Info iconLetter I in a circle

By continuing to use this site you consent to the use of cookies in accordance with our cookie policy

Mule Design’s Erika Hall on interactive design and research

Director of Brand Design, Intercom

Stewart Scott-Curran

@stewartsc

When it comes to cross-disciplinary design, there's perhaps no one more insightful than Erika Hall.

Erika is the co-founder of Mule Design, where she directs research, strategy, and information design. She’s has more than 20 years experience web design and development, and is perhaps best known for authoring Just Enough Research. It’s a must-read for anyone whose work can be improved with better user insights.

I recently sat down with Erika for a wide-ranging chat. We cover the growing importance of language in design, how trust is crucial in a client-designer relationship, and why – when it comes to designing well – research simply isn’t optional.

If you like what you hear, check out more episodes of our podcast. You can subscribe on iTunes, Google Play, SoundCloud, or grab the RSS feed.

What follows is a lightly edited transcript of the episode. Short on time? Here are five quick takeaways:

  1. To design without research is to design without information. It’s absolutely necessary to make informed decisions.
  2. Erika sees three conditions for good design: a clear goal, the right resources and the right information.
  3. Effective language in interactive design should be lively. This is much easier to execute when written the way one human would speak to another, like a conversation.
  4. Users don’t care if their interaction is a bot, another person, or a web app. They just want what’s actually useful, pleasant and, most importantly, fast.
  5. Agencies should look for clients with a clear goal, not a brand that’s simply looking to experiment.
  6. Erika looks for designers who can express themselves clearly, because once hired, the ability to get all the people involved in a project to make a good decision is paramount.

Stewart Scott-Curran: Erika, welcome to the show. We’ve been big admirers of your work for a long time. Can you give us a quick overview of what you do and how you got started with Mule Design?

Erika Hall: Mule Design Studio is an interactive design and strategy outfit here in San Francisco that we started almost 15 years ago, which is a terrifying number to utter at this point. We help people and organizations use design to achieve their goals and work with a really wide variety of clients. I’ve been in client services for a long time, and I really enjoy it because I love solving problems and I love learning about new fields.

Every new client we work with is like a new course of study, which is really exciting to me. I meet some really fantastic people. It’s just a lot of fun. I also have a really short attention span, so I go in for a project, do some work, and get out. I studied philosophy. If you don’t want to be a lawyer, you really have to make up your own field following that. I’ve really found that the abstract thinking and argument creation that I studied in school is fantastically applicable to interactive design.

Stewart: We reference Just Enough Research a lot at Intercom. It’s probably one of the defining books about user research that we’ve seen recently. Why you decided to write that book?

Erika: It’s hard to get clients or business people interested in the idea of research because it just sounds like something that’s academic, like homework, and not applicable to the day-to-day. I came out of a very collaborative, research-driven design practice at the first agency I worked with. It seemed like the obvious way to work.

Then, once I started working with this wide array of clients and seeing the same sort of objections, I found the need for there to be a very, very brief book that explains why it’s important to base your design decisions on evidence and a little bit of how-to, because most research books are written by researchers for other researchers and they’re 500 pages long and cost $75. That’s a hard sell if what you’re trying to say is, “Hey, research is something that anybody can do on any budget in any amount time. Here, I’ve given you 20 pounds of books.” Nobody else had written the book, so I said, “Well, I guess I’m going to have to do it.”

Stewart: That was three years ago. Is there anything in that book where your thinking has advanced or changed on that since writing it.

Erika: When I started thinking about and writing that book, I thought, “I need to convince people to do research.” Then, after thinking about it for some time and sitting with it, I thought, “No, it’s not that people have to do research. It shouldn’t be a separate thing. It’s that what we’re doing is design and we’re getting enough information to make good design decisions.”

If you talk about research separately, people have this sense that the work has to live up to the standards of academic research or pure research, and it doesn’t. The whole reason you’re doing this is to make better design in service of more successful businesses or organizations.

It hasn’t shifted any of the practices. It’s shifted the framework so that people really have top of mind that the goal is a good product or service. The goal is not to meet any sort of research standard. I’ve been writing and talking about it a lot more in that way to say, “This is a necessary part of design,” because as long as we talk about research as an add-on or a separate thing, even if you say it’s a very important separate thing, that makes it sound like you can do design without information. That is not true.

The pillars of good design

Stewart: As a designer, we’ve probably all been guilty about thinking about research as a necessary evil or something that’s trying to push us in a particular direction that the “artist” in us doesn’t want to go. One part of it the book that was interesting to me was when you talk about setting the conditions for good design: a clear goal, the right resources and the right information. Research obviously plays a lot into the latter, but how would you define the others?

