Reinventing the Editorial Experience: Mars + Drupal Canvas at Enterprise Scale

The company behind M&M's and Pedigree runs more than 400 websites. Here's how it rebuilt page editing so its brand teams say the new tools feel like Figma, not a CMS.

Boyan Borisov and Anthony Bojocan's session from DrupalCon Rotterdam 2026, recorded Wednesday 30 September 2026, Success Stories & Innovation track.

 

About this session

Mars is one of the world's largest family-owned companies — 170,000 associates, three business segments, and billion-dollar brands from M&M's to Pedigree. Anthony Bojocan, Global Web Delivery Senior Lead at Mars, opens with the wake-up call: when GDPR arrived in 2018, Mars confronted a fragmented digital estate of overlapping CMSes, duplicated technology, and expensive agency dependence. The transformation program that followed consolidated the company onto Drupal — reaching 100% compliance, serving 54 countries with over 400 websites, cutting total cost of ownership by 40%, and earning the trust of around 70 global brands on what Mars calls its Mars Experiences Platform.

With the foundation in place, the question became which page builder could deliver a best-in-class editorial experience. Boyan Borisov, Director of Digital Solutions at JAKALA and a 20-year Drupal veteran, walks through the bake-off between Layout Builder, Acquia Site Studio, and Drupal Canvas — and why Mars made a strategic bet on Canvas in October 2025, when it was still an unstable dev branch called Experience Builder. The pilot brand, API Fishcare, wasn't a toy project: a full enterprise platform with PIM, DAM, OneTrust, and Salesforce integrations, built on Drupal CMS 2.0 as a starter kit and delivered in about seven months.

The second half shows what changed. A live demo follows a first-time editor updating a homepage unaided; brand teams told Mars that editing in Canvas "feels like Figma." Boyan explains how a machine-readable Mars Agentic Design System let AI agents build components within Mars's established design language, collapsing a 200+ component library to around 80. Anthony closes with the numbers — 40% faster component development, roughly 80% less demand for new components, 30% faster site assembly, around 15% lower TCO — and Boyan with a message for every organization with a mature Drupal estate: don't start from scratch; modernize what you have.

Key takeaways

  • Mars consolidated 400+ websites across 54 countries onto Drupal after GDPR exposed a fragmented, risky digital estate — reaching 100% compliance and cutting TCO by 40%.
  • Canvas won the page builder evaluation over Layout Builder and Site Studio on editorial experience, no-code capability, open source governance by long-standing core contributors, and alignment with the Drupal AI Initiative.
  • Mars committed to Canvas in October 2025 while it was still a pre-beta dev branch — de-risked by talking directly to the Canvas team and building proofs of concept — and launched four API Fishcare sites in about seven months, using Drupal CMS 2.0 as the starter kit.
  • A machine-readable design system (Astro Starlight with manifest.json and skills files) keeps AI agents inside Mars's design language and stays in sync with Drupal via a CLI command; the component library shrank from 200+ to around 80.
  • Reported results: 40% faster component development, ~80% reduction in demand for new components, 30% faster website assembly, ~15% TCO reduction — and editor changes that took a two-week sprint now take minutes without code.

Chapters

  • 0:00 Intro
  • 0:58 Speaker introductions
  • 1:47 Mars at enterprise scale
  • 4:05 A fragmented digital landscape
  • 5:01 From chaos to control: consolidating on Drupal
  • 10:48 Choosing a page builder: Layout Builder vs Site Studio vs Drupal Canvas
  • 16:26 Finding an MVP: API Fishcare
  • 17:46 An enterprise platform, not a static migration
  • 20:47 Launch: four websites in seven months
  • 22:19 Demo: a first-time editor in Drupal Canvas
  • 25:36 What changed for editors, developers, and agents
  • 33:23 Results and KPIs
  • 35:52 Contributing back to the community
  • 39:26 Why Mars stays with Drupal
  • 40:22 Q&A

Transcript

This transcript has been lightly edited for readability (names and Drupal terms corrected, filler trimmed). Statistics and claims are reproduced exactly as spoken. Passages marked [unclear] or [verify] are pending review against the recording.

