You might not think of Home Depot as a leader in human centered design, but 1.5 Billion transactions a year has taught them a few things. Today we'll talk with Kelly Robinson about how they're using UX and content strategy to keep customers happy and coming back for more.

Kelly is a UX content strategist with The Home Depot. She’s spearheaded content and design initiatives across multiple industries, from retailers like eBay to entertainment companies like Blue Man Group. Her specialties include information architecture, UX writing, user research, and front-end development. Kelly is an animal welfare advocate who finds delight in songwriting, camping, and spooky stories.

Follow Kelly on Twitter


Episode Transcript

J. Cornelius: You might not think of Home Depot as a leader in human-centric design, but 1.5 billion transactions a year has taught them a few things. Today, we'll talk with Kelly Robinson about how they're using UX and content strategy to keep customers happy and coming back for more.

J: Hey, everybody. We are happy to have Kelly Robinson from Home Depot with us today. She is a UX content strategist. She's spearheaded content design initiatives across multiple industries from retailers like Ebay to entertainment groups like the Blue Man Group. Her specialties include information architecture, UX writing, user research, and front-end development. She's also an animal welfare advocate who finds delight in songwriting, camping, and spooky stories. Kelly, welcome to the show.

Kelly Robinson: Thank you. Really glad to be chatting today.

J: Yeah, I'm glad you could join us. What's going on at Home Depot? What are you working on? What are you excited about?

Kelly: Some really exciting stuff going on right now. A couple of the projects I'd like to talk about include chat bots. That's reaching a couple different types of audiences. Let's say you're in the store, you're wanting to find something. You're having a hard time. You can just pull up a chat bot on your phone and it's going to direct you to the specific aisle and bay you're seeking your product in. That's coming soon.

J: Cool.

Kelly: From a content strategy perspective, I've been working on some flows for that. What sort of questions are people going to ask and then how do we best address those questions? The idea is really to be conversational, make it feel like maybe you're talking to a real person, but make it obvious that it's a bot at the same time. We obviously don't want to be misleading or cause any types of confusion.

There's also going to be a product-focused chat bot. Right now I'm working on a flow with washers and dryers. For instance, what sort of questions do people have when they're shopping for a washer and dryer on Home Depot? What are their delivery questions? What do they not understand about maybe the specs? And then, surfacing answers to those questions in a way where they can just quickly plug in a question, they would get an immediate response, and then they don't have to necessarily talk to a real person, but they would have the opportunity to if they wanted to at any time during the process. Leaving that line of communication open, but just giving them some more options. I'm really excited about those two things that are in the pipeline. From a content strategy perspective it's really exciting.

J: Yeah, I bet. It seems like riding that line or finding that line between knowing that it's a bot, but still sounding like a friendly human is probably a really tricky thing to do. How are you doing that?

Kelly: It starts with the introduction, maybe we say "I'm the Homer bot", for example, or the store finder might have its own particular name, and then "Here are some ways that I can help you today." Just instead of having these types of robotic, repetitive responses, like "I could not find that item," "I could not find that item," we really try to randomize it and have "I couldn't find that, but maybe my friend in an orange apron can help you today." Or "I'm sorry I couldn't find what you're looking for. Maybe you could try putting in a different type of product."

Definitely randomizing it a little bit. Yeah, just keeping it friendly, keeping a conversational tone. It applies to chat bots. It applies to a lot of what we're doing, really. A big initiative in content strategy with Home Depot over the last year and beyond has really been making the content feel more authentic, personal, conversational. We just want to add warmth to what we're doing from things like instructions to tool tips to error messages even. I've probably rewritten about 200 error messages since I started here. That's a big one.

J: Wow, yeah. Microcopy, super important. I guess that's one of the easy places to start is what are the basic, simple interactions that you have and how do you make those feel more friendly or more human?

Kelly: Let's see. A lot of times it's ... Well, I'll take an error message for example. We want to really help somebody identify what the problem is, explain to them in a human way how they can go about solving the problem, and just expressing sympathy. Sometimes, just "Oops" might sound human, but it's not really the way that you want to talk to somebody when maybe their order wasn't processed and they just spent 30 minutes putting it together. Something like that.

