Transcript

Hello, I am recording this from my home office, right outside of Boston, where the temperature is way too hot. And I haven't had a haircut in six months.

So first I'd like to thank Steve Fisher for having me here at the Design & Content Conference. I wish we could have all been together in Vancouver, but that's not the way of things right now. And I want to thank all of you for showing up despite the end of the world. So welcome and hang in there.

So today I'd like to talk about how we use categorization as an organizational structure on the web and the impact that our organizational choices have on the people who use our websites. We're going to start by talking about Walter Plecker.

In the 1920s, this guy was the registrar of Virginia's Bureau of Vital Statistics, a government agency that controlled birth records and marriage certificates and things like that. And he sucked. He was the worst, real bad, real bad. He was an atrocious white supremacist, and he was particularly terrified of interracial marriage. He believed that most interracial marriage was the result of wrongful categorization. White people were marrying nonwhite people only because the government hadn't correctly identified them as such. So he decided he was going to change that using the power of bureaucracy. He eliminated Virginia's racial categorization system and left just two categories. So one very specific group of people were labeled white and everyone who fell outside of Plecker's very narrow definition were not. And then a bunch of racist laws ran wild with the results.

And suddenly thousands of Virginians' lives turned completely upside down. The government saw them differently and treated them differently. They no longer had access to the same public spaces, the same schools, the same services and safety nets afforded to white people. Marriages were invalidated. Children were taken away from their parents. Indigenous people could no longer legally identify themselves. Virginians lost agency over who they were all because Walter Plecker changed a label.

So changing a label is a design decision. In this case, it was one calculated to disenfranchise specific human beings. And today, most of us don't have Walter Plecker's job. Instead we're designers, information architects, content strategists, product managers, copywriters, we build interfaces and products and apps. We help people read the news or register for classes or order takeout. It seems unthinkable sometimes that our work could carry the same weight as the Bureau of Vital Statistics.

But of course, I'm here to argue that it does. We work on the web where, you know, the entirety of human knowledge lives. So that means we control how people find, understand, and use information in every facet of their lives. That's an enormous amount of power and responsibility. You may already be familiar with this quote from information architect, Richard Saul Wurman, and it's such an important quote because it is the core of our work. The creative organization of information creates new information. Whenever we organize information, when we categorize it, when we arrange it, when we label it, when we connect it, when we link it, we are altering it. We are changing how that information will be perceived. What we sometimes forget though, is that our organizational decisions don't only have the power to create new information. They can also erase information, or maybe this is a better word. Creative organization is a necessary and wonderful thing, but we need to remember that it can be used for evil just as it is for good. If we are not thoughtful about our organizational choices and their potential impact, then we risk causing harm.

So how do we make sure that we're being thoughtful? How do we make sure we're thinking through the unintended consequences of our categorization? So let's take a step back. When we talk about categorization, we're talking about a type of organization in which we create groupings based on shared characteristics. From a very young age, we're taught to identify patterns, to notice how things are similar or different from other things. It's a really fundamental way of ordering and understanding the world around us. Take food, for instance. We have so many different ways of categorizing food. We have fruits and vegetables and meats and grains. We have breakfast foods and dinner foods and desserts. We consult cookbooks or recipe websites that might be organized according to season or main ingredient or flavor profile or method of preparation. So there are lots of ways of thinking categorically about food. And then there's grocery stores. Now, at a high level grocery stores all follow some general mental models, different types of food are grouped together. We can be pretty confident that we'll find bananas near the apples, which will also be near the spinach. Produce is a nice, stable category. That seems pretty easy to define, but that stability doesn't necessarily apply to all product categories. And a lot of grocery stores differ in some of those details.