0:00 Intro

Anthony Bojocan [0:03] Okay, well, good afternoon everyone. Do you still have some energy? Because I don't. I am going to pretend that I have energy. With my friend, we went for a coffee — it's like, yes, one more hour. So are we still okay to take in some more information?

[Unidentified] [0:21] Yeah?

Boyan Borisov [0:22] Yeah.

Anthony Bojocan [0:22] Okay, so we will start. I'm not sure how many of you were in this session and the keynote this morning. Okay, so I guess we don't have to do this anymore if you were there. No, I'm just kidding. So today, as I mentioned earlier, we're going to talk more about our journey at Mars using Drupal Canvas. And this time I have my partner in crime with me, Boyan.

Boyan Borisov [0:46] Hello.

Anthony Bojocan [0:47] From JAKALA, who will explain more of that journey on the technical side. And at the end, you'll have more time to ask if you have any questions.

0:58 Speaker introductions

Boyan Borisov [0:58] Okay. Yes, hello, my name is Boyan. I'm from JAKALA. I'm a Director of Digital Solutions at JAKALA, and I'm having my 20th anniversary this year with Drupal. I have seen many page builders over these years. I have been speaking mainly about best editorial experience — the title of the presentation never changed, but the tool has been different. I think now we are really having something which is very impressive. Just a few words about JAKALA: JAKALA is an AI and data company that builds digital experiences for many big enterprises around the world. And we are partnering with Mars already several years. Tony?

1:47 Mars at enterprise scale

Anthony Bojocan [1:47] Thank you. My name is Anthony Bojocan. I am a Global Web Delivery Senior Lead at Mars, under the Digital Experience team. In terms of background about Mars, if you're not familiar with the company yet: Mars is a private family-owned company that was founded in 1911 by Frank C. Mars. Since then, it has been operating in over 78 markets worldwide with about 170,000 associates or employees. You may know them from some of the billion-dollar brands like M&M's, Snickers, Pedigree — you will see them on this slide. And the thing I wanted to show here is I want to really explain the scale of a huge enterprise like this. Here we have a lot of brands at Mars. We have 3 segments. You have Mars Petcare, where you have your pet nutrition brands; you also have the growing veterinary health brands and digital science and diagnostics. There's also a second segment, which is Mars Snacking. You have Tru Fru, KIND, but also the bigger brands like Pringles, Cheez-Its, which is thanks to the Kellanova acquisition of the company recently. And then finally, there is Mars Food and Nutrition, where you have Ben's Original, Dolmio, et cetera. Now — I am part of Mars Global Services. My team, our function is to support all of these segments across the globe, and sometimes at the group or even the regional level. So we have a lot — we're working with different brands, different stakeholders, different priorities. So the challenge for us is: how do we manage this big landscape? In terms of that journey, it was actually in 2018 when GDPR was put into regulation. You're probably familiar with GDPR, I suppose. At the time when this came out, it forced Mars to face the reality that we have some risk on our digital landscape.

Boyan Borisov [4:04] Yeah.

4:05 A fragmented digital landscape

Anthony Bojocan [4:05] So, we realized we had a fragmented brand ecosystem. Our paid properties are not connecting well to our own properties. Therefore, there is kind of a vacuum in terms of the continuation of this omnichannel journey of our consumers.

Boyan Borisov [4:26] Right.

Anthony Bojocan [4:27] So, there is also a risk on multiple aspects of our digital properties. At the time, we had a lot of overlapping CMSes — WordPress, .NET, Joomla, PHP, you name it, I think we had it at some point. And as you can imagine, that creates a lot of technical debt, duplicative technologies. And it also becomes quite expensive in a way, because there is really an overreliance on external agencies.

[Unidentified] [5:00] Yeah.

