Tent Talks Featuring Sabina Leybold: Content Modeling Meets UX: Designing for Scalability and Reuse

[00:00:32] Chicago Camps: What is content modeling and why should UX professionals care about it?

[00:00:36] Sabina Leybold: Content modeling, and the best way I can describe it is the activity of defining and documenting the structure of your content. It can come in a couple different forms and early phases, it might be more like high level conceptual models, really exploring the types of content that even exists on a platform or in a system and the relationships to each other.

As you progress through the content modeling activity, you will probably get more granular about your structures attributes, so that might be refining CMS fields, looking at taxonomies, content types, components, all those types of structures.

And to me, I am a unique use case in that I am a on paper hybrid UX designer and content strategist that is part of my entire jobs remit. For a UX professional that is more design focused, a true UX designer isn’t responsible for content modeling, I still think that the content structure is incredibly important to understand because there’s a relationship between the content that is stored in a system and how it displays the user on the front end and that relationship, it’s not always a one-to-one relationship. What’s stored in the CMS may not always look like a specific screen that you’re designing, but your UX design will always be better, which is to say more efficient, more user driven if it considers the content that it will contain.

I think most UX designers would agree with that. Knowing more about the content that will sit in a layout helps them establish a layout. But I think the other thing that I know we’re gonna dive into a little bit more is about content modeling and how it’s often done to fuel content reuse across the platforms. Again, understanding that content structure can really help UX designers to create more scalable systems that save time and money and headaches, both for kind of that end user and for the business user that is managing and controlling the content itself.

I think a lot of it too is that UX as an abbreviation has become so commonly used that I think people forget the user experience and the user is experiencing things that are not just boxes and components and frames and this, that, and the other, that are things that I think a lot of UX designers have gotten pigeonholed into focusing on is what does it look like? How will it sit? What are the, how much space is between this thing and that thing?

And that’s not really, to me, the bread and butter of a user’s experience because there are all sorts of platforms and websites and whatever that look very, I don’t know, that just don’t look updated, I guess, that don’t have that sort of aesthetic to them.

But they are still often good systems because the content is really good and people still enjoy going to those places because the content is good. Now, that’s not to say don’t redesign something to look more modern, to look cleaner, to have better spacing, whatever it is, but that’s not, I think the spacing is not what makes somebody come back to an experience ever.

The content is everything.

[00:03:53] Chicago Camps: How do you approach designing for authoring experiences and what should teams consider when structuring content in a CMS.

[00:04:02] Sabina Leybold: My focus is heavy on websites. I’m in an agency setting. My focus tends to be web content management systems. I am far less in digital product app type of land.

Sometimes I do some e-commerce types of experiences too. Sometimes there are interactive journeys and tools and programs to that effect, but for the most part, I’m heavy in websites, so that’s mainly what you’ll hear from me in terms of my approach. But I do think that it’s something that could be applied elsewhere for folks who are not in that world.

For me, I start with workshopping. Again, some of that is agency preference, but always working with the client on what do they want to see? What do they need from us? So usually we’ll start with standard discovery workshops. We have one that we call goals and challenges. Very much an intro. What do we all need here?

We’ll often do jobs to be done, which UXers will be familiar with. Sometimes we’ll do journey mapping. We’ll usually have conversations about what’s your tech stack? That kind of thing. When it comes to content specific workshops, I am usually doing conversations about information architecture, both current state and ideal state.

That can include taxonomies and site mapping, reviewing content inventories and design system inventories of a current state, and again, trying to understand what the ideal state would be. So that’s the standard bunch of workshops that we do in that true kind of more discovery, light UX and content, not that it’s light, but lighter on general UX and content.

Then when we dive in really specific on content modeling, basically the flow goes, I will create some sort of attempted concept model. Or ecosystem map a la Scott Kubie. So fans of him will know sort of his ecosystem mapping approach. So looking at what are just the general relationships between everything that exists in a system. I usually will do that on my own, based on what I’ve heard from those more general UX and content workshops.

Then I’ll present that to a client. We’ll work together in a workshop setting, on kind of, is this what exists? Are there things that exist that I don’t know about? All that, kind of identifying the gaps within the conceptual model.

Then we’ll move into a structural review of existing CMS fields. Or sometimes attempting to deduce structure from page expression, depending on what they come to us with.