So compare these categorical listings from two grocery vendors in my local area, Peapod and Wegmans. Now some of these labels are really, really similar, like produce, right? Nice and stable, they both have a produce category, but Wegmans has a category for meat and seafood while Peapod has a category for meat and a separate category for seafood. Peapod has a category for dairy while Wegmans has a category for cheese, very specific at a high level. Wegmans also relies on just a few of those categories, I assume the most popular ones, while Peapod works on surfacing more of those aisle groupings up front. Now, if we were doing this conference in person, this would be the part where I ask, who can tell me where to find the peanut butter? Because peanut butter, I swear it's in a different category in every single store in the entire world. There is no consistency in how Americans categorize our beloved ground up peanut spread. Now, at my local shop, I usually find it in the bread aisle, usually, at Peapod, you have to click on condiments and sauces. That's where they put peanut butter. Now at Wegmans, the breadcrumb path looks like this. Simplify your shopping, departments, other departments, grocery, and then peanut butter, jelly, and honey. Why, why is this happening? Why did this happen?

Okay, this is a silly example, but my point here is that categorization is not nearly as straightforward as we think it is. We're tricked into thinking it's simple because we make categorical decisions every single day. We keep our lettuce in the crisper with other vegetables. We put our socks away in the drawer with other socks, but the only reason we're able to do those tasks efficiently for varying definitions of efficiently is because each of us intrinsically knows what we mean when we say vegetables or socks. We have our own private definitions. And we automatically look for the qualities that match those definitions. That's a process known as criteria matching, and we are doing it all the time. We just don't necessarily realize it. It's what we're doing any time we're trying to decide if something is like something else. And that can be really tricky to articulate, which you know if you've ever engaged in the infamous argument of whether or not a hotdog is a sandwich or whether or not a 38-year-old counts as a millennial. Please, someone tell me, I really need to know.

So deciding if something is like something else is a process of drawing boundaries, of deciding what to include and what to exclude, which means it's a process fraught with subjectivity and mistakes and bias and politics and cultural expectations. It is not as straightforward as it seems at first, but it is something we can break down and articulate. To understand criteria matching, we have to answer two questions. The first question is, what are the criteria for the category? What defines the category? How do we know what the category means?

So let's walk through a simple example. I like to cook, which means I spend a lot of time collecting and sorting recipes. Salads have been coming up a lot for me lately, because like I said, it's hot here in Boston, but honestly, what is salad? How do we know what salad is? How do we define it? Again, this is the part of the talk where I would ask you lovely people to volunteer some answers. But since this talk is prerecorded, I'll go ahead and offer some criteria of my own. If we're gonna define salad, we might say it's a mix of vegetables. Okay, it's mixed up, it's mostly lettuce. That's usually what we think of when we think of salad or maybe what I think of when I think of salad. It's cold, right? That's why I like it in hot weather. And the pieces are bite-size right? They're like, chunks of mixed cold vegetables. Great, very appetizing sounding. Okay, so that's our working definition.

Now, now that we know what we mean by salad, we can answer the second question. Does the content match the criteria? And we figure this out by evaluating individual pieces of content to determine whether each fits the given category. So let's try this, butter lettuce salad with lemon Dijon dressing. It's mostly vegetables, mostly lettuce. It's cold, it's in bite sized chunks, great. A-plus this fits our criteria, right? This is, this is a salad, and I don't think too many people would argue with me that this is definitely a salad. Okay, so what about this one? Citrus miso slaw with 10 vegetables, very specific. So this one looks like a salad to me, it is cold. It is bite-size pieces of vegetables, but it doesn't seem to have any lettuce in it. And it's not even labeled salad. It's actually, it's called slaw on this recipe. So would this count? Maybe, maybe, we'd have to think about it some more. Celery pear and hazelnut salad. Okay, celery, pear, and hazelnut. None of those are lettuce. Two of the three things aren't even vegetables. But it is cold and it is bite-sized so, and it is a mix, just not of vegetables. Okay, I mean, I think we could still call all of these salads, I think, I think we could do that. Now, this one, roasted sweet potato, wild rice, and arugula salad. Okay, again, not lettuce. And I'm not even sure it's cold this time around. Roasted sweet potato means we have a hot component to our salad here. Okay, grilled zucchini salad with halloumi, tahini and mint. Definitely not cold, definitely no lettuce. A mix of cheese and zucchini. And it is, it's hot, can we have a hot salad? Can we have a hot salad, one wonders. Finally, prosciutto and melon salad. This is a joke, this is a pile of fruit with a pile of meat on top of it. It is not a salad, it is not a salad. Okay, all right, we might call it a salad.