5:01 From chaos to control: consolidating on Drupal

Anthony Bojocan [5:01] So whenever we have to create a new website, it really is a significant cost to consider. And at the same time, there was an agency that one of the brands was working with directly, charging Mars like 5 times more than the other agency that we have whenever we ask them to do some remediation changes. One other factor that we need to consider is compliance. As we all know, that's very important. 100% of Mars's websites have to be compliant. And we realized that at this point, because of all of this chaos, we had the lack of ability to ensure data and security compliance and regulation standards. So we knew we really had to change something. And all of these technical challenges, digital challenges — it also translates into underserving and underutilizing our world-class brands. For example, we have very known brands like Skittles in the US that are not investing the same on their websites in the UK or China, for example. And then there is Dove, which has a very strong TV media presence; however, it doesn't extend or reflect to their digital property. So, as a consumer, if you see the ads — oh, this is nice — and you look it up online and you see a website that is very low quality, it makes you question, you know, what is the brand? It kind of dilutes your perception of the brand. So for us, it's very important that we have this singular perception of our brands. We really took that seriously. And in about 2020, 2022, we decided to implement a massive enterprise and data privacy [transformation — verify wording, 5:01]. And this is actually where we have implemented a new transformation program and where we have selected Drupal as our CMS. We had tremendous success during this transformation program. We were able to reach 100% compliance on the websites. We were able to serve 54 countries with that, with over 400 websites now on our digital platform. And one of the most tangible numbers I can share with you is that we were able to reduce our TCO by 40%. We have followed the Web Accelerator Kit approach, and inside of that, we have a good library of components and connectors. And we were able to earn the trust of about 70 of our global brands and have them onboard on the Mars Experiences Platform. By the way, Mars Experiences Platform is our internal name for the standard CMS platform at Mars — so that's Drupal. So here we were able to take control of the situation. That's why this whole journey, we actually call it from chaos to control at scale to cutting edge. And this whole thing really positioned us — we are already very good, so the next step is just to be better, to be the best. So now in 2025 and beyond, we are setting up our new priorities for the future. Here are our strategic priorities — I'll try to keep it simple. The first one is AEO capabilities. I don't think that's a surprise for anybody. Here we know that we want to be searchable not only by humans and search engines, but by the third audience that Dries says: the bots. Secondly, we want to reach autonomous web — a self-optimizing, self-healing solution. And we believe that the first step towards that is Agentic Web. And that's why I'm really happy joining DrupalCon and hearing about what you guys are doing, especially in the Drupal AI area. That is to me very inspiring, and I'm very excited to bring that back to Mars. The third priority that we have is global brand archetypes. A lot of our brand sites are local — one website per brand per market. As I said, there are like 54 markets, so you can imagine the efficiency we could take advantage of if we consolidated them into one. But we are still understanding what it means. And then finally, one of our priorities is best-in-class editorial experience. Here, what I would say is that when we go in this direction, one of the questions that we ask ourselves first is: which page builder do we use? So we ended up with these 3 options in the end. We have Layout Builder, which is the current page builder that we are using. There's also Site Studio and Drupal Canvas. As part of this exploration, we talked to a lot of partners to get their opinions, and Boyan is actually a really good advisor for us on making the decision on which page builder to select.

