least on our end using it. But we’re in constant communication with the WordPress AI teams and giving feedback and things like that. So hopefully we’ll be able to resolve that. But the team did build all of that themselves, which is pretty amazing.

[00:18:33] Nathan Wrigley: Yeah, that is amazing. But I guess given what Elementor touches, and the fact that you’ve got that enormous staff there that can try to build that, it kind of made sense to go at that because you didn’t know that the team were working on the Abilities API. I remember when that dropped. So I’ve been covering WordPress news for, I don’t know, a decade or something like that. I remember thinking, this is probably the most consequential piece of news that I’ve ever covered.

The idea of binding AI, so AI in the scenario that you described first, where I’m in Elementor, I want an image, I’m in Elementor, I want some text. That’s interesting, but it’s really confined. There’s lots of boxes around that, there’s only a certain permutation of things that you can do. But the minute you start to bind an AI to all the things that WordPress can do, so create a user, delete a user, create a page, schedule it for this time. Oh, I don’t know. Just imagine clicking around inside the WordPress admin interface and wherever you end up, there’s a thing, there’s a box or a field to fill out, and there’s an ability that can be done by that thing.

And uncovering those and making it so that the AI agent can understand what that is, it’s very hard to me to encapsulate in English why I think that’s so impressive. But, do you know what I mean? Can you see the thread of what I’m trying to say there, how important that is?

[00:19:49] Miriam Schwab: It is super impressive. What you’re saying about AI being like, if we had thought about it five years ago, it would be amazing. Even agentic AI, everyone talks about agents as if it’s like obvious that AI can be agentic. But for quite a bit of time in the beginning, and of course AI years are like a million years.

So for the first year and a half or something of us, of the world using AI, something like that, there wasn’t even agentic AI. So the fact that there’s agentic AI, and that it can be applied to WordPress is pretty wild. And I’ll tell you why I also think it’s wild. Because WordPress is 20 years old, 23 years old, whatever it is at this point. It’s a legacy platform. And the fact that it can be, the Abilities API is a really smart approach to it because it kind of slackens the exposure of everything that’s going on behind WordPress.

Instead of needing like a million different APIs and different approaches and that every plugin’s creating their own way to integrate with WordPress. It standardises it in a way that actually is also good for development, but also is good for the AI future, because it means that the way AI will interface with WordPress is straightforward and standard enough that regardless of what the LLMs end up doing going forward, if there’s more capabilities, if we move beyond agentic to who knows what, I don’t know what, like implants in our brains, then we can still interface with WordPress even though it’s over 20 years old. I think that’s amazing. So kudos to the team for accomplishing that.

[00:21:12] Nathan Wrigley: Yeah, and it was a penny that just didn’t drop for me. I saw AI as this interesting thing where it could create content, but I didn’t make the connection with all of the myriad things that WordPress could do. I just thought, okay, it’ll be inside the block editor, or it’ll be inside of Elementor or wherever it lives. That’s where the AI will belong. And then when I first read a summary of what the Abilities API was going to achieve, and that then binds to a whole other layer of APIs surrounding that, that was when the penny dropped. And I thought, okay, so basically what you’ve done there is you’ve opened the door and you’ve said, here’s all the things that a WordPress website can do. Until now, the AI was just in the interface and you could, you know, words and images and what have you. Now it’s, okay, we’ve opened the doors, now you can do the whole range of things that you’d like to do with the AI. And so it opens up all of those capabilities.

Again, the words fail me, but the import of it, the importance of it, I think only time will tell. But it feels like it’s probably one of the more important things. And again, like you said, 20 year old legacy system, there could have been a moment where the two paths split and, you know, AI became more interesting outside of WordPress, and WordPress sudden decline and what have you. But by opening the doors and saying, actually let’s just let all of the AI in and you can explore that if you wish to, kind of interesting for those people that want to.

[00:22:29] Miriam Schwab: Yeah, WordPress could have struggled to keep up with AI and we, I’m not saying it’s not a struggle. I think we’re a bit late to the AI game in general as a, like an ecosystem, which is understandable. We’re open source, things like, you know, move slower and by committee, so that’s just how it goes. But the fact that it picked itself up, the team was established and within six months of being established or something like that, it already had accomplished their four pillars of work that they had set out for themselves, is really amazing. And it’s super important for the future of WordPress, and it’s great to see that they were able to accomplish that.