The point is, when we first defined salad, we established certain criteria, right? We drew lines, we decided what would be included and what would be excluded by the term salad. But as we evaluated those different salads or dishes against the criteria, we realized some of these criteria don't always work. Many of those salads were a mix of vegetables, meats, cheeses, grains, and fruit. Some had no vegetables at all. Only the first one had actual lettuce in it. Several were grilled or roasted, meaning they were not cold. They were definitely hot. And while most of them were arguably bite-sized, those grain salads are gonna have much smaller pieces. So I don't even know if we'd keep that criteria. So where are we, which of those were salads?

Now, what we would do next is come up with stronger, more specific criteria that accommodates those dishes. Or maybe we find a new name for the dishes so we can keep our criteria intact. More likely, some combination of the two. And honestly, after this exercise, I'm not convinced that I can give you a good definition of salad. But what we are seeing here is the process, the journey of criteria matching. It's meant to be a back and forth thing. We make a rule, we test the rule, we tweak the rule. It's good to reevaluate and redraw the boundaries. It means we're able to better articulate why we're making the categorical decisions that we're making, which means we're more likely to be making good decisions. Here's the thing though, salad is easy. It's a low-stakes example that I came up with after spending too much time on Pinterest. And in the real world, we have to categorize much more complicated things.

For example, what would happen if we tried to categorize all the attendees at this conference? We're more complicated than a salad, I think. If we were all together in a conference hall, I could ask you to raise your hands and identify yourselves according to astrological sign. The criteria for that is pretty straightforward so I might ask all the Pisces to raise their hands or all the Geminis and all the Scorpios would sit here being mysterious and brooding. Some of you would be probably really eager to raise your hands because it's fun to talk about astrology, and it's fun to see what other people are doing with that. Or I could categorize this according to the type of place we work. I could ask those of you who work for an agency to raise your hands, or those of you who work at a university, or those of you who are in-house at a corporation, or those of you who work at a nonprofit. And then I could ask who here is from the United States and who here is from Canada and who here is from another country and who here can prove their citizenship?

And, oh, wait, that um, that didn't feel the same at all. That didn't feel cool to ask. That was kind of weird and kind of intense and kind of personal. And I bet I'd get a similar reaction if I asked us to categorize according to marital status or religious practices or health conditions. Unlike sharing astrological signs, most of you would be reluctant to self-identify. You wouldn't want to share those answers with a bunch of strangers or even just with me. You might reasonably wonder why I'm asking for this information and what I'm going to do with it. Suddenly it's not just an easy, breezy exercise in categorization. Suddenly there's vulnerability and discomfort and anxiety because it's never just information, it's our lives.

So let's take another step back, this time to the 18th century. Let's talk about our friend, Carl Linnaeus, the Swedish scientist responsible for the biological classification system. That is, kingdom phylum class order family genus species. It's one of the most important and impactful categorical systems ever created and has helped set the standard for taxonomies ever since. So, as you might imagine, in order to invent all of taxonomy, Linnaeus had to take a lot of notes. Now, scholars at the time took notes in books. Books are great for writing things down, but they have this one tiny problem. The information that you put inside them is fixed in place. You can't move ideas around, you can't swap pages or shuffle concepts. Once you write your idea down on a bunch of bound pages, that's it, that's where that idea lives forever. So books aren't necessarily ideal for managing information that might change over time, like criteria matching. I shouldn't be so mean to Carl, sorry, bud. Linnaeus hated that he couldn't move his notes around. So he started writing on smaller, separate slips of paper that could be rearranged to accommodate new information. Voila, something that eventually became the index card was born.