10:48 Choosing a page builder: Layout Builder vs Site Studio vs Drupal Canvas

Boyan Borisov [10:48] Yeah, and I will be talking a little bit more about how the journey happened. First — apart from these three, we have analyzed many other solutions within Drupal. You know we have more than several, right? But these three were the finalists. And the first one is really Layout Builder, which has been part of the successful story of Mars so far. So we cannot just neglect it and say, okay, let's not use it. We did careful analysis to see what is the current state. Really, once again, it has proved success — it's mature. But what is lacking is the editorial experience: it was given not-positive feedback by the actual editors. Secondly, it has no no-code capabilities, which was one of the main requirements and something that we are trying to improve — to give autonomy to all the editors and site builders so they can achieve functionalities without involving a developer cycle. And the final one was the alignment with the AI initiative — you haven't heard anything like an integration between Layout Builder and Drupal AI, right? And that was one of the main reasons why we decided to go to another one. To make a further selection: the other one was Acquia Site Studio, which actually is already used for particular archetypes within Mars. It was loved by the editors. It really has no-code capabilities, which was something that Mars was looking for, and it was really the main challenger of Drupal Canvas. What was [lacking — verify, ~12:45], compared to Canvas, was the Drupal AI Initiative alignment, and also the other challenge was that it was bound to a license, so it has a lock-in. So then we finally got to Drupal Canvas, which was checking all the checkboxes. It was great — with one caveat. It was October 2025; Drupal Canvas was called Experience Builder back then, and it was a dev branch. I think maybe it was about to have a beta, and it was super unstable. And we had to take a decision. As Tony said, we talked to many partners. But just to give a little bit more light — and I think this is something important when a company is going to dive in and invest in something which is not stable: you need to talk to the Drupal Canvas team to understand the architecture very well. You need to make several proofs of concept to make sure that whatever is coming is going to work for your main use cases. All of this was done, and finally we selected Drupal Canvas as a strategic investment to move forward. The 4 main reasons why Canvas was selected: first, we knew that the Drupal Canvas team put a lot of energy into R&D to research all the competitors — not just to think about what should be the Drupal way of doing things, but to think out of the box about what will be the best editorial experience across all the competitors on the market. This was something very different from any kind of Drupal initiative, which usually starts with "what is the Drupal way?" — and this is something that we believe really makes sense, and we liked. The second topic is really the no-code capabilities. You saw the demonstrations today and you'll see more demonstrations within this presentation, but the site builder and editor can really achieve a lot without a single line of code. They can achieve things in 5 minutes versus 2 weeks of a sprint. The next super important topic: Drupal Canvas is open source, and it's built by Drupal core contributors. Maybe you have heard, but Acquia Source is built on Drupal Canvas. So this also means that one of the biggest contributors in Drupal, which is Acquia, is investing in Canvas. And in fact, we know that the people behind Canvas are Drupal core contributors who have been around 20 years. We know that these people will not go around — it is not just some developers hired to do the job, but people that have been here and will stay here for a while. This is another reason why I believe Drupal Canvas will stay. Finally, the future readiness of Drupal Canvas: we could see that the big investment of the Drupal community is going around Canvas, and this is once again one testimonial that this is the way to go. Finally, the alignment of Drupal Canvas with the Drupal AI Initiative is tremendous. You cannot pretty much see a Drupal AI demo without Drupal Canvas being involved in it — and maybe it's the only page builder with such a deep integration with Drupal AI.

Anthony Bojocan [16:09] Yeah, and I would add that really the last bit was the differentiator for us, because if you remember the strategic priorities I mentioned before, one of them is Agentic Web — and this is really a good fit for our strategy in this case.

Boyan Borisov [16:24] Yep.

16:26 Finding an MVP: API Fishcare

Anthony Bojocan [16:26] Yeah, so there we have it — we decided on which page builder we're going to select. Then what we needed was just the brand that we're going to use for an MVP. And this is how API Fishcare came into the picture. If you're not familiar with this brand, you probably don't have a fish. API Fishcare is a Mars brand that has products used for treatments of your indoor aquarium or your outdoor pond [verify "pond", ~16:50]. Last year, the brand manager from API Fishcare came up to me and said, hey, I have 4 websites from API Fishcare. They were on a technology that was quite a bit problematic, let's say, and the Mars security team was following up with her to address some remediation that had to be done on the websites. So together we discussed how we were going to handle this problem, and ultimately we agreed that we were going to move all of her websites onto the Mars Experiences Platform — to Drupal. And she was also happy to become the first one to use Drupal Canvas. So in the end, we found our guinea pig for this project.

