Strategic Content Work, Domain Models and a bit of RDF: A Dialogue With Larry Swanson
Working with content to connect people, and ideas is hard, takes time, resources, perseverance and above all devotion to staying focused on the very essence of connection: authentic experiences.
This is what Larry Swanson is after: authentic human-generated content and experiences. Working as a content architect, a content modeler, a UX architect, or an information architect, in his own words, Larry is doing something that sounds deceptively simple:
[..] thinking about you and what you need and reminding everyone involved to stay focused on that.
I first crossed paths with Larry in 2020, thanks to an introduction from Andrea Volpini of WordLift, who connected us to talk about Solid. Larry was curious about the project and subsequently we found out that we have a lot of common interests in the fields of content, authentic media experiences and the Semantic Web.
Now, just a few days before I meet Larry in person for the first time at Connected Data London – an event he plays a key role in organizing — I want to share with you Larry’s thoughtful answers to my questions about content modelling, architecture, and of course the place of semantic technologies in that landscape of digital content work and design.
Larry Swanson At A Glance
Larry is a Content and UX Architect who specializes in creating structured content models for omnichannel customer experiences and enterprise-grade solutions for content. He has more than 25 years of experience working at the intersection of technology and content, currently very much interested in the intersection of domain modeling and ontology design. Having been online “even before the web on Unix dial-up accounts and services like Prodigy and Compuserve”, Larry has applied his zest and curiosity for meaningful content experiences in projects that range from consolidating multiple CMSs into a single headless system, to developing semantic layers for hybrid-AI initiatives, to creating content models for knowledge-driven platforms.
Apart from working with clients, Larry is also hosting three (!) podcasts: Content Strategy Insights, Content+AI and Knowledge Graph Insights, walking each and every word from the slogan on his website: “Connecting content, people, and ideas.”
Meet Larry and his exhaustive answers to all the questions I had about the intersection between content modelling, design and architecture and the Semantic Web.
From Journalism Through Book Publishing To Hypertext
Larry, what is your background and what brought you to content matters? You have mentioned several times your mom’s editorial background, tell me more about that print world 🙂
I’m a “military brat,” so I moved a lot as a kid, which resulted in me becoming a sort of social chameleon, always able to get along with new groups of kids and teachers. I think that helps to explain why I’m so comfortable interacting across the range of business, creative, and technology stakeholders that we see in the content world.
My father was an engineer and my mother was a lifelong journalist (as was her sister), and I think that at least partially explains how I have ended up so often at the intersection of technology and content.
I chose journalism as my major in university not so much because of my family heritage but because it was the program that would let me study the most different subjects. That choice also highlights what I have discovered is my core personality trait – curiosity – which also explains my eclectic interests and generalist tendencies.
After college and up until the dot-com boom, I worked in book publishing, mostly as an acquisitions editor, which is a role much like product management in the digital age. I developed and executed product-line strategy, had P&L and operational responsibility for multi-million-dollar business units, and was responsible for signing new authors and keeping the publishing pipeline full. It was very rewarding work, but I left that field as soon as the web arrived. I had been online even before the web on Unix dial-up accounts and services like Prodigy and Compuserve and was excited about the opportunities afforded by digital information, hypertext, and an internet connection.
All Things Content
What is content architecture and what are its main tools?
The way I use the term, content architecture is simply a conflation of my two primary digital-era craft heritages: content strategy and information architecture. The main tools required for both crafts are: people and communication skills, behavioral research, conceptual modeling, implementation planning, and success measurement. Neither practice has access to anything like Figma in UX design, Jira in product and engineering, or the tens of thousands of SaaS products in the marketing tech stack. At the implementation stage, most content ends up in a CMS or similar system, but the design, strategy, and planning work is managed 99% of the time with whiteboards, Post-Its, collaborative documents, and spreadsheets.
Please, distinguish a content architect, a content modeler, a UX architect, and an information architect for me!
[Teodora: I can imagine how hard it is for you to explain your role. In my case, I still have a hard time explaining what a content writer does, let alone a knowledge graph manager :-)]
My definitions, offered with no small amount of trepidation, since few agree on what these terms mean:
- content architect: a role concerned with the overall structure of a content system – the people it serves, the workflows it needs to support, and the technology that powers it – always grounded in the business goals the the system supports and guided by the overarching management and content strategies
- content modeler: a role concerned with identifying the elements in a content domain that need to be understood and structured for multiple uses, enumerating their properties and attributes, describing their interconnections and relationships, and documenting these discoveries in ways that help the engineers who build the system and the content, design, and other system users understand it
- UX architect: a role concerned with bringing the practice of information architecture to UX design projects and programs so that designers and adjacent stakeholders have a clear understanding of the technical and design objects they are working with and how they interact
- information architect (in the context of content management, as opposed to software engineering and other crafts): a role concerned with labeling, categorizing, and organizing content so that readers and users can navigate and search it
If you were to go to your niece’s school to present your work, what would you tell the kids?