Now, the power of index cards comes from their ability to juxtapose information and thus create new information. Remember our quote from Richard Saul Wurman. Index cards were better than books at managing research notes, tracking records, standardizing and structuring small bits of data. So it's really no coincidence that soon after Linnaeus European governments started standardizing birth certificates on small slips of paper, they started standardizing marriage certificates. They made paper money for the first time and libraries started using card catalogs. Now, whenever we find a helpful tool in human history, we also find a weapon. Once Europeans recognized how easy it was to categorize plants and animals and government records, they decided to start categorizing people, too. And bad things happened.

In a 2017 article in "The Atlantic" Daniela Blei reported as soon as anthropologists applied Linnaeus's taxonomical system to humans, the category of race together with the ideology of racism was born. Nazi ideologues compiled a deadlier index card database to classify 500,000 Jewish Germans according to racial and genetic background. Other regimes have employed similar methods, relying on the index cards' simplicity and versatility to catalog enemies real and imagined. The act of organizing information, even notes about plants, is never neutral or objective. So if you take away nothing else from our chat today, please remember that. It doesn't matter if we're talking about groceries or salad or personal identity. If we assume that categorizing information is simple, if we act like it can be neutral, if we treat it as a trivial task that's just standing in the way of some other cooler part of building a website, that's where we run into trouble. That's where we risk exclusion. We risk perpetuating unjust structures and we risk harm.

So back to how not to do that, back to our work on the web. When do we encounter opportunities for categorization online? Where do we see categories popping up in our digital experiences? Frankly, the answer is everywhere. It's everywhere, so here's a few quick examples. We see it a lot in media platforms, right? Anytime we're collecting movies, television shows, podcasts, music we're going to see categories cropping up. Here, Netflix uses categories to express genres, right? Based on their own sense of things or based on my viewing habits. We also see it a lot in shopping, e-commerce, retail. This is Zappos, they use several levels of increasingly granular categorization. You say that word often enough, I swear you'll trip on it. Now we start with this top level, a division between women, men, kids. And then we go into product categories like shoes and then deeper into shoe types like heels, sandals, slippers, sneakers, they even use categories to create promotions for brands and clothing styles, right?

We also see categories come up a lot in journalism and publishing. This is "The Boston Globe," obviously. And like most of their news sites, they use categories to not only divide their content into sections, just like, you know, in a print paper if we remember those, does anyone remember those? But also they use categories to highlight trending topics. And if you look closely here, you'll notice that Dunkin' is on the trending topic list, but COVID isn't, which just goes to show you that people care way more about mediocre donuts than they do about wearing masks. So, no, it's cool, it's fine.

We also see it a lot on social and creative platforms, right? Categorization shows up like here on Pinterest, like we saw in our criteria matching example earlier, systems that let you organize and personalize collections of items are all about categories and labels. This is my dinner Pinterest board dinner basically where I keep savory recipes. And I've mostly organized it into sub boards by primary vegetable, right? Like artichokes or root vegetables or salads, which is too many vegetables at once. But I also could have categorized these recipes according to season or according to preparation method or according to cuisine and flavor profile. So it's a creative space that lets me establish my own criteria and categories. And you can see similar behaviors in Instagram save folders, Twitter lists, Spotify playlists, that sort of thing.

So categories are everywhere, which means you are always working with them in pretty much any digital experience. Maybe you are listing products or grouping materials. Maybe you're developing a site map and trying to section out the content. Maybe you're building a controlled vocabulary or a taxonomic system. You're likely practicing categorization in multiple contexts, even within the same product or site experience and this certainly isn't a complete list. This is just a few of the more immediate contexts that come to mind. Now, no matter which contexts your categories are appearing in, they have two very important roles to play.

The first is that they communicate all sorts of clues for how users should think about your content. The way you choose to organize and label your content can tell users about the scope of that content, the relationship between pieces of content, the hierarchy of information on the page and the hierarchy of information across the whole site. Your categories tell users whether they're going to find what they need on your site or not. Done right, they should tell them exactly where to find the peanut butter, right? Let's look at some examples.