J: Oops sounds a little too casual for that, yeah.

Kelly: Context is important, right. If you can't add something to a list, maybe that's more of an "Oops" scenario, but if something really went wrong, it's going to frustrate somebody. We have to be a lot more empathetic with our tone. We obviously have these space constraints, too. There are scenarios where we maybe can't make an alert or an error message really long, so we have to make it really scannable, easy to absorb quickly, but also keep that empathy in mind with the message. That's something that can be challenging, but it's fun on my end to figure out that right balance.

J: Yeah, I bet. When you and I met some time ago, we were chatting about the composition of the typical Home Depot audience and how some of it are professional contractors and people who are in the store maybe on a daily basis, and then there's the weekend warriors who are just trying to maybe fix up their patio or shop for a washer and dryer. How do those different audiences play into the types of messaging that you use? Is that a consideration or what are your thinking around that?

Kelly: It is. We really try to personalize a lot of the content. Definitely, professionals are huge, so in terms of, let's see, store sales, it's a huge percentage. Not as much online, but what we're really trying to do is have a larger outreach in terms of the types of features that we're presenting on the website, the types of content that we're creating so it really adapts to different types of audiences. For example, with pros right now, we know that they come to potentially wanting to seek out different stores in their area. Somebody in Atlanta, for example, may have four or five different stores that they shop at. We want to be able to service inventory to them quickly, so we're working on ways to do that. We also know that they tend to reorder maybe more frequently than somebody who's just shopping for a single appliance, like the weekend warrior you were talking about. Maybe having a way of surfacing the reorder options to them a lot more easily.

And when it comes to communicating to different types of audiences, we know that some people who are accessing, let's say, a product landing page, product information page, they know exactly what they want. They don't necessarily have any questions about it. The type of content that we present to them may be a little different from somebody we have identified either as a pro or somebody who may be a lot more knowledgeable in a specific area. I guess a common example would be plumbing. Let's say we have enough information about somebody to identify that they are a plumber. Either they have self-identified or just from their purchase history, it's obvious this person knows a lot about plumbing. Well, we're probably not going to do a lot of product identification, like, "Hey, here's some getting started with plumbing types of guides" for them. That's not really going to be useful to them. But somebody who's maybe tackling a project for the very first time, we would maybe want to find ways to communicate to them, "Hey we're here for you. We want to help you figure out the best way to tackle this project on your own. We want to give you confidence and empower you." So that tone is going to change a lot, in addition to the types of content that we might be showing someone.

J: Yeah, sure. You mentioned testing and doing a lot of research in figuring out how you're going to message people and what type of content to display at what point. Can you talk a little bit about that process and what the results have been in terms of maybe higher retention rates or more sales or what kind of results are you looking for?

Kelly: I guess I'll maybe just talk generally about our research process and testing process and how it takes form in the context of our user experience team and then maybe I can go into some specific examples.

J: Yeah, that'd be great.

Kelly: Okay, great. One thing that I'm really excited about, recently we have formed a usability team within the user experience team. Hired a couple usability analysts over the past couple months. They're doing a lot more in-depth research and analysis to really understand our customers. A traditional type of persona might be developed from some of the research they're doing. They do interviews. Some are remote interviews, some are in-person, some are moderated. Types of tests. Let's see, they do shopalongs, which are really cool. That's where you would actually go to someone's home and observe the way that they would interact with different types of retailers, whether they're looking to do a home improvement project or something else. They've been doing a lot of cool stuff.

J: Are you watching them shop other competitive sites or competitive retailers or product or companies, or what are you watching them shop?

Kelly: Yes, yes. It could be. It could be other retailers. It could be Home Depot. Just to understand how they want to interact with certain types of features. Maybe how they would go about adding things to their cart, how they would go about adding lists, storing information. Just some really interesting behavior in terms of talking to people and actually watching them. For example-

J: So you're watching them not just on your own site and your own product, but on not just competitors, but wherever they might be shopping and seeing what their behavior patterns are.

