Bram Wessel is a co-founder of Factor, a Seattle-based consultancy that brings user-centered design principles and practice to enterprise-scale information problems. In this episode, he breaks down the types of taxonomy problems many large orgs face, including problems with findability, site search, and navigation. Bram’s work aims to cater to how customers think, applying user research to ensure everyone is speaking the same language through information.
With more than two decades as a user-centered design and research professional, Bram Wessel believes that technology should enable natural experiences for real people. Bram has developed human-centered digital experience designs and strategies for such high-impact brands as Adobe, Amazon, Backcountry, the City of Seattle, Crate & Barrel, Disney, Expedia, GE, Group Health, Intel, Lionsgate, MasterCard, Microsoft, Nordstrom, Real, Safeco, Sony, Starbucks, Volvo, UW Medicine, and Warner Brothers.
Bram is a founding partner at Factor, an information and interaction design consulting firm. He is a member of the Information Architecture Institute (IAI) and the Seattle chapter of the international Interaction Design Association (IxDA).In his spare time, Bram enjoys fly-fishing. shellfish farming, and making wine.
Kristina: Hello, my friends. Welcome back to The Content Strategy Podcast. I’m your host, Kristina Halvorson. Each episode, I bring someone in from the field of content strategy who is doing amazing work. This week, I asked Bram Wessel, the principal and founding partner of Factor, to join me. We’ll get to our awesome interview in just a minute, but first I have two things to share with you—and they are both favors. That’s right. I need your help. Yes, you.
Here’s the first one—and it’s a super fun one—I am going to be doing a listener Q&A episode sometime before the end of the year. And in order for me to do a Q&A episode, I need your Qs. So, if you have a burning question about content strategy (or even not a burning question—maybe it’s a lukewarm question, but it’s a question nonetheless) what I would like to have you do is call the fancy Google Voice mailbox that we have set up just for you and this purpose and leave me your question. I’m going to give you the number right now and then I’ll also put it on the podcast homepage, which is ContentStrategy.com/Podcast. The number is (510) 858-6927. And if you can be the tenth caller—I’m just kidding. I’m not giving any special privileges to the tenth caller. Call and leave your message and hopefully you will appear on the podcast.
Here’s the second thing: I need four to five minutes of your time to help complete a survey for the podcast because, my friend, I want this podcast to be extra special and amazing just for you. So, could you please visit ContentStrategy.com/Survey and take a few minutes to tell me all your hopes and dreams and constructive criticism—yes, I’m asking for that, too. That information will be enormously helpful as we continue to shape and grow the podcast in the months to come.
Thank you so much. I really appreciate you. All right, here’s Bram.
Kristina: Welcome back to another episode. I’m Kristina. So delighted to have you here. I am equally delighted to have today’s guest, Bram Wessel. Bram is a Seattle-based experience strategist and cofounder of Factor, an information and interaction design consulting firm. He’s been practicing experience design and strategy for 25 years with a particular focus on keeping the humanity in human digital experience—which I love. Bram, welcome to The Content Strategy Podcast.
Bram: It’s delightful to be here and the sun is shining in Seattle, Washington, but by the time your listeners hear this, it will probably be back to our four hours of gray twilight.
Kristina: Every time I’m in Seattle, it’s sunny. I’m not going to lie to you. I am starting to believe—
Bram: You live a charmed life.
Kristina: People in Seattle just tell everybody else it’s raining and dreary all the time so that we’ll stay out.
Bram: Huh? I might know something about that.
Kristina: Do you? All right. I’m going to get to the bottom of this, Bram. Wait, was this the first time we met at the Enterprise Experience Conference this spring?
Bram: Not really. We’ve been at a lot of the same events. We have a lot of mutual friends. I think we probably met one—
Kristina: That’s when we really sat down and threw down about content strategy.
Bram: Definitely. That’s the first time we’ve ever had an in-depth conversation. I think you came to Seattle on a book tour six or seven years ago. That’s probably the first time—but we didn’t really meet at that point.