[00:23:01] Nathan Wrigley: Yeah, and also it feels like the moment has arrived where a lot of the people that write in the WordPress space, so content creators and what have you, they’ve sort of got this now and they’re now creating content. And some of the people that no doubt, you and I both follow, who may be writing on their own blog or on social media or what have you, there seems to have been over the last couple of months, a real, a sense that, okay, WordPress has a place. We don’t need to worry about that anymore. We’ve now got a path with this AI team. State of the Word AI team sort gathered there and pressed the big red button and all that kind of stuff. There seems to have been so much focus that it’s built itself into relevance again. And there seems to be less worry. People now writing content about the things that WordPress can do with these abilities bound to it.

Okay, so let’s move on to the third strand of what you mentioned, which was, forgive me if I get this wrong. You said Site Planner, right? Was that correct?

[00:23:51] Miriam Schwab: Site planner, yeah.

[00:23:51] Nathan Wrigley: Okay. And this is a conversation that you have. This is something that I always find kind of interesting. You know, getting into a conversation with something which is essentially an AI, that’s kind of curious. But you mentioned that whole process was a back and forth. So it’s not, I want a website. You write this great, big, long prompt, click the button, wait for five minutes and kind of cross your fingers and hope for the best. This is more of a dialogue, is it? Where you repeatedly get asked a series of smaller questions, which hopefully then build up to the sort of final thing. Tell us a little bit more.

[00:24:19] Miriam Schwab: The way that works is, yes, you can like paste in some kind of giant prompt, but even then, whether you start with like one sentence or a prompt, the Site Planner will ask you questions, clarifying questions, and also suggestive questions. Do you want any of these types of pages, or content on your site? And things like that. It’s really good for guiding non-professionals.

It then creates actually a brief, like it writes out a brief for your website based on this conversation. So it’s really good for non-professionals who don’t know the questions to ask, right? They don’t know what they should be even considering. So it asks questions for them.

But it’s also great for professionals. It cuts down a lot of time in the research stage of building a site for a client. And the AI can actually like give them added value in terms of asking questions they might not have thought of. That conversation process can be cut short whenever the user wants. Like it can go on for as long as they want, but they can also say, that’s enough, let’s build the site based on what we discussed.

When the AI is creating the brief, it will tell you the strength of the brief, like in the top right corner of Site Planner. So once you get to the stage where it says strong brief, you could stop there, but you could keep going. The more information you give it, the better the outcome will be. Just like with AI in general.

So once you finish that conversation, you say, let’s move to the next stage. It creates a site map. So it shows the hierarchy of the site based on the conversation. But also, on each page it shows you like what type of content it will have, and it actually gives you the content. The nice thing about the way the AI builds the site is that, first of all, content is king, which it always has been on websites. And here you really get to see the strength of that.

So the content that it creates is really good because it’s based on this conversation that you had with it. At that site map stage you can drag and drop things around, add different pages that it didn’t include, remove sections from pages, but it’s just content chunks. Once you are happy with that stage, you go to the next stage, which is the website stage.

We call it the wireframe stage because it’s not like a fully designed site, it’s more like the site with a structure, but it’s like 80% of the way there. It could have like testimonials section, contact page form, gallery. I created a demo site for a cafe that also has community events. So it has an events page. It’s really great. It like really gets you 80% of the way there. It saves a lot of time, and if you’re a professional, you can show your client that at that stage and work off of that.

If anyone works with clients, you know that communicating and getting to a point where everyone’s happy can be very challenging when you don’t have something to show. But here, when you have something to show, you get past that blank canvas stage, which is very hard for people to like even imagine anything, and then you can work from there.

So once you have the wire frames done, also at that wire frame stage, by the way, you can also there edit things, remove things, et cetera. Change the colour palette, upload a logo to be used across the site, and things like that.

Then you can publish the site. You can publish it into our own hosting environment in one click, but you can also publish it into any existing Elementor account. You just kind of add your It hooks up to your account and it will send it to wherever that site is hosted. Or you can download and upload it somewhere else. So it gets you from zero to 80, or 90% of the way with a site through a conversation and then you can like tweak it, polish it, and finalise it from there. So it’s really useful in my opinion.