17:46 An enterprise platform, not a static migration

Boyan Borisov [17:46] Yes. Thank you, Jenny — whenever you are hearing this, we very much appreciate your commitment. Jenny is the marketing director of API Fishcare. Just to highlight something, because maybe this is not very visible: this is not a migration of 4 static websites onto Drupal Canvas. This is a real enterprise platform at scale, which has all the enterprise features that you can think about. We can start with the PIM integration, the DAM integration, OneTrust — we talk about Salesforce integration, all these enterprise integrations that you see within the [Canvas? unclear, 17:46] platform. Also the features themselves: the pilot has a customer portal functionality with more than seventy thousand [euros — verify against video, ~18:40]. The customer portal has features where you can book a device and measure the temperature and the salt within your aquarium, or something like this. It really is an enterprise platform on an enterprise scale. And just to continue forward, to give you a sneak peek of the architecture — I will not go deep-dive here, but if you have any questions, you can ask me afterwards. The whole platform was running on Acquia MEO [verify product name, ~19:10]. This is the latest-generation Site Factory provided by Acquia. This was also a kind of innovation coming within the project — we said, okay, we go with Drupal Canvas, let's go with something new. We love to have fun. Finally, something that I also want to call out: the usage of Drupal CMS. Drupal CMS 2.0 was about to be live, and actually, we used it as a starter kit to start the whole journey. We didn't have the need to configure Drupal Canvas on our own from scratch — we used Drupal CMS as a starter kit, which was quite a successful journey. If you ask me if I would do this again, I would do it again. And thank you to the Drupal CMS team for doing this. Then on the left side of the diagram, you will see all the enterprise third-party integrations that you would usually expect in such a platform. We also have an integration with the agentic layer which is available within Mars — we are prepared to further expand these capabilities. And finally, we are also ready to be connected with the omnichannel capabilities and further initiatives within Mars. All this is built on the foundation of the compliance and security policies within Mars. Everything passed a thorough cybersecurity and enterprise architecture validation, which actually took quite some time in the whole process.

20:47 Launch: four websites in seven months

Anthony Bojocan [20:47] So I think at this point, everything was ready. We rolled up our sleeves and got to work. Thanks to our JAKALA team, we delivered the project in about 7 months. We started around October in 2025, and then we went live with the 4 websites in April of this year. It was a really successful website launch. The brand manager was very happy with the outcome of this new website. Of course, one of the main pain points for her was security, so we were able to make it compliant. And one other thing that she mentioned is that the cost now for this website is way lower than what she was having before. So that was really a great win for us. And I understand it was also a big celebration from—

Boyan Borisov [21:36] It was a celebration — I'm talking about the JAKALA team. We were extremely happy with the success that we had, even though I will talk in a bit about the struggles along the journey. It was so successful that today the standard way to approach any new project within JAKALA — which, as you know, is a company that is working with many Drupal enterprise end users — is Drupal Canvas. There is no other way, and we are very proud that we have made this choice, and we are happy to be following this path.

Anthony Bojocan [22:12] Okay, shall we go for the demo?

Boyan Borisov [22:15] Yeah, let's go for a quick demo.

22:19 Demo: a first-time editor in Drupal Canvas

Demo narration [22:19] Charlie, a content editor on the API Fishcare team, has been assigned a task to update the homepage of the API Fishcare website. He has little experience with Drupal — he's just had a few demos — and now this is his first attempt at updating a homepage. He just walks into Drupal and spots an Edit button on the homepage, so he clicks on it. Here you see the Drupal Canvas editorial experience, but he's still unfamiliar with where to start. Luckily, the Tour module is activated, so we get a quick introduction to how everything works. He can see where components can be added and how to view the different layers of the page he is currently editing. He can also see all the pages that can be edited, the templates that can be updated, and how to preview and review changes. The tour also shows how to edit page metadata, such as page titles, SEO metadata, and more. And that's the end of the tour. Now he actually has to add a button to the hero section of the homepage. He clicks on the button and notices a duplicate action, so he clicks on it. The button has now been duplicated. The buttons aren't next to each other, but he remembers the container determines how the components are aligned. He clicks on the container of these elements. He can't remember exactly what to do, but luckily the components tour is here as well. He clicks on it to see more helpful information about how the configuration works. This is exactly what he needs to see: the inline layout that can be changed from here. He clicks on it, and the buttons automatically align next to each other. Now they look well aligned, but he would also like to change the text of this button to reference a new product. He would also like to change how the component looks and the type of the button. There are different options for how the buttons can be styled, small or big. Currently, this is something that he really likes — he will keep it as is. Now Charlie clicks Preview to review how the button looks. He can see it in different resolutions: mobile, desktop, and tablet. Everything looks good. Now Charlie can select the homepage for publishing, and Charlie is done with this.