Kristina: Now I think you’re just being polite because my suspicion is that I have met you 22 times. Don’t tell me if that’s true— even though I know it’s probably true.
Bram: Oh, I’ll keep that quiet.
Kristina: You do it. Let’s just high five. Hey, Bram, tell me about Factor.
Bram: Well, Gary Carlson was a taxonomy consultant. He had previously worked at a taxonomy tool maker called SchemaLogic. It’s now part of a company called Smartlogic. I was an interaction design experience strategy consultant. We had known each other from a past life because we went to college at about the same time at the same school. Gary called me up one day and he said, “Hey, I’m doing this project. This is the kind of work I do and I need some help with user research.”
I did some user research for him. We talked to golfers. The client was a string of golf courses in the Seattle area founded by Scott Oki, who is a Microsoft alum. We were really happy with how it turned out. Gary had been lamenting that there was not enough of a user-research driven focus in the work that he was doing for his customers. And I was encountering more and more information-dense challenges.
We did a few projects together. We did a big project for UW Medicine and it occurred to both of us that we should join forces. We cofounded Factor in 2013 explicitly to focus on this challenge of bringing user-centered design principles and practice to enterprise-scale information problems. That’s what we do. We’ve been doing it now for almost seven years and it’s going really well.
Kristina: Bram, I have a lot of different folks in the audience for our podcast. Some of them may not know what “taxonomy” means. Would you do me the favor of just giving a little insight into the practice of taxonomy: What it is. Why we have it. Why it’s important.
Bram: Sure. A taxonomy, at its very basic form, is a list. A lot of taxonomies have terms. They have concepts. If you’ve ever heard of a controlled vocabulary that just means, “Okay, we’re going to say that this is what this list contains and what it doesn’t contain.” And maybe there’s a subject that unifies all the things in that list. The things in that list can have relationships with each other. You can have a tiered, hierarchical list that has parents and children, or you can have a bunch of different lists that you interconnect. That’s sometimes called an “ontology,” which is a word that we only use if our customer uses it first.
In e-commerce, for example, this is how taxonomy manifests itself: You have a customer-facing product taxonomy. That’s a list of all your products. There may be categories and subcategories within that list that apply to the categories, subcategories, concepts, or terms. In this case, products in that list. All those things really make up the taxonomy of a company’s products. What we do at Factor is try to make sure to model these taxonomies in a way that customers can understand—that reflects the mental models that people have in their minds about this set of products or this subject matter or what have you. That’s, at its basis, what taxonomy is and how it gets applied to the work that we do.
Kristina: What happens if the taxonomy sucks or if there isn’t a taxonomy? Can you give me some scenarios in which we should have a taxonomy when it comes to dense information sets but we don’t?
Bram: A lot of companies have accidental taxonomies. They have lists and groups of things and categories—sometimes they even have attributes—but nobody has sat down and figured out, “Oh, okay, what’s the best structure for all of this stuff?” It just evolves organically in the organization. When a client or a customer comes to us with a situation like that, we take a really deep look at it and try to expose some of the structure. We also do user research to figure out how that company’s customers, or how that organization’s constituents, think about this information so that we can remodel the taxonomy based on how they might think.
If something sucks, you’re going to have a lot of bad side effects. Navigation might fail. Findability might be difficult. Search might not work. A really good example from some of our work in the past is a company that many people are familiar with: Crate & Barrel. Crate & Barrel had a really well-optimized process for what happens after you put something in a shopping cart. Great. They knew all the inefficiencies in that funnel and could have an A/B test and usability study regime that would allow them to keep that really well functioning. What they didn’t understand is why people couldn’t find products to put in the cart in the first place. The first rule of e-commerce … I’m just picking on e-commerce right now—
Kristina: It’s so easy.
Bram: People can’t buy products they can’t find, right?
Bram: Yeah, right? Crazy talk. What we worked on with Crate & Barrel was pre-cart findability: What are some of the reasons why people can’t find things? One of the things we discovered is that the attributes they had were either incomplete, not detailed enough, or some of the products they had were misattributed. By remodeling their attribute sets and cleaning up some of the poorer tagging quality of the attributes, we were able to increase product findability. They could then start to measure that part of the funnel—everything that happens before a product gets carted. That’s an example. There are some other examples from search and navigation we can get into if you want to, or we can move on to another topic.