Kelly: Right. Sometimes we're surprised by some of the patterns that we see. They don't necessarily always go along with our expectations. That can be really educational. And then what I do more on my end as a content strategist, when I'm trying to research in preparation for a test or a preparation for a certain type of initiative, is maybe talk to some call center associates to figure out what their pain points are, what are people calling them about, looking through some of the most frequently asked questions through the chat dialogs, doing some stakeholder research, talking to different business units within Home Depot, doing some competitive analysis, looking at what's been tested before, looking at online data through things like Omniture, 4C, Clicktale. 4C is, just for context, what we use to get feedback from our customers. There's a wealth of content through there.

Omniture is what we use for a lot of our web analytics. Just an example of the way that I recently used Omniture, I wanted to figure out what sorts of questions are people asking their phone in order to learn about or find out about certain types of products. So many people are asking, "What is the name of this thing? I have this chandelier and there's this thing on it. I'm not sure what it's called." "What is that thing that I put on my door?" We're taking these types of interactions and thinking how can we shape an experience based on that? From my point of view, that's a lot of the types of research that I might do. And then the usability team would be more in-depth.

When it comes to testing, they're also involved in helping us really set up tests. The UX team, in terms of designers and content strategists, here in-house, we do a lot of unmoderated tests through, a couple other platforms we're experimenting with, but how that tends to take shape is for something small, like I want to maybe test out some of the new fulfillment language we're using, when I say fulfillment, things like pick up in store, have it shipped to you, all these different options, I've tested a lot of that because I want to make sure that we're communicating it in a way that people understand. Just putting together a prototype, maybe having people compare different types of messaging, compare different types of interactions with the content to get a sense of is it clear? What do we need to change? Are we maybe assuming too much of people? Not assuming enough? Those types of things. And then reshaping the content, the design, based on what we see, based on how people are struggling.

J: Yeah. Are you testing all the way down to individual words and phrases on pages or on different screens in the mobile app?

Kelly: We are. We are. Sometimes it's a lot more comprehensive. It's more testing. Are people able to understand and interact with a certain type of design? Is it intuitive enough? Other times, what I really love to do is get a sense of maybe comparing different types of phrases for how we would describe something. An example would be we have this experiment we're doing, we're trying to think of a way to offer a same-day delivery to certain types of markets, and so how do we go about communicating that without potentially misrepresenting the process or over-delivering, or saying something that wouldn't necessarily be accurate, or lead to misperceptions.

That's a big part of it, just testing out a few different combinations of even how we label these types of things to ask people "How do you perceive this? What does this mean to you?" And then seeing how they're able to interact with it. Does it meet their expectations? Once they get to the cart, do they see what they thought they would see or is it something completely different? If we are describing a process in terms of a few bullet points, is there anything in that set of bullet points that makes people go, "Whoa. I didn't expect that." Or "That doesn't really seem like how I would want the process to work out." And then sometimes we'll adjust things based on those perceptions.

I think we have to be careful in how we interpret the results, too. A big part of it is that we want to make sure that, based on, let's say the first two or three people, they may struggle with something, it doesn't necessarily mean that that group is representative of everyone who's going to be interacting with the experience. We try to use the unmoderated testing results as guidance on how we can help improve something, to help maybe identify a struggle, identify where something is unclear, but we also don't want to make these grand assumptions, like well, because we observed this in two people, that means that people think this way or people approach shopping this way because we could be totally off if we were to do a larger survey.

J: Because two people is a pretty small sample, right?

Kelly: Right. A lot of times, we may do a test and we put it in front of, I would say, on average somewhere between eight and 15 people for a particular type of test, if we're testing, maybe, two things against each other, it would be maybe six to eight people in one group, six to eight people in the other group.

J: Sure.

Kelly: That tends to be a pretty average sample size for the types of tests that I might run.

J: And all of those people, I'm assuming, would have similar personas, right? You're not testing somebody who's a professional electrician against somebody who can't figure out what that thing on their chandelier is called.