Please get out your phone, tablet, or laptop and go to your favorite social media app or website. See all of those numbers and words and pictures and videos and how they all make sense and inform and entertain you and help you buy things (with your parents’ permission, of course) and share stuff? I work on systems that help designers create those screens and experiences. There’s also a bunch of other stuff that happens before you see those screens, and I help the engineers who build those systems understand what they have to work with and how it fits together. Most importantly, because all of those people have lots of other stuff on their mind, I’m always thinking about you and what you need and reminding everyone involved to stay focused on that.
What are some of the seminal books that changed your perception of how you work with and design content?
There are literally hundreds I could mention, but a few that have deeply influenced me:
- Computer Lib/Dream Machines by Ted Nelson introduced me to hypertext and its infinite possibilities, as well as important concepts like transclusion and intertwingularity
- Being Digital by Nicolas Negroponte helped me start to think about the many ways in which content could be shared independent of a physical artifact like a book or magazine
- Information Architecture for the World Wide Web by Lou Rosenfeld and Peter Morville helped me tame the organizational anxiety I felt as I was building content systems in the late 90s
- How to Make Sense of Any Mess by Abby Covert further calmed my info anxiety with its succinct and more visual approach to IA
- Content Strategy for the Web by Kristina Halvorson gave me a name and an intellectual foundation for the strategic content work I’d been doing for years
- Content Strategy: Connect the Dots between Business, Brand, and Benefits by Rahel Anne Baiie and Noz Urbina helped me better connect content practice and business prerogatives
- Thinking in Systems by Donella Meadows gave me a conceptual framework for the CMS and other system-design work I do
- The Back of the Napkin by Dan Roam helped me develop the visual communication skills that are so crucial in conceptual modeling work
- The Core Model by Are Halland was only published last year, but I have been using this elegant content framework since I first learned about it 10 years ago
- Semantic Modeling for Data by Panos Alexopoulos, my current reading, is inspiring me as I discover even more connections between semantic and content practice
And last in this section, Larry, please elaborate on what you have shared in a private conversation about having a core content model, like events, people etc. with few, I will call them properties – correct me if I am wrong, and then a model that is unfolding, based on the use case, that is on the informational needs of the user. How are thse two different? And what should be account for when building a content model with that mindset.
Like most ontologists, I start any project – in my case, usually a content modeling project – with a domain scoping exercise. That exercise always involves both a 1) top-down approach that considers the project’s domain, the organization’s priorities, and an understanding of its current architectures and systems and a 2) bottom-up approach that looks at instances of the things, both native domain entities and the content and information assemblies about them, as well as the language that people use to describe both those things and their properties and attributes. The result is a domain model that informs the construction of the content model. These domain models almost always include common, high-level concepts like people, organizations, and events, as well as other concepts that are unique to the domain.
The content and information assemblies that my content models deliver must consider both 1) the purpose of the content elements that are modeled and 2) the intent expressed by the users and customers who consume the messages and content experiences built with them. It’s that consideration of both content purpose and user intent in your content modeling, along with a solid metadata strategy, that enable efficient content re-use and the orchestration of truly user-centered content experiences.
All Roads Lead To …RDF
When did you first meet the Semantic Web?
I used to subscribe to the print edition of Scientific American, and I read Ora, Jim, and Sir Tim’s famous article as soon as I received the magazine, probably April of 2001 (back then subscribers typically got the magazine the month before the publication date). I re-read it at least once a year.
Have you worked with RDF for a client?
Alas, no. I’ve done some personal projects as I’ve studied SPARQL and OWL, but RDF is still very scarce in the CMS and content systems worlds.
How can the processes of content strategy be aligned with the processes of expressing information in RDF?
SEOs have done this for years in the content marketing realm with RDF linked data, but it’s not that common in other parts of the content management world.
There are a few folks I know of who have done work that combines content strategy and RDF-formatted information. When he was at the BBC, Mike Atherton did a lot of systems-level work with content and linked data. Aaron Bradley did some very interesting content personalization work with knowledge graphs when he was at Electronic Arts. More recently, Michael Iantosca’s DOM Graph RAG work at Avalara has highlighted how the RDF metadata and DITA content W3C standards can work together. Places like Amazon and Netflix have big teams of ontologists, and I’m pretty sure they’re using RDF, but I don’t know how they interact with their content strategy peers.
All of those examples point to the ability of RDF to elevate your content metadata strategy. I haven’t used RDF per se in my content modeling work yet, but the ontological thinking that my study of semantic web practices has given me has improved my practice a lot.
Could it be that we can put in practice the knowledge from the Polar Bear book (i.e.Information Architecture for the World Wide Web ) and the book Semantic Web for the Working Ontologist ?