Kristina: Let’s do it. Examples are great. Let me be perfectly transparent: My areas of expertise are not in taxonomy. The first person to ever explain taxonomy to me was actually Rachel Lovinger many, many years ago. Whenever somebody comes to us and says, “We’re interested in taxonomy.” I go, “Uh, I feel dumb.” This is actually very helpful for me to hear different ways that people explain it. A big part of why I wanted to talk to you about it is that you come at it with that user-centered focus. It’s not just, “Okay, let’s dig into meaning.” It is really linking basic UX and IA principles: “Well how do people actually think about these things? Let’s do the deep research to find out.” Lay the examples on me. I want to know the different ways in which taxonomy is important, how it helps, and how it can backfire.
Bram: Sure. One of the projects that we did recently was a large auto parts manufacturer that many people have probably heard of. They brought us in to help them find a new search engine. Once we got there, we discovered that the search problems they were encountering—searches being unsuccessful, searches being abandoned, poor clickthrough rates on search engine result pages—those were actually a trailing indicator of the real problem, which was that the information that the search was looking at—the index—was just not very sophisticated or complex enough to reflect the true complexity of the product line in a way that was useful for their customers.
We looked at search logs and we did some user research. We looked at search logs as the starting point to see where vocabulary was not matching with search results. Then we did some user research to figure out what mental processes were going on that were leading to this experience not performing as well as they wanted it to. One of the things that often happens is that search engines can be very finicky and you have to teach them what in the taxonomy world we call our “preferred terms” versus “alternate terms.” Another way of saying that is “synonyms.”
If you don’t, as a user, type in exactly the right word and therefore can’t find something, you need to teach the search engine, “Oh, there are alternate words that mean similar things.” There are other signals we can look at that can help us figure out what the user might need if they aren’t using the word that exactly matches what the search engine believes that concept represents. Another example of this is from some work that we did with Adobe. We’ve done a lot of user research with Adobe. We’ve tried really hard to understand how users think about Adobe products and the concepts within them.
Adobe products are notoriously high-learning curve. They’re difficult to really achieve basic competency or mastery with—like Photoshop. Over and over and over again, we had participants in our user research study say, “I don’t have the words. I don’t know the word for that. I don’t know what the Adobe word for that is,” when they tried to find support content to help them figure out how to solve a problem or how to do something in Photoshop. That’s really the crux of the issue: What kind of vocabulary do customers have and how do you tune your information infrastructure to be responsive to the vocabulary that they have?
Kristina: It’s got to be a shifting vocabulary too, right? They’re still going to have 50 different words that they use to describe what it is that they need.
Bram: Right. It’s exciting right now because we can start to teach machines to interpret a variety of different signals and learn what some of the alternate terms might be on their own. We can use natural language processing and machine learning to help search engines and other digital experiences get smarter over time about vocabulary. That’s hopefully never going to replace the need for a good librarian or an information scientist who really understands information structure well and can curate it. But we can harness the power of those machines to make our experiences a little better.
Kristina: That’s AI. That’s a little scary—it’s not really. We kind of were chatting a little bit about information management in the enterprise environment. You talked about how we’re starting to teach machines to recognize these different nuances within language and so on, but what are some common challenges that you see within enterprise information management? What opportunities for solutions are you excited about in terms of the work that you’ve been doing?
Bram: Yeah, that’s a great question. We encounter a lot of organizational misalignment that is reflected in the information ecosystems of our customers. A really good example of this—let’s say you’re a big semiconductor manufacturer and you have a marketing budget of, oh, I don’t know, $2 billion dollars a year—
Kristina: I want those people to call me for content strategy.
Bram: Perhaps we can arrange that.
Kristina: Yeah, right.
