Chris Bryce: Okay, we are live, clicking on a variety of links here. Okay. Welcome to the show, everyone. Thanks for joining us here for Beyond the CMS, a podcast video cast, where we discuss everything to do with content, content management, CMSs. I got the idea from the original Car Guys show, where people could call in and ask us anything.

Chris Bryce: And we'll do our best to give great answers. And we have a special guest today, from Dotfusion, Janice Lee, who is our customer success lead here. Janice, welcome. Thanks for making some time to join the show. It's a pleasure to have you here. Do you wanna just tell everyone a little bit about yourself?

Janice Lee: Sure. So I am the partner that looks after growth and customer success at Dotfusion. What that means is I'm responsible for making sure all the accounts are moving smoothly, thinking about growth opportunities for them, always looking for ways to elevate our service levels. But also, I like to say I'm a bit of a digital translator as well, 'cause I work with a lot of our marketing folks and clients, just making sure that we're bridging the conversations between our technical teams and the marketing teams, and basically showing them how certain things can really help with their brand initiatives.

Chris Bryce: Super cool. And today we're really, oh, we got a little thing here, just got a little popup. One thing we do is monitor the social channel, so if anyone has any questions, I'll be looking frequently over to the side to make sure we answer any of those questions. And today we're gonna talk about how AI is impacting content strategy, kind of across three main themes: on websites themselves, for our customers and the authors and the folks that are using different content management tools, and then if we have time, talk a little bit about the impact that it's had on our team in general. Janice, do you wanna maybe kick things off? Let's start with kind of a customer lens. Talk a little bit about what you've seen over really the past six months on how AI is impacting customers, and we've got kind of AI terminology and agents and all kinds of stuff.

Janice Lee: Yeah, yeah. I think it could be a little bit overwhelming, and even for myself, confusing in the beginning when you're just getting into it. Everyone is just, you know, the buzzword of the day is AI, then you're hearing about agents, and it can be a little confusing. So to me it was really helpful to kind of break it down with an analogy. I think from a marketing perspective, what's important to know is, you've heard of generative AI, so everybody's probably used ChatGPT once, or Claude, or whatever tooling they're most comfortable with, Gemini, Copilot, things like that. And then there's the agents, which is different, but based on AI.

Janice Lee: So I kind of use the analogy that generative AI is like a creative freelancer. You give it a brief, you have it write some amazing copy or a campaign, and it does that super fast. But once it's completed its task, it's gone. With agents, what you're talking about is like having that creative freelancer, but superpowered. So imagine you have a very dedicated customer service account executive working on your campaign. Not only do they make sure all that content is great, but they're also helping you manage your budgets, looking after your campaigns, doing work in the background autonomously without you necessarily having to prompt them. And so that's kind of the evolution. Where I think most people are at today, or at least from a marketing standpoint, they're probably more thinking of AI as the generative AI, which is great for individual productivity. But what we're looking to, forward thinking, is how can agents also help marketers and our clients out there.

Chris Bryce: Yeah, so I have an example actually of what an agent looks like. I'll try out this screen sharing idea here. This is one of our agents, this is one that helps us with our marketing and our outreach, for example. And what you're seeing on the screen is a sequence of events that different agents activate and deploy, and there's AI involved in this process. We can get into more detail on the demo day sometime, but essentially there's a daily trigger at eight that goes out and takes a list of leads, runs an AI generated prompt, validates it, goes through an editorial workflow, and then sends out an email.

Chris Bryce: This is also something that we're starting to integrate into content platforms, and it might be a great time to discuss how AI is impacting authoring journeys. What we're seeing is that one of the, not challenges, but maybe opportunities in streamlining content production and workflow, is to help authors manage the environments that they're creating their content in. Most of the CMS platforms these days have the ability to massage your content with AI once you're inside the platform. A little bit of feedback that we've got from customers is that they've already done that work, so they're using Claude or OpenAI or whatever tool that they're using, they're creating their content, they're putting it into Google Sheets or a Word document, someone's approving it, and then they're going through a process of moving it into the CMS and publishing it.

Chris Bryce: Well, what agents can do is connect that native origin source of the content directly into the CMS. One example of that that's really awesome, I'm not sure how many books that are listening to this now or later use a headless CMS, we love Agility CMS, and Agility allows us to use its APIs to connect to other systems, like the one that I just showed you, that allows the authors to be managing and creating their content in one place that they're really used to, and we can connect that directly to go into Agility and get published, which is really awesome. So we're always trying to figure out what is the unique situation that our customers are in, where they're creating their content, and how can we make their life easier, streamline their workflows. That's a great use of an agent situation, and essentially we're looking at streamlining any kind of operations that are done more than once with some kind of agent or workflow. What are you seeing out there, Janice?