Here, this is the website for the MBTA, Boston's transit system. The categories here tell us that there are, first, multiple types of transit available in this system. We've got trains, we've got ferries. We can also see that there are lots of tools for self-service, right? We've got maps, we've got a list of subway stations, a trip planner, and then there are categories for support, alerts, big service change banner at the top. The prominence of that alerts category actually is pretty telling about how often alerts have to happen in the system. And this is where I would normally make a joke about the T always being on fire because it is, but instead I'm just going to sigh deeply and wonder if I'll ever go outside again. Anyway, quick categorical bonus. If you scroll all the way down on this homepage, you actually end up at a series of buttons that look like this. Ah, important links. All of the other links must have been unimportant. I joke, but a category label like this actually demonstrates indecision or maybe internal politics gumming up the works. It's not the intentional categorization that we want to work towards, right?

Right, let's look at "The Boston Globe" again. All those categories up there at the top. Let me blow those up for us so we can get a better look. That's better, wait, wait, oh, wait, okay, all right. I think we're done now, great. Okay, the thing about having just so many categories, especially at the top level of navigation on a news site is that it has elevated all of these categories to the same level of importance and the same context. So what happens in this list is that Rhode Island becomes conceptually equivalent with sports and business yet separate from metro for some reason. Marijuana is conceptually equivalent with arts, and magazine might as well be a topic like cars rather than an alternate channel. As users, we're being told that these things have parity even though we know they don't logically match and there's too many of them to take in the whole picture anyway. On the other hand, sometimes overwhelm is what we're looking for, right

This is a series of screenshots that I've put together from a website called Ravelry, which is an online community for knitters and crocheters. And what we see here is the advanced search options. And just part of them, it keeps going, but I couldn't fit everything on the screen. So what we learn about this content just from seeing this system of categorization is that it's extensive, right? It is just massive and detail oriented. It is in depth and specific. These advanced search options are just wildly detailed, which is what people want on this website, actually. What we can see from this example is that we're not just getting clues about the content itself, but we're also starting to get clues about the users, right? Like for instance, they like detail. They like to be able to get very specific. And we can also tell that they're primarily thin, 30-something, white women who like being cozy, which brings us to my next point.

Categories also tell your users what you think about them. The language that you use, the hierarchies that you establish, the scope of your categories, these are all bits of data that reveal who you think is using or should be using your website. So this example, I love this example, spotted by Erin White on Twitter. What does Cannondale bike company think about their users just from this system of categorization? Well, first of all, they sell road bikes, cyclocross bikes, mountain bikes, women's bikes, which means that all the other bikes must be men's bikes. Or maybe what we mean is bikes for bad boys. Bikes for hooligans, I don't like this. I don't like this. What this says is that women don't ride bikes. Women don't use this website. And well, we certainly don't now.

Another example, this is Harney & Sons. This is a tea company that I love. I love their tea, and what we learned from this site, and I know it's hard to see, but this is the page for black teas, all the different varieties of black teas, categories of black teas. Let me blow one up here. This is teas from other regions, that's the category. And it says nowadays tea is coming from more than the usual suspects. Please explore these teas from Africa, Indonesia, Southern India and Vietnam. Talk about drawing boundaries. What I take from this aside from some weird assumptions about usual suspects is that they expect their users have expertise on black tea. They expect that their users, first of all, know what the usual suspects are. They assume that their users know where the tea they're looking for comes from, whether they come from one of these countries or from a different region. They think their users know exactly where to find the tea they want. And I love their tea, and I love black tea. I love tea in general. I'm not an expert, I don't know where my teas come from. And maybe that's something I should learn about. But in the meantime, I don't know where to go to find the tea I'm looking for. I don't know if it's in this one or in a different category. So that's an assumption they've made about their users.