Bram: You need to understand the effect that your marketing campaigns are having on sales. The problem is: what’s a campaign? You ask seven different people and they give you seven different answers or they say, “Well, I mean everybody knows what a campaign is, right?” If you probe a little more deeply, nobody really knows. When it comes time to assess and measure the impact of marketing on sales, if you didn’t know what a campaign is, how are you going to do that? In the case of our customer, it took a mob of people and weeks and weeks to come up with an approximation of what the impact might be. If you’re spending $2 billion dollars a year on marketing, that’s a hundred million dollar a quarter kind of problem.
The reason that this situation existed is because nobody had gotten together and said, “As an organization that does campaigns—because we’re a marketing organization—we’re going to define a campaign. We’re going to define all the different types of campaigns that we do. We’re going to define the core, structural elements of a campaign. And we’re going to give campaigns unique IDs that can travel with them through all the systems, processes, and operations that we have so that we can compare apples to apples instead of oranges ... and pineapples ... and grapefruits as we track the progress of our campaigns throughout our organization.”
That’s a situation where multiple, different accidental taxonomies have led to this kind of Tower of Babel and an organization can’t really measure how well it’s performing. That scale of problem is really exciting for us. It’s fiendishly complex because, in this case, we had about 30 taxonomies—or 30 different systems and 25 taxonomies. I don’t even remember where it was and it’s only increased since then. We had to normalize all that stuff. We had to do that not just with campaigns, but with hundreds of fundamental concepts that are really important to the business that they need to be able to measure. To me, I think every organization is waking up to the reality that their information is a core strategic asset that has tangible value. In a lot of cases, it’s a big mess.
Kristina: Well, I think the other big mess is the way in which people are—and this gets a little bit meta—but the way in which people within the organization are communicating about that information, right?
Kristina: Isn’t that ultimately one of the huge purposes of taxonomy—not just to make things manageable and findable, but make it so people can literally be speaking the same language when they’re talking about what they need to know, or what they need to accomplish, or where they need to collaborate, or how they’re going to measure?
Bram: Right. Absolutely. As more and more companies go through digital transformation, that communication is conducted in systems or via digital means—not just email and Slack. You’re actually using systems to communicate between groups. A lot of that is semi-automated. Every organization has an information supply chain. The first place I ever heard this was Nate Silver speaking at the 2013 Information Architecture Summit in an interview with Lou Rosenfeld. Every organization has an information supply chain. Nate was talking about the information supply chain of pulling data and election forecasting, but the larger your organization is, the more stratified or fragmented or distributed your information supply chain is going to be. Everybody needs to define their terms and have a common vocabulary just to talk about all of that information.
Kristina: Talk to me about this word “content.” You have been talking about information systems, information complexity, managing information, communicating about information, and information architects. Somehow, within the realm of content strategy or content design, content became the word that we talked about. When you’re in organizations, are they differentiating between what they’re calling information systems and content systems or content ecosystems? Can you unpack that a little bit for me?
Bram: Yeah, absolutely. There are a lot of people who have defined content much better than I’m going to. I’ll just say that for our purposes, content is information that is grouped to have meaning for a specific purpose. In a lot of cases, what we’re modeling is both the full taxonomy, the full ontology, the full information model, but what it’s addressing is pieces of content. For example, going back to Adobe, they had learn, help, and support content that had kind of evolved organically. It was a huge mess. They didn’t have any kind of governance over who could create content for what purpose and didn’t have any redundancy checking built in. They didn’t have any way of describing that content that would allow them to figure out when things were redundant. The work that we did with them was really about creating a content strategy that was driven by taxonomy and attribution so that there was a system of meaning.
We were describing the content of the content. We were saying, “Okay, we’ve got all these videos. We’ve got all these articles. We’ve got all these little fragments of texts that are floating around. We’ve got images. We’ve got interactive tutorials.” But what is being talked about or being addressed inside all of these pieces of content? We created a model—a content model inspired by the work of people like Rachel, Karen McGrane, and Sophia Vochesky Prater—which defined the content of the content. You had products. Products appeared in all of these different content types. You had tools and tools within products. You had problems. You had questions. You had step-by-step techniques. By creating this model to describe what was inside all of this content, we could assemble it in more sophisticated ways.