Janice Lee: I'm seeing within some CMS partners where they're integrating, if not AI, some sort of automation tooling to kind of make things faster. This is more on the development side, but I know one of our partners recently launched an integration with Figma. So in the past, what would happen is a developer would take a Figma design and then manually build out all the components into a CMS, and that's the part that takes the most time. As a marketer out there you're like, well, we have the design, why does it take so long? A developer has to build everything in, within some CMS partners where they're integrating that.

Janice Lee: I think it's a really cool development and just really forward thinking, because what happens is the developer now can just go into Figma, I believe there's some sort of public plugin that's native to the CMS, and then the plugin actually allows them to just lift the design and port it straight into the CMS for them to develop. I actually asked my developers, but we haven't tried it yet, 'cause apparently that was a very new feature, but developers would love that. Like, hey, that makes my life easier, I don't have to manually compare this Figma design to where I'm coding, I go straight in and it does. Those create a lot of time savings, and ultimately for clients, you get your website faster, or your application factors. So there are things also happening in the backend that maybe, from a client facing, marketer standpoint, they may not see is happening, but it's good to know about these changes, because you can understand and see, maybe ask your teams, hey, what software are you using, are they using some sort of automation tooling that can expedite these things. So it's good to be aware of these things, not necessarily to understand how it works, because I don't understand how all the details work, but I need to know enough so that I can have these conversations. So I think that's important to know.

Chris Bryce: From your perspective and feedback that you get from customers, how do they feel about AI? Are they embracing it, are they tentative, what's the weather out there?

Janice Lee: So people that I've spoken to individually, I think they're excited. They're also a little bit, I would say, gun shy and maybe intimidated, because I think a lot of the language out there is geared towards very technical folk, MCP, RAG, all these terms that seem like, oh, I can't get ahead of it. Simply put, if you have an interest in AI, or you just wanna know how it can make work life better, it's good to just kind of play and be comfortable in finding spaces that allow that. Those are just amongst the people that I have talked to, I think they seem interested, they just kind of don't know where to start.

Janice Lee: If I'm talking on a bigger stage, last week I was at an event that was full of enterprise CMOs, so we're talking like td, craft, Canadian Tire, there was a panel, this is public information. But when asked about where their organizations are in the process of adopting AI or integrating it into their systems, a lot of them kind of took a wait and see attitude, which is interesting, like they know it's important and I think they understand how relevant it could be, but they're also massive organizations, and I think the reality is to implement a small change is a huge undertaking. So they're kind of trying to figure out what's the best way to implement these tools. I say wait and see, obviously I'm not in those conversations behind closed doors, there may be some other discussions going on. I suspect that they're probably integrating it in pockets of the organization, but maybe it's not broadly distributed. My guess is that they're probably enabling people to use, for enterprise, like, use Copilot, use Gemini, or whatever it is, the tools that come as part of their systems and platforms. But beyond that, I don't know how deeply that automation tooling is being integrated.

Chris Bryce: Well, one of the examples that I know of, we have some friends at one of the major telcos in Canada, and they really adopted AI early. Their main concern was governance over their information, so they didn't wanna be just uploading all their corporate documents and customer service information into a model that didn't have really strict governance. They actually have their own internal enterprise model, and they've really encouraged all teams to be experimenting and exploring. What are your thoughts on governance, or concerns, or what should CMOs be thinking about when they're now tasked to take on implementation of AI?

Janice Lee: Yeah, that's an interesting conversation. That came up as a debate with some AI enthusiasts and colleagues, because as we know, there's actually no public form of governance today. I think with the most recent administration, they said they're going to start looking into AI in Canada, but there's no policy, there's no set rules. It's kind of like the wild, wild west right now, so it would say self-governed. And so I would think that it's up to the organization's leadership team to really set the tone for what that means. There's this balance between how much should you govern and be strict versus how much creativity do you want to allow to flourish. I don't know if my perspective is to let creativity flourish as much as possible without doing any overt harm to anyone or anything, but that's also very subjective, in some cases. So I think that's kind of hard to distinguish. I know you and I probably have had conversations about that, like what would you say around governance for AI?