Now, this is from a profile form that one can fill out if they are a member of the Association of Writers & Writing Programs. What this section is for, this is the third question they ask in your profile. It says gender, it says, this section is optional. This demographic information will never be displayed in your online profile or in the online conference schedule. Gender information is only used in the aggregate to help AWP report to our funders and membership, and it may be released when a representative sample is achieved. This information is not used as part of the proposal process and is not shared with the subcommittee. And then it offers us, I think, 13 options here for gender, as well as a gender not identified here option. So what we learn from this is that they believe that their users have a wide range of identities, care about expressing those identities in the context of this association, and they also understand the potential for harm that comes with some of these identities. By displaying, the information of this section is optional and this demographic information will never be displayed, they're saying, we understand our users. We understand that transparency is important. We understand that optionality is important. We understand that not everyone wants to give this information. Those who do can, and they can give the information that is most accurate to their own identity. So this is a great example of categorization demonstrating what a company knows about their users and what they think about their users.

So examples are one thing, but how do we actually do this work? What are we doing when we sit down to create categories for our products or services or search systems or site maps? What is it that's guiding our thought process? There are four factors that influence our categorization.

First, there's the content itself, your primary source of information. It tells you what's missing, what's unnecessary, what's working well, what's not. You learn this information through a content audit and other discovery activities. And that forms the foundation of your organizational decision making. You have to start with what you have. The content will lend itself to some initial natural groupings.

From that starting point, the business goals start to add a little color. What is it that your company or your organization is trying to accomplish? Why is this product or experience or website being built? Or I guess another way of saying this is, how is the money getting made? Because unfortunately we are all still working in a capitalist system. So that bottom line is actually going to influence your categories and labels a little bit. What systems of categorization will enhance the business? And are those systems in conflict with our next criteria, the needs of the users?

This is the one I think most of us are trying to focus on most of the time. This is where we really spend our energy. We think about what people need, what mental models do they use, what language and labels will be familiar to them? How can our categories give them what they want quickly and efficiently and effectively? User needs and business goals are this sort of twin set of factors that get a lot of attention.

But lately we're starting to hear a little bit more about this final factor. And that is, how can our choices benefit the community? What's the impact on the larger world? Because design decisions can certainly meet user needs and business goals while still having a deeply negative impact on society, Facebook. But if we can incorporate that question of greater impact into our categorization decisions, into all of our decisions, we can start to build better systems.

In the real world of technology, Ursula Franklin calls for redemptive technologies, technologies that are driven by their impact on people, on society, on the natural environment. How do we build technologies that promote justice, that right wrongs, that uplift the human condition? How do we use our work to protect community and climate? Erika Hall argues for a similar approach in her article, "Thinking in Triplicate," which borrows the concept of the triple bottom line from economic thinking. The triple bottom line suggests that a company's profit and loss isn't an accurate measurement of the company's value. To truly capture value, we also have to consider the social impact and the environmental impact. Only by lining up all three of those things, profit, people and planet can we gauge the true bottom line. Shockingly, this idea has not really caught on with most companies, but it certainly is an idea we can try to incorporate into our own work on the web, or as Erika put it, the concept I'd like us to all agree on is that we need to design products and services that make their users better off, make money, and don't fuck up society or the planet. This is actually a tall order, but I think we can start small. We can start with the categorization of information, a very tiny thing that you can influence with your work. How can your categorical systems advance the greater good?

So I want to look at a few more examples that start to head in that direction. The first is the homepage for a nonprofit called Legal Services of Northern California. The first thing we notice on this page, the first thing our eye is drawn to are these brightly colored buttons in the middle of the page, right? They each represent a content category, a topic of legal assistance that they can provide their clients. Now, they could have made a list of these categories as a dropdown from the main menu, which would have hidden them a bit more. They could have put them in a text sidebar, which wouldn't have been as attention-getting. They could've put these items in a carousel, which, let's not start talking about carousels. All of those decisions would still have met user needs and business goals. Any other organization would have seen this primo real estate on the homepage and they would have said, great, let's display our news. Let's put up glossy hero photos, testimonials, a letter from the director, but Legal Services of Northern California chose to use this space to display categories to shorten the distance between landing on the homepage and actually getting help, bravo.