Kelly: It could be a pretty random sample or we could actually set up the testing in a way where there are screening questions to get the type of, let's say, group that we may be seeking, so, for example, maybe we want to talk to people who have just started a kitchen remodel because we want to get a sense of if they go through the remodeling process, does the information feel accessible to them? Are they finding what they're seeking?

J: Right.

Kelly: We're actually able to ask the type of screening questions, what sorts of home improvement projects do you plan on tackling in the next three months, would be a common one. Or what sort of remodeling projects might you currently be undergoing? If we're looking to basically test just homeowners, then that could be a screening question, too. If we want to really expand our audience, then maybe we wouldn't have as many specific screening questions, but we do try to get to some pros through that way, also, just in terms of trying to narrow down the test takers. If there's a really big pool of people out there who are wanting to take these types of tests, then we can get pretty specific.

J: Thinking through how a design's going to lay out, once you get some testing results back, how solid is the content? Are you still dealing with working content or placeholder content at that point? Obviously, you're not doing something like lorem ipsum all over the app or the website, but what are you doing in terms of developing content and using placeholders to get you through the process so you can make sure you're on the right track?

Kelly: That's a great question. We typically start off, when I'm thinking about a design and how the content is going to fit into the larger story with a designer, I'm definitely not thinking about placeholder content. We're thinking about more functional content. The idea is when we're putting it into a prototype to begin with, we want to experiment with it first, just to see does it fit into the space the way that we expected, for example. Maybe a tool tip isn't the best use of this space. That kind of a thing. That's why using working content is really important then maybe content goes here, lorem ipsum or paragraph goes here. We get a sense of we just have a rough draft to begin with.

Okay, so what information do we want to convey? We know that, from the business stakeholder's point of view, we have to really communicate these three points, and then we know that the users are really caring about these three points. We have, let's say, six pieces of information that we want to convey. Let's actually figure out how to put those pieces of information together in the prototype and have a somewhat polished, but not perfect, way of communicating it.

First it would take on a very draft-like form. We would work with it to maybe figure out how to clarify the message more, figure out how to cut down unnecessary things, figure out how to make it more scannable from that point, and then when we feel pretty good about it, that's a point where we would start maybe putting it in front of some people to get their take on it. Yeah, there's that whole sense of using working content from the very beginning. I've heard the term proto-content also used before. I like that, too. Just saying, it doesn't have to be polished, but it has to be functional, relevant, it has to be something that would be somewhat close to what you would actually be using in reality. Once we put that-

J: Yeah, so you're kind of testing content length and, as you were saying, scannability, how well it fits into the overall aesthetic of that particular screen and the message you're trying to communicate, right?

Kelly: Exactly. Once we put it in front of people, we realized how we might want to refine some of that messaging, then we just go from there. In some cases, do some followup testing if it's a particularly complicated type of project.

J: Right. When you're thinking about the content and the role that it plays in the overall experience that you have, how are you looking at the success of a particular piece of content or the voice and tone as a whole? How are you measuring that and really, how do you think about success beyond a conversion or a successful checkout?

Kelly: Yeah, that's been something that we're really looking into. It's valuable for us to look beyond conversion. Conversion is so important. It's a very traditional and measurable and tangible way of measuring success. It can sound really great if you're like we released this feature and conversion increased by this many basis points. It sounds exciting, but in reality, we really want to think about measuring our success beyond that. We actually had a great speaker from Clicktale come in and talk to us last week, and something that really resonated with me that she said was, "Conversion does not equal customer experiences." She mentioned that somebody could have a transaction, but they could actually have a really negative experience that led up to that transaction.

We're thinking about it a lot more holistically and some ways that we may be able to get a sense of have the designs that we put out there actually led to a better experience, a change in perception, all these different things? It could be we see a decrease in returns, we see a decrease in cancellations. Maybe when it comes to the call volume at the call center or the chat volume, there's a decrease in certain areas. We can measure the types of interactions the people have with different areas of the site, with the app. For example, we released a one page checkout recently and continuing to release that in various stages, but just being able to see the way that people are interacting with that, that their checkout time has decreased so significantly. We consider that a success regardless of whether it's actually increasing conversion. It has increased conversion, but just looking at it a lot beyond that.