Anthony Bojocan [24:29] Yeah, I think this is really impressive. In my experience, this really makes a big difference based on how we were working before. And during that time, we also did a demo with other people at Mars, especially those from the brand teams who are looking after their content. And actually, one piece of feedback that we can share with you today: they said that working with Canvas is like you don't work with Drupal. In fact, it's like you're not working with a CMS at all. It feels like Figma. We were really happy to hear this.

Boyan Borisov [25:05] We didn't expect this, honestly. I mean, we thought it's good, but we thought, okay, we are engineers, we don't know. But when she said it — wow. Yeah, amazing.

Anthony Bojocan [25:14] Yeah, because this is really empowering the brand teams. As you can imagine, there is this reliance in terms of workflow — what you can do and what you need to log a ticket for. In this way, by giving them access to Drupal Canvas, they themselves directly have more control over their content.

25:36 What changed for editors, developers, and agents

Boyan Borisov [25:36] Yeah, but let's deep-dive more into what has changed for the different audiences, because here we have multiple perspectives. First, we will start talking about the editorial perspective. Imagine you have a request: I want a new accordion component. The normal lifecycle of such a project using Layout Builder, Paragraphs, any kind of code-driven tool is: you create a ticket, you assign it to the dev team, maybe some grooming is done — in 2 weeks, you have the component on production. Within the Canvas way, this is not the case. As part of your design system, you have the base components that you need available, and you can create such a component without any line of code. You can just have it. But maybe someone will say, well, when we start a new platform, we will think about it, and maybe these basic accordion components will already be there. But let's think about the fact that at some point a brand — which always happens in a multi-brand or multi-site organization — will come and ask: I want to have an image on the right side. And once again, easy — this is maybe a 4-hour developer task. But a 4-hour developer task turns into grooming, 2 weeks of waiting until it's released on production. Whereas in Canvas, you have the image base component, you can drag and drop it within the accordion item that you need — or maybe only one of the items — and then you have these features. But someone will say: I am a smart developer, I will predict this type of layout and this type of use case, because we have so much experience. But can you predict the next request? No one can predict what the business and marketing will come with as a request. For 20 years within Drupal, I know that I wouldn't know what the marketing team will come with. And we don't want to spend all of our time building components. We would like to spend our time building complicated integrations, playing with AI features — this is what the developers want, and this is what the agency wants to focus on. So having this approach, we can achieve many complicated use cases without involving the developers at all, and focus on what really matters. Let's take a look at the other perspective: the developer's perspective. As you know, the developer cycle in Drupal is: there is a task, it goes to the backend developer, they configure the paragraph or the block or whatever entity it is, and they hand off this work to the frontend developer. The frontend developer styles it. Maybe there will be some questions and requests that have to be changed — like an extra field to be exposed — so there is some back and forth. Luckily, within the sprint cycle, this will be released on production. Within the Canvas way, this process is super simplified. The frontend developer can autonomously create a component without the need of a backend developer. This is possible because of the architecture of the single-directory component itself — a single folder, a single directory as the name says, where you have all the needed assets, and all of them are controllable only by the editor [verify "editor", ~29:30]. Once the component is created, you don't have any configuration — you can just commit it, and the component becomes automatically available within Drupal Canvas. This saves time, and you can do more with less, which is really something that everyone is working on nowadays. And of course, we shouldn't forget the third audience: the agents' perspective. We have seen already many developers doing the straightforward way, where you have a Figma file created by our designers; each of our developers usually has an AI assistant, an agent locally, and with the needed skills created, they can start generating single-directory components based on the design. But what wasn't enough is that we were lacking context. How do we make sure that this agent will follow the design system of Mars, which has been established for many years now? This was a challenge, because as an end result, what we could see is that the agents would create components which are not needed — which duplicate functionality but don't reuse base components that we could reuse. And that's something where we were not happy with the results. So we created the so-called Mars Agentic Design System. This is a design system built on a technology called Astro Starlight. It is a documentation framework similar to Storybook, but the difference here is that this design system exposes manifest.json and skills files which are machine-readable. So the agents can read them, can understand them, and can use them within the development lifecycle. As a result, we migrated the platform with 200+ components to end up with more like 80 components. So it was a massive reduction. And just to mention, some of these 80 components are form elements like input, dropdown — so as you can see, the real components we end up with, achieving the same functionality as against 200 components, is quite low. And the final addition, which I think is very important: we have seen this as a challenge — how can I make sure that my website, my platform, and my design system are in sync? That's something that we were also able to achieve thanks to the single-directory components and thanks to Drupal Canvas, which makes this task much easier. Now, for example, we have a CLI command which we can execute at any time, and we can sync all components which are in Drupal with the design system. So you always have an up-to-date design system. You could do this for all the brands and for all the sites — currently we do this only for the brands, but you can do it even on the site level. And for example, if you have a customization for any particular brand — if a component lives only in that particular brand — this will be represented within the design system.