Chris Bryce: Well, you need to protect your data, and quite frankly your prompts as well. Those prompts become an important piece of your content arsenal in a way. A lot of the ways that we're seeing the future of the web is that a lot of the content is going to be generated to a large degree by AI. Right now we're in a state of time where copy and images are being manually produced and getting to publishing state more quickly, but we see the future is that someone hits the website and agents are spinning up ultra hyper-personalized, unique content on the fly, referencing what we would call the brand prompts. I think that brand prompts library needs to be as secure as the information that it's understanding as a knowledge base. But speaking about personalization, I know that was on your mind, do you wanna talk a little bit about how AI is helping in personalization in general?

Janice Lee: For sure. What we've seen, and what we've been doing, taking it back to websites specifically, there's been some conversation out there about how does AI or agents impact websites, are websites still relevant in this world where AI is changing things a lot. My short answer to that is yes. I think for any market out there, your website still is your platform, the place where you go to showcase your brand, talk about the personality, surface things like that. I think what AI is gonna help do is start engaging your audience a lot more in a different kind of way.

Janice Lee: So one of the things that we've talked about here at Dotfusion, and what's been really cool, is how can we enhance a website that's already there. Maybe you can fire up and I can talk a little bit about these voice agents, but the idea is you can create engaging experiences with your users or people that visit your site that you wouldn't have had in the past. So what we looked for was an opportunity to take that old idea of, hey, you have a chatbot sitting on your website. I think a lot of times when people think about chatbots, at least I used to, it's like you type in text and then you get a list of prompts that give you, choose one of these three things. It's not really engaging much, it's very transactional almost. But I think a lot of the AI today has unlocked a lot of capabilities to make that a more human-like interaction, and it's pretty cool.

Chris Bryce: Yeah, and to your point, voice is now part of your content library, and how do you manage and govern all that as well? We've been doing some experimenting with very advanced voice agents, and we've actually got to the point where we can have the agents change languages on the fly, and it's really cool. Maybe I'll just show a quick example of how this all works, just hoping that my audio works. So we're tooling a demonstration for a shopping center, and I'll kind of just show you an example of how this is not only replacing humans from a customer service perspective, but it just has infinite knowledge. Let me just, gimme a sign that you can hear this.

Chris Bryce: Welcome to Yorkdale Shopping Center, Toronto's premier fashion destination. I'm Samantha, your personal concierge. I can help you with store locations, dining options, event details, parking information and more. What can I help you with today? Oh, could you tell me the center hours today? Sure. The hours today are Monday to Saturday, 10:00 AM to 9:00 PM. Sunday, AM to 7:00 PM. Please note we are closed on statutory holidays such as Canada Day, Labor Day, Thanksgiving, Christmas and New Year's Day. Is there anything else I can assist you with? Yeah, I'm just really curious, how do I get to the Apple store? Certainly, the Apple Store is located on the ground level of Yorkdale Shopping Center. To find it easily, you can use our online map, or if you're already here, you can look for directional signs. The store is close to the central area of the mall. If you need more detailed directions, you can visit yorkdale.com or contact guest services at 647-749-7467. Can I assist you with anything else? Yeah, speak Mandarin.

Chris Bryce: So all that to say, things are progressing really quickly along all forms of content, and how AI is impacting the entire experience. The sky is really the limit. Janice, do you wanna talk a little bit about how that's impacting teams, and we can maybe speak about our teams internally and some of the really big changes that have happened over the past six months?

Janice Lee: Sure. Internally we are using a lot of AI tooling and automation just to speed up productivity and just like hyper, I would say, what do you call that, ludic drive our team, which is basically, I'll give you an example. I was submitting for an RFP, and you know how long these documents are, usually it will take you, I don't know, like a week and a half to just get through the material, respond to it. I did an RFP response in about three days, and that's three days not dedicated to it, but a combination of being able to feed AI the contents of the RFP, the requests, and then doing some very intelligent prompting to understand, since I knew what I needed to create in terms of an output.

Janice Lee: So it was almost having a RFP writer or a right hand person really helping me do all the, I wanna say the grunt work, but really all the heavy lifting. What I've found is that AI has really helped us, our team is, I call it small and mighty, we have a lot of senior people. So what AI is providing is almost like a junior development, junior content, junior designer team, where we can quickly spin up prototypes and materials, and then we use our strategic level and senior level people to really review, see, and adjust things as needed. It'd just be like if you had junior team members on your team, and we're just performing in the same way, and it's just happening a lot faster.