Another example, Patagonia. This is a lot of white space, right? Because I've zoomed way in here. This is, what you're seeing is the drop down menu that appears when you click on the label inside Patagonia. What I love about this design in particular is that their main navigation is only two categories, shop and inside Patagonia, that's it, those two at the top of the page. You click on this, this is what you get. What that categorical decision says to me is that the information about their mission, their activism, their environmental and social responsibility is exactly as important to them as their sales. Other retails display that top level product categories as their main navigation, right? Like we saw in Zappos, women, men, kids, or they might bury their mission information in the footer or put it in a utility navigation. But by Patagonia condensing their entire catalog into this one category of shop and putting it on equal footing with their mission information, that's sending a very different message. And one that's much more in line with that triple bottom line approach.

I also really love the example set by clothing company, Universal Standard. This is just an item sold on their site, but look at the categories they've set up for the sizing. I'm gonna make that bigger. They're not the same clothing labels found on other retailers. First, the range of sizes is much more expansive. It runs from size 00 to size 40. Typical women's clothing retailers, at least in America, usually offer size two to maybe 16, 14. Second, they've recontextualized those sizes by labeling the middle size, 18 to 20, that's medium. That's not true in other retailers if you're not familiar with women's clothing sizes. That's not how it works in other places. They've started with the middle size that they offer, labeled that medium and then adjusted the sizes out from that center point, which is a much more progressive and inclusive way of looking at sizing and shopping. Now their website says, if 67% of women in the U.S. wear a size 14 or above, why were their options so dismal? It was clear that all women weren't given the same level of quality, style, or even respect. We wanted a size 40 to shop in the same way as a size 00 using style as her only filter, and we wanted to be the ones to make that happen. So this very simple change in a label, in a category, by doing this to their sizing, they've advanced the conversation of inclusion and body diversity. Using the same sizing as every other retailer would have still met business goals and user needs, but it wouldn't have had the same impact.

Now, last fall, the racial justice organization, Color of Change, launched a campaign, asking wedding planning websites to stop promoting former slave plantations as wedding venues. As they put it, you have a multimillion dollar industry that makes money off of glorifying sites of human rights atrocities. Now one site, Brides, responded by immediately removing plantations from their feature articles and vendor list. You can see here searching for the term plantation wedding doesn't show those venues, nothing comes up for that. They said in a statement, Brides is an inclusive place where everyone can feel celebrated. Content glorifying plantations is not in line with our core values. In order for Brides to be successful in this effort, two things had to happen. First, they had to be able to recognize that plantation or plantation wedding was a category, something they could define, something they could criteria match. Second, they had to want to act on that definition. They had to make a decision, an active choice to change the way they and their users interact with that category. They took a stance. They recognized this category is not neutral.

Now, contrast that with The Knot's response to Color of Change's request. They've decided not to remove plantations from their site, but they have promised to revisit the language that they use to promote those venues. Language, I guess, like paradise, breathtaking, sophisticated. Clearly they haven't revisited the language yet. How about the phrase, the landscape boasts a rich history and culture. Hm, what history do we think they're referring to? The Knot has argued, even if we remove these venues from our marketplace, they of course still exist and will continue to operate and potentially still market themselves using language that is insensitive to the history of plantations, which doesn't solve the problem. But the thing is, Color of Change didn't ask them to solve the problem of plantations existing. Color of Change asked them to take one small step in the right direction. One step toward reducing harm to users.

One change in categorization. We can make those small changes because our categorical choices have consequences. They impact our users and our communities in both good ways and bad. Every decision we make about how to present information, how to structure it, how to design it, how to categorize it will ultimately change it. So let's make sure that we're changing it for the better.

If you like what you've heard here today, you can learn more in my book, "Everyday Information Architecture" from A Book Apart. Thank you again so much for being here and continuing to fight for justice through your work on the web. Please stay safe, thank you.