I mentioned 4C earlier also as just a way to get what sort of feedback is coming through. Are people complaining about the same things that they were complaining about before? A lot of those types of behavioral things. A lot of the work that I do, it actually ties into brand perception. I was talking about things like warmth, for example, making content have more authenticity and warmth, I think those things get especially hard to measure, but I think when we're looking in the grander scheme of things and how it all fits together, even when we look at some stuff that the marketing teams are looking into, they have sentiment types of data that they collect over time. We can probably get some clues from them when they're looking at these types of sentiment streams and how people are perceiving our brand, that it can come from different types of experiences that sentiment could come from a store experience or it could come from what we're doing in online UX.

J: So, obviously, you're having better content can have a lot of positive repercussions. Is the decision to maybe work on a particular group of content or particular screens or how you're doing some messaging, where is that typically driven by? Is that somebody from customer service coming to you and saying, "Hey we get a lot of complaints in this area, can you do something to make it better?" Or is it by revenue management who are looking to increase sales or increase per cart averages or is it all of those things? What's the motivator?

Kelly: It's pretty complex. It comes from several different sources. A common type of process might be we have a product manager that might be focused on a particular work stream, for example, and when I say work stream, it would be like commerce, so cart and checkout on would be an example of a work stream. The one that I'm most focused on is, let's see, browse and online shopping. It would be like search, product information pages, product detail pages. Those dedicated product managers would either be doing competitive research, they would be looking to see what the customer's pain points are. They would be maybe talking to different types of business stakeholders to understand their needs. Let's see, just to get a sense of what they would want to put on the product roadmap. Then, once something is on the roadmap, then it gets translated to us. That is the most common path, I would say, to how a project materializes. Of course there are all sorts of different ways.

Sometimes, UX will drive a particular type of project. We'll identify a problem, think about how we can solve it, and then go from there. I think, for my end, yeah, I might see how people are struggling just based on what I see through Omniture or 4C or from talking to call center associates. An example would be we were looking into really making some improvements on the Help section in terms of making content easier to find, also just getting rid of duplicate content, improving some of the design there, and just actually going to the call center and listening to some of the calls. They kind of have this tag along thing you can do where you put on headphones and understand how people are struggling, and that actually translated into some user stories eventually.

The research that the usability team is doing really ties into that, too. There's this big appliance process overhaul that we're looking into right now. A lot of that was one of our usability analysts went out and just talked to a lot of people who had purchased appliances either on Home Depot or beyond, and really understood where some of the struggles were coming from, and then we were able to translate those into stories. That was really cool. We were looking at things from more of an end-to-end perspective then rather than a work stream perspective. I think that's one of the areas where we're trying to really strike the right type of balance, too, is that we have these work streams that are a little more focused, more granular, and I think that has its importance. It certainly has its place. Then we have to think about things from an end-to-end point of view, too. It's kind of like, think about how your heart and lungs and digestive system and brain all function together and how it's important to maybe focus on one sometimes, but you can't ignore how they're all interconnected and they affect each other in so many different ways.

J: Right because the entire system has to work and if one piece breaks down, obviously that's a problem, but if the entire system breaks, that's a much, much bigger problem.

Kelly: Exactly.

J: Speaking about how the teams work together, talk about your process a little bit. Talk about your team structure and what things you're doing there, and how maybe experimenting with some of that stuff is yielding some good results.

Kelly: Yeah, I'm lucky to be on this team that's really experimental in terms of how we approach different types of things even in terms of how the team itself is structured. Let's see, when I joined, the work streams were a little different than they are now. They've tried not just shuffling the people around, but how we really categorize different types of the work streams that we set up. In terms of things like how we handle reviews, so, for instance, we used to have these Friday alignment meetings-

J: You mean like product reviews? Or team reviews?

Kelly: How we're reviewing each other's work, yeah.