Janice Lee: That's one thing I noticed from a productivity standpoint, and that is across marketing, whether it's us promoting our podcasts, getting our socials up to date, creating case studies, design, or even development and deployment. Just lots of great productivity. And on top of that, we're all really excited, 'cause I think we are in a unique position where we get excited by new technology things, we wanna know what's at the forefront. We get that not everyone is, but it's almost like we're happy to be those technology guides for folks that maybe don't have the time or the technical expertise to dive into things. We love doing that.

Chris Bryce: Yeah, and when you talk about the facets of our business that are being impacted, from marketing, operations, design and dev, and strategy, that's kind of our core pillars here, it's really impacted all of those. Something that I've found really interesting is how quickly these tools are maturing. We've gone through, hey, look, I made a cool image of a dragon or something eating ice cream, to fully deploying prototypes that are now the assistance to our UX team, or fast tracking code from our development team as they've kind of all adopted AI tools to make things move quickly.

Chris Bryce: And I think we're in this interesting stage where the automation aspect of agents that are supporting our team are kind of, they're like half baked. But it went from zero to half baked in no time at all. So we need to, it's almost given us a little bit of extra work to do, because not only do we need to be shipping products for our customers, we need to have one eye kind of on, oh, what's next and what's new. And we're starting to develop processes around those, because there's so many options and different tool sets out there that you need to kind of have a bit of a focus and a plan to make sure that you're staying current, but you're still getting work done.

Chris Bryce: So I think we're in this stage right now where there's a lot of excitement and a lot of new things to be engaging with. We're probably for the next year or so still operating in old school mode, where we need to think through design, develop, and ship, but with an eye to new school mode, which is prompt agents, get prototypes and first drafts of things ready, and then we just polish off. But it won't be long before things are, you know, 90% baked, I think.

Chris Bryce: Did you wanna, I know we were talking a little bit earlier about MCP servers, and we spoke to our CMS partners and they're looking at developing MCP servers, and I think this is a buzzword that's been thrown around a little bit. Do you wanna explain that a bit to anyone who's listening?

Janice Lee: Okay, I will do my best to describe what that means. So, MCP servers, it's a fancy term, it stands for model context protocol. And so what it is, is a protocol that allows AI models to securely connect to external data sources through some standard server connections. And I sourced that from Claude, so just letting everybody know that I actually did that, 'cause I wanted to explain it. It's still new to me personally, and I know that developers are starting, some forward thinking developers are already starting to use MCP servers. They're basically connectors that allow agents to talk to other agents.

Janice Lee: Yeah, and so you're basically leapfrogging, like, oh, do I need to build an integration from this CMS to this particular thing. I don't know, the example I gave you was the Figma one, but can Figma talk directly to a headless CMS without a dev spending hours coupling those things together? It's funny 'cause a lot of these MCP servers are now out there in the wild. They're open source, they're free, 'cause they want people to start using them, adopting them. So there's this other world of MCP servers that are allowing anybody who's technically inclined to connect different agents together to actually do work faster.

Janice Lee: So that is a benefit to any clients that are looking to build websites. You should be asking your partners, hey, do you work with MCP servers, how can you make this faster for me? 'cause maybe in the past, let's say a website would take, I don't know, six to eight months to build, it's just time, it's hours and effort, development is usually the longest stretch. But now maybe you can reduce that to half, because so much of the heavy lifting is already being done by MCP servers, some automation tools, whatever it is. So you're not sacrificing the product quality, you're just getting it in half the time.

Chris Bryce: Yeah, and we're starting to see that from a customer perspective, in that they're aware of what's going on and we have to react to that, 'cause there's an expectation now that we're shipping products faster, but at higher or better quality, because they know that we have these tools, so we have to manage that properly. One example that we were helping a customer out with is a model where they could continue to use Trello as their content editing environment. So they'd have content being moved through Trello cards. We set up an MCP server to connect Trello directly to be deployed onto the CMS platform and onto the website. So you move a card into ready to go, the agents are waiting for that signal, take the content, update the image and the copy there.

Chris Bryce: So this is all happening, I think right now there's less development overhead required, but there's still a little bit of development that's required, so we need senior folks to be kind of managing that, and I think there's great opportunities for folks that are new to the technology industry to be jumping in as prompt engineers and providing oversight. And there's a lot of discussion around jobs, job security, and what are the jobs of the future. And while there may not be such an emphasis on becoming a JavaScript expert, let's say, you still need to know how it all works, as if you would be a manager, knowing how someone newer is doing their work and how to validate all of that as well.