[00:27:34] Nathan Wrigley: So prior to the publication of it, you mentioned that you could show it to your clients. Is the capacity to have a public facing yet not public facing, if you know what I mean? So a URL, which is visible to you, but not the wider world for SEO and what have you. Is it possible to sort of take an up and running version, give it to a client so that they can have a little bit of a poke around?

[00:27:52] Miriam Schwab: That’s an interesting question. At that point, when it’s still in site planner, I think it’s only visible to you. So you would have to download and put it somewhere. But then you could put it in like a staging site or wherever you want to put it. You could share your screen with the client. But yeah, I think you can’t share the URL to that particular project within a third party.

[00:28:09] Nathan Wrigley: Because that’s all so very instantaneous compared to, you know, it may be that you get into this dialogue, which takes half an hour or what have you. But, you know, compared to the old way of building things, it’s more or less instantaneous. I had this notion that a service like that would be kind of interesting if it spat out four or five variations of the same thing at the same time.

So I realise there’s a sort of, you know, a burden in terms of technology and overhead and all of that, cPU time and what have you. But the idea of entering that client meeting with five versions of the same basic thing. So, okay, I have a, I don’t know, a bricks and mortar store and I sell widgets. So I build the one version, but Elementor, for example, in this scenario, builds me four others as well. And then I can go through and cherry pick which bits of this one do you like or, I like the colour scheme of that, if you know what I mean? So I don’t know if it’s possible to do that. In other words, is it possible to spin up multiple versions of the same thing, or would you have to sort of start from scratch each time?

[00:29:02] Miriam Schwab: Well, what you can do within the project is you can ask it to regenerate a section. Like, you could be like, I don’t like that. So you could say like, do it differently and then it will suggest something else. You could theoretically, I guess like create a second page, let’s say a second about page and then be like, okay, suggest a different format for me for this. Like a different structure because it’s not exactly design.

The design afterwards can be worked on with a client or whatever. It’s more like the structure of the page. So, and even though it’s a live website, it is like structure. So you could be like, no, I want the testimonials to be a different style, like suggest something else. And then you could show, I guess the customer various pages within the site. It’s an interesting idea.

I know that the team is working on Site Planner, and it will be more rolled into the WordPress Elementor environment, like in the site. Right now, it’s an external tool, so once it’s in there, you could like duplicate pages, you could regenerate pages, things like that. I don’t have an ETA on when that’s happening but that’s the general direction.

[00:29:58] Nathan Wrigley: It’s a fascinating conversation. And the other thing I was going to say is that there’s obviously, in your community, there’s a lot of people who are, they have a lot of history of building things with Elementor. They’ve learned the UI inside and out. They’ve become experts. They know where every checkbox is. If you show them something in Figma and then give them 10 minutes, they’ll mimic that perfectly.

And I was wondering what kind of capabilities the AI has within the Elementor interface. So in other words, does it do a sort of cookie cutter job of creating rows and sections and what have you, based upon things that have already been prebuilt? Or can it, in its AI wire frame stage, can it do, I don’t know, unique padding or crazy CSS things? Can it get into all of the bits and pieces that Elementor offers, all of those rich experiences? Or is it more kind of, okay, somebody in the team built these testimonial elements, and so we’re going to basically pick one of those and mimic that onto this wire frame?

[00:30:52] Miriam Schwab: So there’s Angie as it exists now and there’s Angies that will be over the next few months. So I’m going to be giving like, I guess kind of a sneak preview into what’s coming.

At the moment it can do a lot within Elementor, but it is kind of constrained within what Elementor classically does. However, v4 is going to enable the AI, or let’s say the v4 AI partnership will give it a lot more creative capabilities, and extend basically to whatever you want it to do, that you can do with Angie. From custom code snippets, to custom widgets, to custom anything, or working within the Elementor. There’s going to be a lot of interesting things that people can do, basically in some ways just limited by their creative ideas.