When I think about enterprise content strategy and I think about content, I think that this is also a universal challenge that a lot of companies have similar to what I described earlier with the information ecosystem and supply chain. Companies and organizations are waking up to the fact that they’re not just in the business they thought they were—whether that’s selling widgets or providing services to constituents. They are also in the content business. They have all of this content that they need to provide really rich, digital experiences for people to access and use. By creating taxonomies and ontologies that describe all of this content, we can give that content handles.
We can say, “All right, we’ve got an attribute set that allows you to manipulate this in a certain way. We’ve got relationships between different pieces of content that allow you to order them or arrange them in a certain way.” When we’re interacting with content, that’s what we’re focusing on. There’s almost no case in any organization or company that we work with where the content that they’re managing, the content they have—even if they haven’t really thought very carefully about it—there’s almost no case where that isn’t part of our work. The information model that we’re building needs to address content.
Kristina: As you’re talking, I keep coming back to what you were saying about the accidental taxonomies that pop up everywhere—whether it is sets of information or pieces of content. The work that you do is to create order to that. When you’re talking about making choices around sets of content, I’m thinking, “Okay, well somebody—whoever is in charge of that or whoever’s doing it—has got to have some kind of information management training or background. How do you maintain an established governance for this stuff? Once you’ve got things sorted, how do you keep them sorted?
Bram: Right. Two things here. We encounter situations where there’s a team of full-time taxonomists. That’s one end of the spectrum. Then there are other situations where nobody has any kind of information management or information strategy expertise. In both cases, usually we are brought in because we have an independent, outside perspective. Whether it’s the expertise that they need or the independence, the role that we play is to take an unaided look at what exists, help people create clarity around it, and create order around it.
The second part of your question is governance. Governance is one of the things that is a huge part of our capability set. We do governance planning and create governance strategies. In governance, the challenge is figuring out what is the best model for basically change management. You’ve got a set of information. How often does it need to change? What’s the procedure for changing it? Who needs to be consulted when it’s changed? In highly regulated environments—we have some customers in healthcare—there are all kinds of rules and regulations imposed by the government, like HIPAA, about what can and can’t be touched. We need to factor that into creating a governance framework that’s flexible enough so that the information is usable and you can manipulate it and make changes to it, but rigid enough so you don’t have people making changes willy nilly and degrading the quality of the information and the content.
That’s the situation that Adobe had. There were too many people with too many rights to be able to create content anytime they felt like it. As an organization, they needed to come together and establish a consensus about what the ground rules were for who could create what, when, and why, and create an editorial loop so that people with eyes on the entire set of content could determine if it was fitting into the overall content strategy. Governance is the most non-one-size-fits-all thing there is. Every organization has different needs when it comes to governance.
As a consulting practice, it’s not really that repeatable. We just need to get in there and see how the organization is structured, what some of the information workflows are—whether that’s content publishing or whether that is making changes to a core vocabulary—and then figure out what governance—whether that’s rigid or whether that is loosely coupled—needs to be for that organization. The real key is that we don’t want something that’s brittle. Brittle and rigid are not the same thing. You can have a semi-rigid governance framework without it being brittle. If it is brittle, you’re not going to be able to make the changes that you need to in the amount of time that you need to to support whatever organizational functions the information needs to support.
Kristina: Isn’t that interesting that Factor does human-centered taxonomy per se, but that you’re also in the business of governance. Almost like, “Look, we can give you this work, but if we just walk away, it’s not going to make any difference. We want to stick around and help you put a framework in place to make sure that it can be sustained over time.”
Bram: Absolutely. Yeah. Sustainability in our world means: How durable is the information model that we deliver you and how maintainable is it? What we find often is that we are organizational alignment consultants. We’re going into an organization and we’re figuring out where all of the centers of power are. We’re helping those organizations—through sort of the side door of their information—restructure their power center so that they can use their information more effectively. This is one of the meta things about the work we do. We find that when we’re at the level of information, there’s a lot of power involved in that.