That would be a fun and rewarding mash-up! I’ve been doing that for the past five years in my content modeling practice, but I’d love to more intentionally and deliberately collaborate with people like the folks I mention above as well as others like Jeff Eaton, Karen McGrane, Cruce Saunders, Carrie Hane, Noz Urbina, and Rafaela Ellensburg.
All Things Content In Action
You and Rahel and Cruce are working on an alliance, can you tell me a little more about that. Are there any standardization fights related to this? Or Planned?
The current team is Rahel Anne Bailie, Cruce Saunders, Deane Barker, Hilary Marsh, and me. We have created the Kinetic Council as a formation entity for a proposed new “big tent” association for professionals across content, data, and semantics disciplines. The likely scope will include programs and activities around:
- professional networking and community building
- education and certification
- industry leadership and advocacy
- career advancement
We’ve been working on this for well over a year now, and Rahel and others had worked on a similar initiative before. We’ve had a lot of conversations at gatherings like Semantics, Lavacon, and Janus Boye events, and the feedback has been very positive. We just need now to find the time in our busy schedules to formally establish the group and take the conversation out to a bigger audience.
If folks would like to follow us or get involved, there’s a sign-up form on the Kinetic Council website.
You are working on two podcasts, bringing together awesome professionals of various disciplines, tell me more about the discipline behind of sustaining such a dynamic and intense content workflow.
I started Content Strategy Insights seven years ago. I knew I needed to be doing more to regularly share my discoveries in the content world. Ironically, for a lifelong word nerd, I hate writing, at least under deadline pressure. But I do love a good conversation, and Zoom made it easy enough to capture the conversations I was already having. So I settled on an interview-format podcast, and set out to bring a little more rigor and curation to my chats. That has turned out to be a good decision. It’s definitely a lot of work, but I have streamlined my workflows and am able to spread out my workload when client and community obligations take over my schedule.
Do you think that podcast rhythm is replicable, scalable in an enterprise?
In theory, and speaking only to the craft-democratization approach that I take, absolutely yes, but in practice I don’t think any enterprise would ever dedicate the resources it takes to sustain a long-term program like mine. When I add up all of my prep work, interviewing, and production time (not including book or manuscript reading, which I would do anyway), it takes me about 4-7 hours to produce each episode (which explains why I drop back to one episode every other week when I’m on a full-time contract). I can see why managers would object to allocating that much of a practitioner’s time to what looks like a side project, and I can imagine colleagues taking exception to someone getting to do such fun work. This, of course, assumes that it’s a one-person project. It’s fun to imagine a programmatic approach to this, an enterprise professional-development program that shares its practice discoveries in a podcast or similar format. Again, though, there are few incentives for any one organization to do that.
What do you think about those AI generated podcasts?
They’re clever, and sort of interesting, but I abhor them, along with the other attempts of the tech plutocracy to replace our humanity with technology. When it comes to the role of tech in society, I’m firmly in the Doug Engelbart camp. Technology and computers should augment human intelligence, not try to replace it. When it comes to conversation, I vastly prefer actual one-to-one human connection, hence my regular podcast interviews with actual human guests.
Elless media – where Elless comes from?
My initials are LS. Phonetically, that can be pronounced ell-ess. Several people have noticed this over the years. My favorite example was when Deane Barker and Corey Vilhauer figured it out from reading my brand name aloud on a Zoom screen during a podcast interview 🙂
Who’s gonna interview the interviewer?
Teodora: This is a rubric where I ask the people in the Dialogue to spare several minutes and ask me a question or two. In this case, the answer wrote itself, following my belief that text writes itself, it is not us writing it. That said, Larry and me some time ago, went down the content and concepts meanders in one of his podcasts: Teodora Petkova: Dialogic Communication for the Semantic Web – Episode 7
Instead Of An Epilogue: The Future Of Content
Larry, what’s the future of content and why is it so hard to see it from the trees of day-to-day content operations?
This topic is definitely in the air. I help organize a new meetup in Amsterdam and when we had to name it we ended up calling it (shout-out to Claudia Mueller for the idea) The Future of Content. A couple of months ago at Lavacon, Jack Molisani invited me to join the conference’s closing panel entitled “The Future of Content.” There’s also a podcast by the same name. Apparently a lot of people are thinking about this.
My prediction, maybe more of a hope, is that content and content practices will become even more important as misinformation and disinformation sweep across the social, political, and business worlds. In a world where the default assumption is that any content you see is fake, I think people will come to crave authentic, human-generated, factual content. There’s obviously a lot of work to do on that front, but content people and ontologists are the right folks for the job.
With that I will leave you to Larry Swanson’s website and LinkedIn profile links, for your intertextual pleasure 🙂
For more modelling and semantic web adventures, you might also want to dive in Music, Taxonomies and What Defines a Sandwich: A Dialogue With Bob Kasenchak | Teodora Petkova