Erika: I’d say we talk a lot about the goal, having a clear goal. That sounds so obvious, but what we find when we work with our client organizations is you might have a goal like increase the number of users, increase customer retention, something like that, but everybody’s going to have their own individual goals. It is actually really challenging to clearly distill what you’re trying to accomplish into something that everybody working towards that goal is in agreement on.

That’s absolutely necessary, because if you don’t do that, and you see this in a lot of organizations, even brand new startups, it’s like, “What are we here to accomplish?” That actually takes some thinking and some investigation. A lot of people who want to move fast say, “We’ll just do a lot of work and then our goal will emerge.” That’s not how people conduct any other aspect of their lives.

Stewart: There’s an attitude about continually moving fast, especially with smaller companies, but it doesn’t necessarily always lead to a goal or intended outcome.

Erika: You get nowhere fast.

Stewart: As a designer and thinking about design work, we often think good design just sells itself. It’s going to speak for itself and everybody’s going to be on board with that. Obviously that’s not true. What makes a compelling design story, and are there any nuances that are especially important with dealing with clients?

Erika: Absolutely. This isn’t often a topic that in traditional design education you touch on at all. You learn to talk about design with other designers. You have to figure out how your work fits into what is important for your client or the people you’re working with.

There’s only good with reference to what you’re trying to accomplish.

Sometimes designers create what they think is good design against some abstract set of criteria for good design, but they don’t really ask, “How is this good design for the business I’m working for? How does it help them really meet their goals?” You have to start with their goals and their language and their sense of what will make something successful and fit how you’re describing your work into that. That is an essential part of design because there’s no good, there’s no absolute good design. There’s only good with reference to what you’re trying to accomplish.

Stewart: Designers spend a lot of time talking about design to other designers, which really doesn’t help. I think it just perpetuates an echo chamber type situation.

Erika: It really does. I’m really interested in solving problems. If you’re helping solve somebody else’s problem, the way to do that is not to make yourself feel comfortable or make yourself feel superior. It takes a lot of personal introspection and work to get over yourself as a designer and to say that just because you have a particular practice and you have a particular domain that you’re comfortable in and the people you’re working with are comfortable in a different domain, that somehow sets you apart or makes you better. It’s on you to close that gap because your work would not exist without the business or the client who requires that design.

The growing importance of language

Stewart: Earlier this year you stopped by the Intercom office to give a Creative Mornings talk, “Stop Writing for the Web”. You explained how interactive design falls a little bit flat when we create language that’s literary rather than what real people use in real life. There’s an interesting crossover with our mission at Intercom, which is to make internet business personal – language is a big focus of ours. What’s wrong with the status quo, and how should we be approaching language in interactive design?

Erika: Humans, fundamentally, we interact through language, but a weird thing happens as soon as we start to think about what we’re doing as writing. This whole other set of rules and restrictions and standards comes into play. There’s been this historic split in interactive design. There’s visual design and there’s coding and there’s interaction design, and then there’s content.

Content is a thing that you fit into a system you’ve already created, but the actual truth is that meaning is what drives interaction. The other stuff is just the framework for that. What’s happened is that when people think about language in terms of creating a set of documents that you pipe into an interactive experience, it’s really easy to create things that are really flat. That’s because you’re writing something that’s a thing, not a lively interaction. It really should be a lot more like a conversation.

The conversation should be the thing driving the interaction because it should feel like a natural conversational flow. If you don’t start from that place, if you start from the place of like, “Oh, how do the interface elements behave or move around, or what’s the layout?” and you don’t start from, “What does it mean, and what’s the interaction at every point from a verbal perspective?” everything gets really, really disjointed and creates a lot of problems. Virtually no teams are set up such that it starts with the interaction and language and then drives towards how that looks or how interactive elements behave.

Stewart: Would you make any suggestions to a designer or a team who wanted to try and think more along those lines?

Erika: Start with a role-play, because people always do this role-play of like, “We’re going to create personas, or we’re going to create user stories, and then we’re going to run them through this interaction” So many of the systems that we’re creating are proxies for human interaction. We don’t interact with travel agents any more. We get on Orbitz or Travelocity. If your system is a service, it’s a proxy for somebody else. Actually sit down and have a conversation and say, “What would somebody ask, and what would they expect the response to be?”

Write it like a play to start with and then say, “What do we need to provide at every stage?” and then, “How do we use visual design or interactive design to support that conversation that a person is having with a system that’s standing in for another human?” Start from there. Don’t start from drawing a lot of boxes and then trying to fill those boxes with words. That is a very uncomfortable situation for designers that have been brought up through graphic design or visual design with a sense of, “We create a form and then fill the form with words.”