[00:31:38] Nathan Wrigley: Right, so much more granular. Because I have this vision that at some point in the future, and maybe some tool out there can do this. But I have this vision that, let’s say you’re inside Elementor, you’ve got 99% of the way there, and then you suddenly realise that that picture of the cat is the wrong cat. And so you just sort of say, can we have a different cat? No, not that one. Okay, that cat. And now, could we make the borders rounder on that particular image and give it a bit of a box shadow? That would be nice. Oh, and then swap it around so the text is on the other side.

So you end up in this sort of dialogue, almost like you’re chatting to the designer and you are watching over their shoulder as they build it. That seems to be the kind of place where we would love to get to. And then of course we enter this curious, difficult moment for the web developers and web designers, where these things become so dreadfully straightforward and easy that we then have to start questioning, how do you offer this as a client service if almost everybody can speak to a website and get it to do whatever you would like it to do?

So there’s two things in there. I sort of smuggled the last one in. But the first one is this sort of dialogue with the website, and then the second one is whether or not these tools are making it more difficult to be in the industry that we all love so much.

[00:32:44] Miriam Schwab: With regards to the dialogue description, so AI in general, it’s become like our team members for certain things. You know, in the past, if I wanted to, let’s say, post something on social media on behalf of Elementor, I would go to the design team and ask them to design something and go through that process, and then get what they gave me and then post it.

And now my design team is Nano Banana, right? Or ChatGPT. I get two versions, they compete against each other. Whichever one works best for me is what I take. And so it’s like that in almost everything we do if we’re using AI, and also in the world of website creation, design, development, and management. Our team for many aspects of it will become AI, which will create greater efficiency and also it seems like greater redundancy.

But I think like anything with a tool, the results will be defined by the abilities and skills of the person directing it. AI needs direction and that conversation to happen. And the quality of the results will be dependent on the skillset of the user, not skillset in terms of, how do I get AI to do stuff? But skillset of, what should an excellent website be?

So the human in the loop, I think, will continue to be a very important part. It’s just we’re kicking the can down the road kind of thing in terms of where we bring value. So instead of us bringing value, sitting and clicking on an image and changing the background using Photoshop, our value will be in being the director and producer of what output we want it to have and getting it there.

So we’re not like the tinkerers and like the hammer or nail people, we’re the contractor or the strategist, or the blueprint creators. So the conversations will take place, but the human in the loop part will still be really important. We all, whether we’re web creators or whatever profession we have, have to figure out how to use AI to help us do what we’re doing better.

I actually don’t exactly agree with the idea that AI helps us do more. It does, a lot of things become faster, but I think the value is that a lot of things get done better and a higher quality. So there’s that.

So how do we use AI to the best of our abilities to help us do our work even better. But also, how do we continue to provide value in a world that’s dominated by AI? And that’s like, I don’t know what the answers are. 2026 is going to be the year of, who knows? Because AI changes every second. They could invent a new model or release a new model tomorrow that changes this whole conversation.

[00:35:19] Nathan Wrigley: One of the things that, sort of round it off with this one maybe. I’m thinking about your support system. So your support system, which has built up over 10 years, and presumably worked to help your clients. But the clients themselves probably started with a blank canvas and then, you know, made mistakes along the way, but were able to describe the mistakes that they made along the way to your support agents. Like, okay, I was doing this and then something went wrong. Can you unpick that for me? Where did I go wrong?

The realignment that needs to take place for your support team with the advent of AI, I suppose is fairly profound because you’ll probably have a lot of users who will just know that they did a prompt. And now this thing happened and I didn’t want that thing to happen. How do I go back? I mean there’s obviously a case of click undo, or don’t save the changes or whatever. But I’m also imagining the support that you provide has to pivot a lot as well.

[00:36:06] Miriam Schwab: The development of AI has to include implementing a ton of guardrails to prevent users from doing things that can damage their site. I’ll give you an example. In the early days of Angie, I generally love pushing tools to their limits, so I was pushing Angie to its limit and I got it to kick me out of my own site and remove my user. And I was like, well that’s not cool. So guardrails were put in place around that.

There’s concepts of, first of all, not allowing a user to implement something until they’re sure. Like meaning, not just be like, hey AI, can you do this? And the AI is like, done. No. AI is like, okay, here’s what it will look like, here’s what it will do, that kind of thing. Do you want to go ahead? Then, of course, undo buttons.