We’ll discuss gaps or challenges or opportunities that exist there. Usually I take all of those inputs, go away, come back, propose a content model. Usually this is all in Miro with stickies, but it would be like a sticky for a field plus maybe some notes about it and going through how might this look in the CMS. Still in a format where we can move stuff around and usually then it’s one to two workshops to review that model.

So that’s like asking questions about things like free text versus taxonomy. What would be better? Do you want a totally open text field? Do you want some structure here asking about things like how many characters should we limit this to?

What are the default values? What are the required fields? All those really like granular. Types of questions about specific fields. All of that then serves as the input for what is usually the final deliverable, which is like a spreadsheet basically of all of the fields that we’ve identified within the content management system and all of their attributes, their limitations, anything that kind of needs to be coded as such, validations, things like that.

There’s a lot of considerations. Obviously, I mentioned a handful of them in there, but to me a lot of the considerations that I’m always asking the sort of like top 10 questions, if you will, are like, what is the source of this content? Is it the CMS? Is it within this content type or another one that you’re drawing information in from?

Are you pulling from another system entirely through an API through JSON objects, this is where it gets a little techier and maybe that’s the part that I think maybe freaks some UXers out, occasionally, just because it’s, again, not usually part of the expectation of somebody doing UX to have that background. But again, I think it’s really important.

Other types of questions I like to ask or concepts I like to get at are things like what needs to be structured for reuse? What does it mean to be, what other content exists that we could pull in to reuse? Does that content structure work here? Do we need to transform it? Where might we expect this content to be reused? And does this structure support that reuse all that kind of question.

Then again, I think this is very much more in what UXers would typically think of as being a content strategist job, but what information do we need to store that doesn’t get displayed on the page itself or on the app screen itself or whatever it is. So like the published status or the updated date or teaser text that’s not on this page but is on the card version of this page, the SEO metadata, that kind of thing.

And then I think another last one that a lot of UXers I think don’t deal in, but the ones who do are gonna be like, ah, the thorn in my side is, does this content need to be translated or localized in another way?

And will we store those translations or will we rely on like front end Google Translate, auto translation types of approaches. So that’s my quick list of all the really just easy stuff to go through.

But it, it’s complex. It’s really complex, but. There’s a lot to be thinking about. And again, I think there’s a lot of overlap between what a content strategist often is thinking about and what New Xers can and should be thinking about even within that world.

So there’s your hit list.

[00:09:50] Chicago Camps: How do you see AI tools or AI agents influencing content creation and authoring experiences? And what should teams consider when designing for both human and AI authors?

[00:10:02] Sabina Leybold: What I see most from my clients is that AI is a distraction at best and actively harmful at worst. And that is, as you say, not the world’s most popular opinion in the industry at the moment. But it’s also not a popular opinion in the industry at the moment either. And I think it, it is very nuanced and there’s no way to approach the topic without some level of nuance.

For me, what I mainly see is tons of businesses who in particular. Want to have an always on chat bot that can do all of their customer service for them, because it seems easier to them to have a chat bot do that rather than a person doing it.

And when I say easier, realistically cheaper in the long run. But also again easier in a variety of ways, right? Hypothetically, training could be easier in that way, easier to ramp up, possibly easier to control. Again, you have to think about money as being sort of the, the impetus for a lot of that.

But I think the things that my clients often forget about that sort of chatbot approach is two things. The first is if the goal is reducing customer service time, what ifs? Your products or services were easier to understand in the first place. So the customer never encountered the issues that they’re contacting a service person about. Right? The second thing, and I think you’re with me on this, is that your AI is only as good as the material it’s trained on.

And so for both of those a robust content strategy is essential in the path forward. Even you as a more pro AI person, and me as a more anti AI person, I think would agree on that content strategy as being essential to making it not be a garbage in, garbage out situation.

I think when organizations seek out an an AI chatbot, again, in particular, but they don’t want to include serious investment in structured content on that roadmap, it’s saying you wanna look like a bodybuilder, but you don’t wanna lift weights. And basically it’s saying, Hey, I’m all about the vision, but the path there seems mundane or like irritatingly pedantic, or a side quest to the main journey.

But I think they just don’t realize that the vision truly just isn’t possible without the wildly complicated steps along the way. I think that’s part of where that nuance comes from for me, is that I am often asked by people who do not understand what AI takes to make it make sense when things don’t always already make sense.

Now, I was talking primarily about chatbots just to start there. That’s the application I hear the most often because that just happens to be what my clients tend to come to me about in terms of AI applications. But there are other uses that I think are interesting. But again, require a level of nuance to be applied to them.