Chris Bryce: From a user engagement perspective, do you think there's demands from users for different kinds of content to be on websites, what's their expectation? Is the demand higher because they're assuming that brands are using AI, so they expect something new all the time?

Janice Lee: Yeah, I definitely think it depends who you ask. Let's talk about age groups and demographics. I was talking to a friend the other day and she's like, oh, I still use Google for all online searches. I was like, oh, interesting. But I think customers in general are getting more savvy, and they know that AI is being used in a lot of places. So I think the expectation, if you were to engage a user on your website, definitely I would expect some level of personalization, some way to engage me more than just share, you know, it's not a one way knowledge transfer. And I think with a lot of websites, and if I think about B2B today, that's kind of how it is.

Janice Lee: Is there a way to kind of bring in your user more, to engage with you, to showcase your brand personality more, and get the customers, that's how you're gonna build loyalty. Your website is basically now your frontline, almost staff member that is engaging your audience. If they've arrived at your website, they got there somehow, so they want to know more about you. So how do you do that? Oh, sorry, go ahead.

Chris Bryce: No, that's what I wanna say. I think there's so many things that we've started looking at that there's definitely gonna be more, and as time goes by and we know what agents are capable of, there's gonna be more opportunities to unlock different things. Personalization is already kind of the starting point. But people come to a site at all times of the day, you're not gonna be just working there nine to five and expecting somebody to be able to answer your questions. It'll be instantaneous. I think the demand is gonna be a lot higher.

Chris Bryce: Yeah, it's true, lots to be thinking about. Man, time goes by pretty quickly, but I've got a lot of questions about AI's impact on our kind of shipping process, and how do you get prototypes to the market really quickly. I just wanted to highlight one product we're working on, which is essentially a website that will help people with boats find someone to fix their boat. There's a problem in the market, and normally what we would do is go through a fairly intensive intake period with the customer, spend six to eight weeks building a prototype, do all the data architecture.

Chris Bryce: And the example I'm gonna show right now is something that took us less than a week to build, and this is the entire prototype for this product. So there's an AI chat that answers your questions, there's this idea of finding an expert and linking to them, and search, sort, and filtering their model out. You can get a video consultation, there's a dashboard, and then what we've thrown in here for demonstration purposes on our end is we've created an area where we can be constantly updating our prompts, because the prompts are really governing everything and the behavior of this dynamic tool.

Chris Bryce: So what I'm highlighting here is not only did AI build this, although the prompt engineering and the systems architecture needed to be pumped into AI, but it's also operating the website. It's not just looking through hardcoded instructions on how to behave, but it's actually able to dynamically interact with the user on the front end. So that's pretty cool stuff too, and it just speaks to this amazing point in time where folks with ideas for digital products can get them out in the market fairly quickly, just to test the proof of their concept, and if they have some signals that there's traction, then they can polish it off and finish it and take it to market. So it's a really great time, where you're not blocked by large development teams to get an idea out there.

Chris Bryce: Just wanted to say thanks to everyone who made a moment to drop by today. Janice, any closing comments that you might have?

Janice Lee: No, I would say for anyone who's interested or feeling a little bit hesitant about just jumping in with two feet, I would say just start small. You don't need to know the next craziest thing, but just start with what you know. And if you don't know, you can always ask a friend, I'm here, I can kind of direct you as a starting point. Because if I think about my journey, AI's been around for a little while now, but I think it's really just rapidly gained momentum in the last six months or so. There's a lot going on, tech week in Toronto is next week, so there's gonna be a lot going on there as well.

Janice Lee: I would say, if you YouTube search anything, just get the basics right, don't try to learn the next craziest thing, because you'll never be able to catch up. But I think just start small, like with anything, you'll take your first step, and then from that first step you'll find out what is interesting to you, that will unlock another step. I don't think there's any real AI experts out there, because the landscape is rapidly changing so much, but I call myself an AI enthusiast.

Chris Bryce: That's a good way to put it. We're out of time. I just wanted to thank you for making some time in your busy schedule to join us for a special edition. Thank you, this was fun. We were talking about how AI impacts content and websites, and even our internal team workflows. It's been a pleasure hosting. We go live every Thursday with Beyond the CMS, where we talk about anything to do with the current and future state of content. Thanks for listening and have a great Thursday. Bye everyone.

Janice Lee: Bye-bye. Thank you.