[Unidentified] [33:09] Yep.

Boyan Borisov [33:10] And the agents are happy — I mean, I hope as much as the editors. We don't have a quote, but I believe that they're quite happy, because they do their tasks faster and cheaper, which usually is what they want.

33:23 Results and KPIs

Anthony Bojocan [33:23] When JAKALA presented to us this design system, I found it really impressive and promising. First, because of course, we were maintaining Storybook — and one of the things that we need to always remember is that every time there is a new component developed, or any adjustment, we have to manually update Storybook. And that's a lot of tedious work, as you can imagine. So this really saves us time and makes sure that everything is quick and efficient. And on the other hand, with this manifest.json — we had ideas on several agentic use cases that we wanted to employ in our landscape. But of course, we didn't have MCP yet; we didn't have a lot of these things that we would require. And this manifest.json file is really promising for unlocking a lot of those agentic use cases that we were thinking of. So it was really good when JAKALA proposed this solution. Now, here are just some KPIs I'd like to share with you today. After utilizing Drupal Canvas, we were able to be 40% faster in component development, and we see a reduction of about 80% in demand for new components overall. We were faster as well on website assembly, by 30%, and we were able to reduce our TCO by about 15%. And of course, all of that is just an overall increase in customer satisfaction. Actually, looking at this journey — these numbers and all of those changes that we talked about earlier — it validated the assumptions that we made at the beginning, when we took this bet to go with Drupal Canvas, knowing it was still very new at the time. And in fact, we said we have 4 websites — actually, by now, we have already deployed one more website. I think one more should be coming soon, and then potentially the strategy now is to look into what the next strategic transformations are that the other segments I showed earlier will plan in the coming years. And at the same time, when that happens, we're going to move them into Drupal Canvas. So I think for us it was really the right approach, and we're really happy to continue with it. Thank you.