So there’s a couple. AI for alt text is something that I’ve heard of folks doing. I am so adamantly in favor of accessibility and working accessibility into your process and training on accessibility and, and that accessible design is good design and inclusive. Inclusive design is essential for creating the best design possible, all that kind of stuff.

So the idea of, hey, in the absence of time to be writing our own alt text, could AI fill that in for us and help to contribute to the sort of accessibility of the platform or of the web. Nice idea. Great. In theory, I can definitely see how that could be better than none. Right? AI generated alt text being better than no alt text.

Sure. But aI can be inaccurate and it can be biased. And in fact, we’ve seen lots of cases of both of those things. There’s a really great talk from the conference axe-con about AI for alt text specifically, like maybe two years ago, could be three years ago if you’re looking for it. And it basically, the person who presented posits that doing AI for all texts is actually sometimes the something is better than nothing, is not always true.

And that speaker has lots of really good examples of that. I’ve also heard of applications for AI, for things like text versioning. So I mentioned earlier like creating teaser copy for a card to represent a page, right? As being an example of thinking of content that sits outside of a specific expression for content modeling.

Again, AI, sometimes the idea is, okay, let’s make it faster to create that teaser text that represents the page. My response to that is, is it really that hard to write teaser copy? But that’s coming from my perspective of really not feeling like there’s a ton of great use cases, at least that I have seen for AI and feel writing a sentence should not be wildly complicated.

I also worry about AI for larger generative content, so things like blog posts. Honestly, my opinion on that is like the world doesn’t need more generic blog posts with very little point of view, and they don’t need more like boring images for the sake of decoration. I don’t know. I’m, like I said, I’m a little pessimistic about AI.

I hate to be so negative about something that others find really super promising, but I just see so many drawbacks around like ethics and sustainability and just like the general flattening effects that I have found generative AI to have. Then again, that piece I spoke about at the beginning of the business’s eyes are bigger than their stomachs on like the actual content strategy work that’s required for effective AI implementation.

I’ve seen very little good come out of the AI hype, so until I see that, that’s gonna be my take on it. To me there is a lot of things that people are doing with AI right now that is pure laziness and I think even you as a person who like sees value in AI you probably agree you’ve also seen lots of people do some lazy shit with AI.

At the same time, I do think that something like, I wanna be focused in a meeting. I would like to take notes, but if I’m taking notes and listening at the same time, neither my listening nor my note taking are gonna be great. So having somebody else there to take the notes for me. And it doesn’t have to be a person, it can be a system that is actually in my mind, an okay use of AI despite the questions that I have around, again, ethics and sustainability of it in general, but in terms of things like note taking, spitting out, that kind of response.

Cool. Cool with it for the most part. It’s just so much of the, as you say, the like hot garbage that the sort of flattening effect that we talked about of just another boring take that wasn’t written by a person. But I just think so much thinking is writing and vice versa. I know that not everybody feels that way.

I understand that at the same time, like how do you know what you think about something until you’re writing it out? I can’t fathom not feeling that way. When I see people not writing, I see them not thinking so, but that’s why I worry about AI in particular.

[00:17:05] Chicago Camps: What are some common misconceptions about content reuse and scaling within design systems?

[00:17:10] Sabina Leybold: I got a couple. I hear a lot about content reuse, again from my clients, and they often come in with a preconceived notion of what they’re gonna get out of content modeling as it pertains to content reuse. So there’s a couple, I think the first day is creating systems that allow for effective reuse.

Takes a lot of work, which maybe shouldn’t surprise anybody, but to make something good, repeatable, scalable, all of that. It takes a lot of breaking down all the little pieces and not as a lot of work. I think, not just on the content modeling side. Is it a lot of work, but also internal politics around content reuse in particular.

So I’ve got a good example that I was working with a children’s hospital system and they were working with us on department pages, the department content type for their website and basically what we landed on is every department needs to have four different name fields. So the first is department name and that’s like the friendly name.

It’s the page title, it’s the search title. Pediatric orthopedics and sports medicine. Right. Okay. That’s your example. Then they have something called the department official name, which is the first mention in the body copy. It has the full brand name, so the example is like Best Hospital, TM Williams Center for Orthopedics and Sports Medicine.

Right. So it’s like the very official name that they use for official purposes. Then you’ve got like the department sponsored name or the department nickname. So sometimes that serves as the page’s subtitle. Sometimes it’s throughout the body. So in my Greek example, we’ve got William Center for Orthopedics and Sports Medicine.