When you start to make changes to that, you run into a lot of conflict and a lot of friction. Once you resolve that, there’s great power in the flexibility that it gives you. In terms of sustainability, another thing that we do is implementation consulting, where all of these information models that we deliver need to get implemented somehow. Even if they’re just managed in a spreadsheet, they’re probably going to need to get implemented in a content management system or a DAM, or maybe in ERP or a CRM. There’s acronym soup when it comes to these big enterprise class systems. There’s a million small details because each of those pieces of software have a different set of modeling capabilities—ways that the information can be modeled and structured.
Another thing that we do is—I just had a phone call about this earlier today—help organizations get a tool that’s specifically designed for that. There’s a handful of tools—Synaptica, PoolParty, Semaphore, TopBraid—that are called taxonomy or ontology management tools that have really sophisticated modeling capabilities and real-time API-level integrations. All of this information can be shared across these systems really, really efficiently and don’t have to be imported and exported by hand with a spreadsheet.
Kristina: Bram, what kind of person do you have to be to do the kind of work that you do? What are some of the personality traits?
Bram: Well, I was an English major in college. I’m really fond of wine and fly fishing and Indie rock. I don’t know—
Kristina: I did not ask for your Tinder profile, Bram.
Bram: I don’t know if there’s a personality type. We have people who have come from a lot of different backgrounds working for us at Factor. We have somebody who changed careers after working for a medical device manufacturer. We have people who have library science degrees. I think you have to have a passion for really, really wicked problems. We tend to go really rigorous on the work that we do. That’s one of the personality traits: You have to be curious. You have to be really interested in human behavior and how people think in psychology to be interested in this work. I think you have to not be afraid of scale and complexity. You have to think of those as fun challenges rather than daunting ones. I don’t know how I managed to get to that place. I think just doing it for years and years led me in that direction. I’ve always kind of been fascinated by technical, complex things.
Kristina: Well, I think anyone who is interested in technical, complex things and comes at them through the lens of making people’s experiences better when they are encountering that thing—whether it’s an information asset or a product or whatever—those are the people who are helping lift up our discipline and our shared practices throughout content and UX and so on. I just think you’re one of those people. That’s why I was excited to have you on the show. Thanks a lot. Thank you for joining me. I appreciate it.
Bram: That’s kind of you to say. It’s been my pleasure.
Kristina: Bram, where can people find you online?
Bram: FactorFirm.com. That’s all one word. F-A-C-T-O-R-F-I-R-M dot com. All of my profiles are boringly just my first and last name. It’s Bram Wessel, B-R-A-M-W-E-S-S-E-L. On Twitter, on Instagram, on LinkedIn.
Kristina: Well, aren’t you everywhere?
Kristina: Great. Well, I hope the sun continues to shine for you, Bram.
Bram: I think it will. I think it will.
Kristina: That was the best closing line I’ve had on all of my podcast episodes.
Kristina: Yeah, I know. It’s a low bar. All right, thanks.
Bram: Thank you so much.
Kristina: Thank you so much for joining us today. This podcast is produced by Brain Traffic, a content strategy consultancy, and makers of fine conferences and workshops. Please visit BrainTraffic.com for more details and sign up for our mailing list to hear about new workshops, dates, and locations, as well as content strategy insights and little personal notes from me with hilarious jokes.
If you enjoyed this episode, please share it with a friend or leave us a review on your favorite podcast catcher. Our podcast is produced by Podcast Press. Transcription services by Rev.com and Heather Heigel. Show administration by Bailey Miller and Amy Pletch. Show art and music by Sean Tubridy.
You can find even more episodes including transcripts and links to resources mentioned in the episodes at ContentStrategy.com. Thanks, we’ll see you next time.
The Content Strategy Podcast is a show for people who care about content. Join host Kristina Halvorson and guests for a show dedicated to the practice (and occasional art form) of content strategy. Listen in as they discuss hot topics in digital content and share their expert insight on making content work. Brought to you by Brain Traffic, the world’s leading content strategy agency.
Follow @BrainTraffic and @halvorson on Twitter for new episode releases.