J: Right, so team reviews, yeah.

Kelly: Team reviews, right. We used to have these long Friday afternoon meetings. As the team grew, it became a little bit challenging because not everyone is able to showcase their stuff-

J: For context, how big is the team?

Kelly: Let's see, I want to say we have around 18 designers right now, two content strategists, and then five people on the usability teams. It's really grown significantly.

J: Yeah, that's a good-sized team.

Kelly: Yeah, and then a few managers also.

J: Right, of course.

Kelly: Yes.

J: You gotta have the managers.

Kelly: Gotta have them, absolutely. I think as the team grows, you can't have this one size fits all approach anymore to how we structure things like our review sessions. If it gets too big, then not enough people are really getting their voice heard, not enough people are able to showcase what they're doing. We've really tried to restructure how we collaborate. We use Slack a lot and so we had one Slack channel that was set up for just reviews, just showcasing your work, getting feedback on it. We still use that a little bit, but what we're experimenting now is more of a structure of on a weekly basis, we have a team that showcases their work to a collection of representatives that come from other teams.

For example, today I was looking at what the digital décor team was doing. They have some really innovative stuff going on. It's a way to give your feedback, but also really get an eye on what the other teams are working on. I think, as we have continued to grow, it's easy and it's just the nature of a large company, of a large operation, to start to feel a little bit siloed, like you don't really have the type of insight into each other's work that you would prefer to have. I've been guilty of that, too, maybe just in not sharing things with everyone who I could have shared them with, just because it didn't occur to me, in a lot of cases.

I think setting up that process where things are going to naturally occur to you more, and getting more in the swing of having spotlights on different types of projects, being able to really have the types of information sharing sessions, where we can exchange ideas, where we can really foster each other's creativity, that's really important. We've experimented with different types of structures for doing that sort of thing, in addition to the way that the teams themselves are structured.

J: It's really just kind of building collaboration into the culture and building that into your regular workflow.

Kelly: Exactly, yes. Yes. And it goes beyond even UX, so just to give you an example, there's the enterprise UX team at the Home Depot. We work on a lot of store types of operations, maybe like merchandising software, for example, and then there's the online UX team, we're actually in different buildings, so we don't interact face-to-face that much, but that doesn't mean that we can't have the wealth of resources to share with each other. We really try to collaborate with them through Slack, for example, or through even just grabbing some coffee with somebody and being like, "Hey what are you working on?"

J: Right.

Kelly: I collaborate with marketing on some of their projects even though they're in a different building. Sometimes it wouldn't even occur to me on a daily basis to get marketing input on something, or maybe align with them on the types of style guides that they're using, but then just definitely trying to make an effort to make that more of my process when I'm analyzing content, when I'm thinking about it from the larger picture.

J: Yeah. Speaking of style guides, how standardized is a lot of the stuff that you're working with? Obviously, Home Depot's got a very strong brand standard and that color orange is everywhere, but you also have to be careful that you're not oversaturating things with too much brand. You still let the content come through. I'm curious how you work with maybe the design teams, the marketing team, the UX teams, to find that right balance to where things feel right when you're using the app or when you're using the website.

Kelly: Yeah, in terms of standards, yeah, there's been a lot of work on standards the past couple years. There were some guidelines when I came in, in terms of content standards initially, that were inherited from different areas, some from branding, some from more of the editorial side, and the other content strategist and I, her name also happens to be Kelly, so we're Team Kelly, we wanted to-

J: Nice.

Kelly: We really wanted to take those, but really interpret, okay, what does this mean for UX? And so, we actually put together what was forming to be called the orange library. It's a combination really of content standards, but also the design standards that other people have worked on, are factored into the orange library. It's basically a place where, ideally, there is an intersection of content standards, design standards.