It’s basically the same as the official name, but without the hospital system brand in it. And then we’ve got the department’s discipline name. So this is now, it’s not the something center, now we just have orthopedics and sports medicine. No pediatric or adolescent indicator. Right? So they need four different types of name fields.

So in order to reuse the name of our department across our system, in case the name of the department changes in case that person who was once sponsoring this department no longer is sponsoring this department. Whatever it is. In order to edit that in one place and have it apply everywhere means you have to break things down into incredibly granular parts.

And that piece of it is not just difficult sometimes to make those decisions on the content modeling side, but also the politics side. So you can imagine getting to four different types of names had a lot to do with people’s feelings internally in the hospital system . It’s not just the alignment there, but then, okay, now every department also has things like what is a summary versus what is a value proposition?

Does the marketing language ever belong in the summary? Does it only belong in the value proposition? Is the summary always very like dry language in the value proposition very flowery. Is there were from both types in both all of those types of considerations, editorial sort of style, if you will, comes in a lot when you think about content reuse. ‘Cause it’s not just any longer, hey, this term or this word or this image makes sense in context. But now it’s what are all the context that could possibly exist? And if I want to reuse pieces of information, they may now be in different contexts. We need to break them way down.

Everybody needs to agree on them. So that’s a solid like web example of how much work it can be. I think a lot of UX writer oriented UXers have attempted to really like control strings of copy, right? And doing that based on specific user behaviors or user interactions, it’s like a really fantastic idea to say if the user does this, input, this type of behavior in the text string that is otherwise consistent, it can really work, but it is very complicated.

The implementation has to be very thought through. You have to do lots of user testing to make sure that when people do things, it appropriately interacts with the content that’s being reused. And we, I think, have all seen it go wrong. So there’s a lot of opportunity for error there.

I think a second misconception that my clients often come to me with is that everything on their system should be built for reuse. That’s, I think, not true. I think it’s really important to be deliberate about what should be reused and what shouldn’t. In the same children’s hospital system example, they wanted to make their SEO content, which is to say like their meta title, meta description, that kind of stuff.

They wanted to structure that in the same way as they structured all their reusable content, even though none of that necessarily. Needs to be reused and in fact shouldn’t be. So things like that, not everything needs to be structured for reuse. Not all text on a page needs to be reused either.

You can have, there’s text that’s specific to a page. That’s not a bad thing. People’s names don’t always need to be reused. One person has, you know, only one context in it, and an instance. So I think being thoughtful about what should and shouldn’t be built for reuse is just as important as attempting to build things for reuse that are not currently reusable.

I think the last big one is that all of the journey towards reuse and systematizing within a design system, all that kind of stuff that all of that needs to happen all at once. In the Children’s hospital example, I actually took over the project from another colleague who had done the content modeling for other parts of the site.

So she did their condition pages and their treatment pages, describing conditions that a patient can come in with, and the treatments that the hospital offers. Which meant, because those pages referenced department names, she had to essentially create like shell content types for the departments themselves.

So like the title, the URL, maybe the contact information, that kind of just like very basic information, but none of the summary and value proposition and the four different names that I had to work with. But basically she did a very light model so that she had things to reference and then. You can come in later and build that out.

So that’s essentially where I came in and figured out what other types of reference fields exist and let’s expand things out and let’s bring in other data sources as needed. So it’s not a one and done system. It doesn’t all need to happen at once. Your content will evolve, that is okay. You can evolve with the content, it can evolve with you.

All that kind of stuff.

[00:23:45] Chicago Camps: What advice do you have for UX designers who want to work more effectively with content strategists or CMS developers?

[00:23:53] Sabina Leybold: I think there’s a couple things that UX designers can do on their own to work more effectively with content strategists for CMS developers, and that also, of course, in collaboration with those folks.

So first and foremost, UX designers who are like, I need to know more about content modeling. I need to understand more about. Structured content, how these things actually get built, how these things are managed.

My sort of handful of tips there are to constantly be asking yourself the question, how might the author have to enter this information? Or how might the author control this information? Similarly, you could ask yourself, what can we do to make this element easier to edit?

The nice thing about that is when you’re thinking about design systems and scale. Free text. Allowing people to enter whatever they want is not always great for design, and it’s also really not always great for the content strategy and the content management of it all.

So are there places where instead of allowing somebody to enter whatever they want, we can control it with taxonomy, right? That will make it easier for the UX designer to plan for what might actually be in that container. Thus, now how can I design that container? It also helps the author.