But then even with all of that, yeah, support could be different. We can have logs and things like that about what happened and why it happened and things like that. And that can help troubleshoot things. But on the other hand, I think support can become much better because while still the human is in the loop, getting the answers to issues, you know, you end up with this huge knowledge base of stuff, and a support agent can only know so much. Our brains are finite. We don’t know what someone discovered yesterday as an issue and as a solution and all that kind of stuff. So once you have AI kind of like analysing your corpus of knowledge and then pulling out what can help you or the user, then that becomes much better.

But I’ll give you another interesting angle of where support becomes complicated, where you have the Abilities API, or Angie or both, and a user is interacting with Angie, let’s say, and Elementor and also a third party plugin that is also working with AI. And then there’s something borks. And then the user is like, was it Elementor that was a problem? Was it WordPress that was a problem? Was it this third party plugin that was a problem? How do I even ask for support when I don’t know what just caused that issue? I guess it’s another version of plugin conflicts, but it’s just taking it to another level.

[00:38:01] Nathan Wrigley: Yeah, it’s a different level altogether. I said that was my final question. Turns out I was being insincere. I have one more. And that is, I know that Elementor has done a lot of work trying to make websites more accessible in the recent past. And so that’s the final question, I promise. The idea that we hand over the agency of this to our AI builder and so things like accessibility, there may be other things that we could smuggle in there as well, but we’ll go with accessibility. Is work being done to ensure that the output follows WCAG guidelines and things along those lines?

[00:38:32] Miriam Schwab: So something that I’ve seen as we’re working on implementing AI capabilities is that there’s this idea that I had, and I think others do as well, that AI is the all powerful whatever. And if you just say to it, make my site accessible, it will make my site accessible. So it turns out it’s not like that. Maybe one day it will be. But AI needs tools, needs direction, needs the capabilities, and it doesn’t just have it just because it’s AI.

So it’s kind of the same with accessibility. So actually Elementor created an accessibility plugin called Ally, which is very useful because it goes through the website, analyses it, tells you where there’s issues, but doesn’t just tell you where there is issues, it gives you solutions. And it can, with AI, implement those solutions. Like you can click and be like, fix that.

And that kind of ongoing accessibility assistance means that a person can, first of all do like a once through of their site and bring it to a better level of accessibility. We don’t guarantee like 100% accessibility, I don’t know if anyone can, but it improves accessibility on many levels.

And then as you’re going along, it can keep an eye on things or you can like trigger it and it can help make sure that whatever you’re doing going forward also stays within accessibility guidelines. So it’s like a really useful assistant, slash solution, slash tool that users can implement on an ongoing basis with their site. So worth checking out.

[00:39:59] Nathan Wrigley: Yeah, I’ll put that into the show notes as well. And if memory serves, I could be wrong about this, if memory serves it’s not an Elementor specific solution.

[00:40:06] Miriam Schwab: Oh, right, exactly. It’s for all sites, all WordPress sites.

[00:40:08] Nathan Wrigley: So any WordPress website can benefit from that. Okay, that’s interesting. Okay, so who knows what the year 2026 will bring. It no doubt will be chaos, turmoil.

[00:40:18] Miriam Schwab: It won’t be boring.

[00:40:19] Nathan Wrigley: That’s the words. Yeah, it definitely won’t be boring. And, yeah, good luck for Elementor in the year 2026.

[00:40:26] Miriam Schwab: Thank you. I’m excited, by the way, for 2026. It’s like a whole new world, and I think in some ways, at least for me, and I see this with others, it like reignited a spark for innovation in WordPress. WordPress is amazing, it’s been amazing for over 20 years and we’ve always, you know, been seeking, looking for ways to innovate. But AI takes it to a whole other level and makes innovation more accessible, if we’re going to use that word. And it’s really fun to see what people are creating around AI for WordPress. And I think we’re going to see some amazing things released over the course of 2026, not just by Elementor, that will make WordPress, wow. Anyways, that’s just my thoughts about 2026.

[00:41:00] Nathan Wrigley: Miriam Schwab, thank you for chatting to me today.

[00:41:02] Miriam Schwab: Thank you.