A lot of what we were doing was taking some of these editorial guidelines, but also really trying to define, okay, how does the voice and tone that we want to use fit into this larger context? An example would be we came up with these mini guides, it would be like best practices for writing error messages, tool tips, transactional emails, modals, that type of thing, how that all fits into the way that the design team is coming up with the designs for those. There are different types of templatized things, different types of standards that the design team can grab from, and their documentation. For example, buttons, modals, forms, let's see, what does this mean for full-width browse, that type of thing. How it all fits together in terms of how we're presenting everything out visually, but then also how we communicate it. We don't want to necessarily see those things as absolutely distinct entities. We want them to feel like they're coming together in a set of standards that everyone can access.

J: Right. You have visual standards and you've got interaction pattern standards and you've got your voice and tone guidelines and all of those things kind of fit together to create this overarching standard of how things are going to be communicated and displayed regardless of the properties, small screen, large screen, or whatever.

Kelly: Exactly, yes.

J: I think we've all, at least here in Atlanta, we've all been familiar with the work that's been done over at MailChimp with their voice and tone standards. I think Kate's done a great job over there. Do you have something similar that's ... Is your stuff available to the public or how much of that is something that you can talk about or share with the outside world in order to kind of get an idea of how a company at the scale of Home Depot is actually solving its problems?

Kelly: Right, so the standards that we've put together are currently just in place on a private server, but that doesn't mean that we wouldn't be able to eventually share some of it out. I think, especially some of the mini guides that we put together, I can see those, at some point, maybe being publicly accessible. I don't really see that there's any proprietary issues with some of those. Yeah, a lot of its industry best practices, too, so it's not necessarily something that would be super unique to Home Depot, but it's really a combination of taking some of those industry best practices when it comes to tone, the way you convey a message, and then thinking about how that relates to the actual Home Depot brand to make it more specific.

J: Yeah, we're seeing more companies do that, like SalesForce has released their design system. I think LinkedIn is starting to be a little bit more open with the way that they're approaching things. The concept of industry standards is actually really interesting. If we think about the web as a whole, still a pretty young industry, right, but construction has been around since the beginning of time, and it's a lot of times those standards and the things ... You have an inspector come and look at your building to make sure that it's up to code before you can actually have anybody occupy it. I'm kind of interested to see where the web goes in terms of developing those types of standards for things like accessibility and interaction patterns to make sure that we're building things in a way that's going to be durable when it actually gets out into the wild. Seeing how a company like Home Depot is approaching the way they're releasing product is obviously going to be a big contributing factor to how those overall standards emerge.

Kelly: Yes. I think it's really important for standards to be flexible and adaptable. The more that we would learn about, for example, accessibility, that might change how we approach a certain type of standard. Last year, my team did an accessibility audit where we just took some of the accessibility guidelines, just split them up among the different teams, and then really took a deep dive into the website to see how we may be adhering to those or maybe not adhering as much as we would like, and then coming up with a plan to address those. Yeah, with content or design, really, as we learn more about people, as we learn more about our customer base, about how people are interacting with what we're doing, we really want to be able to take things and adapt them and make them flexible, make the documentation flexible, too. Just always think of it as a working document that we would always want to be receptive to change with.

J: Yeah, it's a living guide. I think that's a really important thing to understand is that you have to always be learning and always be adapting and always be updating your approach and your standards based on the current state of things.

Kelly: Definitely, yes.

J: Well, Kelly, it's been great having you on the show today. I would love to have you back at some point and just kind of get an update about how things are working at the Depot, maybe talk about some stuff that you've released recently. I think that'd be a lot of fun. If somebody wanted to get in touch with you and learn more about what you're doing or connect and kind of chat some of these things through, what's the best way to get in touch?

Kelly: Let's see, they could contact me on LinkedIn, if you just search for Kelly Home Depot, I will pop up there. On Twitter, I am kellyreverie, all one word. Definitely feel free to reach out. I would love to talk to anyone about content strategy, design, how they intersect, any of that good stuff. Yeah, so really love what you're doing with the podcast. Thank you.

J: Oh, thanks. Thanks. Well, I'm glad to have you on the show and like I said, it'd be great to have you back on again in the future.

Kelly: Absolutely.

J: That's it for today and we'll talk to you again soon.

Kelly: All right, thanks a lot, J.