I see all the time just inconsistencies in content entry. So plurals on some things that not on others, and now they’re considered two terms, as an example. So allowing free text doesn’t always make this system easier to manage on their side either, which again has impacts back on the user experience. To me, we really cannot separate making a really solid author experience that allows an author to provide consistent content to a user from the idea of the design having consistency in and of itself.

Which is why I’m so passionate about UX folks, really getting a little more into the mindset of content strategy and content modeling specifically. But again, what can we do to make this element easier to add? How might the author have to enter this information? How might we make that better for them and more controlled for them, but also of course, allowing them freedom in places where it makes sense, too.

That’s why a lot of the questions and like considerations that I mentioned way earlier in this conversation about does there need to be a character limit? What validations need to exist? Should this field be required? They are not just making it easier for the author to be entering information correctly, but of course then for the designer to be expecting what information is gonna be somewhere and being able to design for it.

I think the other thing that I would really love to see more UXers do is just get exposure to different CMSs and other platforms that are managing the content that you are working within. So see if you can access like sandbox accounts or test environments. Do trainings and certifications on specific technologies.

Talk to a developer and have them share their screen on whatever it is that they’re building. Ask for a walkthrough on what it actually looks like to the person managing the content so that again, you can think about what it looks like to the user who would be receiving that. On the other end, I find that the best content folks and UX folks that I have worked with are people who at one point were the content marketer who’s actually entering the information, or they were once a web editor at an e-commerce company and they were just doing, basically rote content entry or managing production of a page specifically because they’ve seen it, they know what works, they know what doesn’t. They know what’s confusing and what’s not for that author experience.

So if you can get exposure to more CMSs that are built in different ways. All the better. That also means different technologies too. So I mostly do content modeling for Contentful and for Drupal CMSs. I have some experience with WordPress, and every time a client comes to me with a different CMS than I am used to experiencing, I often ask for credentials, a test account for theirs, or ask for a walkthrough just because it can really help to see it in action.

So those are things that UX folks can do really on their own. In terms of collaboration, I would say ask to be involved in reviews or, ask for a specific walkthrough of the content model and the other sorts of parts of the content strategy in particular. So taxonomy, development, that kind of stuff. So I find that honestly, a lot of that work is very.

Invisible because UXers are often so in the world of, here’s the layout, it’s a very like visual type of job people can imagine. I design pages and they know what that looks like when I try to explain what I do, so much of my work is in spreadsheets and whiteboards and conversations, and it’s very hard to make that tangible to somebody.

So I find that the work can be thankless, but we’re often the most passionate people about our work because it is so invisible. You taking an interest in what it is that they are building, what they’re working on, especially if you’re asking about things like content modeling that are in the weeds, it will, I think, A, make their day, but B, really show that you’re establishing a bridge between the UX and the content, which that bridge always needs to be stronger, I think. And also understanding how to read a content model in whatever form they do it. Like I say, I do mostly spreadsheets, but there are other systems that folks use. So just understanding what it is that you’re looking at.

I think the second thing on the sort of collaboration front is just looking for opportunities to connect your work with their work. So there’s a really solid experience that I had with a visual designer who would annotate at the components in Figma that she had designed, not just with design specs, like hex codes and font sizes and colors and interactive elements and whatnot, but also she would map which field in the content model was populating the information that was shown in the designed field in the Figma file. So really connecting the visual display to the content model that’s controlling what information goes there. And that was like one of my favorite collaborations on a project ever in my career because I just, I felt very seen. She felt very seen. We were able to really work together on how is this thing actually gonna come to life? Both on the visual and on the sort of content editor experience.

I mentioned earlier that the relationship between the expression of the content and the content as it is stored is not always obvious. So having that visual connection and that annotation and opportunities to connect the visual to the editing experience, making sure that is connected. That can go a long way to making sure that it gets understood by developers and built properly. So taking an interest in each other’s work is always key.

My last tip, frankly, is just don’t worry about getting it wrong. Don’t worry about asking silly questions, and that’s such like common advice, but most UX designers have never been asked to understand content structure. They’ve never been asked to develop it themselves or to think through it or to even read a content model.

So just have a beginner’s mindset. You could be doing this for the first time in your career, even if you feel like a rock star UX designer in a thousand other ways, and that’s okay. And you can just be humble and you can get it wrong. And that’s fine. And every person has gotten it wrong before you, and they will get it wrong after you.

So just not worrying about that is my last tip.