35:52 Contributing back to the community

Boyan Borisov [35:52] Awesome. Thank you, Tony. Those are really great results. I just want to talk about something super important. Mars and JAKALA — we are not just takers, we are makers. And we have been very proud of contributing back to the community. I will start first with: when you're an early adopter, someone has to be the guinea pig. Mars has been happy to take this role. We hit a lot of edges, but we were happy to work together with the Drupal Canvas team to submit our feedback, so they could fix all the problems that we had, together. Overall, this helped us accelerate the development cycle. Another one is the multilingual capabilities of Canvas. This wasn't something that was in a stable mode before the beginning of June, and we had been behind testing these early betas, making sure that we could provide enough feedback so the Drupal Canvas team could improve how this functionality works. Finally — this is something that is happening today — we are supporting the Canvas Web Form Builder initiative. This is still in a UX/UI phase, but we have interviews with some of our experienced site builders and experts to provide such feedback. Secondly, let's say what code we can contribute. You may have seen earlier today there is one feature in Canvas which is quite interesting, where you can copy one component from one page to another. In fact, there is an issue queue right now which allows us to copy one component or a whole section — a whole page — from one website to another. And that's quite tremendous: I have heard many editors asking about these features, and I think they will be very happy to have something like this. Secondly, you saw in the demo the Tour module. The Tour module is something that we have created. Currently it is still not part of the Drupal Canvas module, but the merge request is there, and it's ready to be merged — for everyone to benefit from the Tour module as part of the Canvas experience. Finally, just a final message — the final takeaway that I believe everyone should take from here. Many organizations have invested in Drupal, building great digital estates. And sometimes you hit a moment where — I have seen these stakeholders — they think, okay, we need to change, and they change the technology. I believe this is very wrong. Really, Drupal has the modernization and the capabilities that can achieve this modernization and be best in class for your digital estates. So do not be in a hurry to change your digital estates with something that you don't understand and start from scratch. Drupal Canvas, Drupal AI — all these great initiatives are easily adaptable to your existing platforms. And that's something that you need to have in mind. Do not start from scratch — just make it the next step for your innovation.

39:26 Why Mars stays with Drupal

Anthony Bojocan [39:26] Yeah. And I would say, I think for me, one of the reasons that we have this strength is because of the Drupal community — a lot of very engaged, passionate people who really like what they're doing and continue to improve. And I think this is why you can see the evolution of the platform, of the solution. I remember last year there was a product that we were discussing that was kind of introduced, and when I looked at it and how we could use it at Mars, to me it was like: okay, this is not enough, it's not at the scale that Mars requires. However, it was reintroduced this year, and it was already fitting that big-scale requirement that I'm talking about. So I think this is really one of the strengths, and one of the reasons for Mars, for example, to have decided — and continue to decide — to stay with Drupal.

40:22 Q&A

Boyan Borisov [40:22] Yes, thank you very much, Tony. And thank you to everyone for being so late with us. Thank you very much. If you have some strength still, you can ask some questions — I can give you the microphone.

Audience member [40:22] How do you handle Git configs?

Boyan Borisov [40:22] Yes — I mean, the Drupal Canvas configurations, whenever it's needed — which is needed, let's say, for things like Canvas templates, for example, because as you heard about single-directory components, this is not needed really — it is handled in the very same way as standard Drupal configuration. It uses the standard Drupal configuration CI/CD workflow, usually. Nothing new. Any other questions?

Anthony Bojocan [41:13] There is one there.

Boyan Borisov [41:14] Oh, sorry.

Audience member [41:18] Yeah. So you mentioned in a slide that you have basically a copy-paste capability between sites. And I'm curious, because usually in large enterprises every brand wants to be unique — they have to be unique — and then every country and market also wants to be unique and have something different. How do you actually approach this kind of content strategy? How much content are you able to share and reuse between different markets and brands, et cetera?

Anthony Bojocan [41:34] I think it's a combination of different aspects — and I was discussing with Julie [name unverified, 41:34] about this earlier — because it's true that there are some unique requirements from each brand. But of course, from the technical perspective, you also have to consider what would be easier to govern, to maintain. So what we normally do in this case is have a conversation with the brand manager to understand: is this a functionality that is really important for the brand? For example, one of our brands is Perfect Fit, and one of their marketing ideas is this kind of "5 reasons why you should choose this product" — and we had to build a component, because that's really the core of their marketing strategy. But for the others, it's really a discussion of: is it really important and unique, a differentiator for your brand? Because if we have to do it that way, then this is what it means — there's cost to consider, there's maintenance to consider. Normally, if you talk about what it means — like cost — it becomes easier to align with them. But sometimes it can also be complicated. I hope that answers your question.

Boyan Borisov [43:01] Thank you. Any other questions? If not, thank you very much — and have a last hour of DrupalCon Rotterdam.

Anthony Bojocan [43:09] Thank you.