Where bots win and lose

Stewart: There’s an added layer to that with the rise of the bots. Many of these early chatbot designs are falling a little bit flat. They’re designed with the aim of being too human or trying to replicate the human experience rather than solving a particular problem. What’s your point of view on those, and what a good direction to take them in would be?

Erika: The reason why bots are suddenly so popular is because messaging apps are so incredibly popular. People do not call each other anymore. People text each other. They’re spending all their time on these messaging apps. People are using Slack now instead of email, and so everything is going to this interactive chat messaging place. There was an idea of, “Everything’s going to be a chatbot now.” That’s because these hype cycles all depend on, “This is the next big thing. This is the silver bullet.”

I’m surprised we’re still talking about bots and not like everything is an augmented-reality, monster-catching game now. What it comes down to is nobody actually cares. No individual human cares if their interaction is a bot or is another person or is a web app. They just want to do the thing that’s actually useful and actually pleasant and interesting and, most importantly, fast to use. There are times when having a conversation really, really is fast, but if you’re not having a conversation with another human the chances of the bot being wrong are really, really high because a human can adjust immediately to interaction.

A customer service person, if they hear you’re frustrated on the phone, they can immediately respond. If you ask them a question, they have this immense wealth of knowledge that is impossible to program into a bot. It’s not just the texting that’s exciting to people. It’s so many other things about the efficiency of that interaction and about getting that little drop of dopamine that comes from texting and all of that that keeps you addicted to it. The problem with bots isn’t that they’re being too human; it’s that they’re optimizing for the wrong things.

The problem with bots is they’re optimizing for the wrong things.

I played around with one the other day that was supposed to be a “give your coworkers props for things” bot, and it was super chipper, and I had a hypothesis. I thought, “I’m going to put this in our Slack room and I’m going to put it in a personal slack channel I’m on, and I bet people are going to turn homicidal in like five minutes.” I was absolutely correct. The bot was only human in the most shallow way, because it didn’t talk like any actual human would talk.

Say a person in your office came up and was super chipper and said, “Hey, I just want to let you know that Bob thinks you’re awesome and super swell!” If every interaction you had with this person was in that tone of voice, you would want to punch them in the face. The people creating this bot had no deep understanding of how humans actually want to interact or what makes humans feel validated at work. They’re like, “We can just make a bot for this and it’ll be super cheerful and people will like that.” The answer’s no. They got funding for this thing, and I think it was because everybody was like, “Everything’s bots now because people really enjoy texting.”

You have to look a little bit further to say they don’t just enjoy texting for its own sake. It’s because there’s a payoff, and there has to be a payoff. It has to be faster. It’s not faster to order a pizza using a bot that’s chatting with you than it is to just go online and click things in a menu. I think it’s that failure of really understanding what problem they’re solving and really understanding what makes something human. What makes something human is the fact that it recognizes your needs and responds to your needs and is very quick and very efficient.

It’s those values that are really, really important, not something that chats with you and has a few automated canned responses. I talk about Douglas Adams a lot because he was so right about so much in Hitchhiker’s Guide to the Galaxy. He talked about robots with real people personalities. There was an elevator that was incredibly cheerful and chipper and like, “I’m so happy to be taking up to the next floor,” and doors that would sigh happily when they opened for you, and how this made all the actual humans full of rage.

Everybody working on an interface absolutely just needs to go back and read Hitchhiker’s Guide again, because Adams was right about all of this stuff way before it was getting real. Now it’s getting real, and everybody’s like, “This is going to be the next big thing. Everybody likes to chat with each other, so people are going to want to chat with really, really stupid, cheerful canned robots.” It’s not true because what people want to do is get things done really fast.

Stewart: And easier. The temptation is that with this new technology it becomes a trend and we want to just implement it everywhere, where it’s not required or where it’s not actually useful.

Erika: Exactly. Alexa turns on our lights now because we have the Amazon Echo. It’s working now, but for a while you’re saying, “Alexa, turn on the lights,” and Alexa’s like, “I’m sorry, I can’t play you that John Major song.” It’s the time it takes and the ambiguity and all of those things. Ambiguity and uncertainty increase the perceived amount of work. A lot of these bot makers are missing how much work they now require on the part of the user to figure out what answer they need to give to the bot to get what they need.

The right kind of clients

Stewart: You’ve said previously that a studio is its clients. Obviously, there’s a lot of subtleties to client creative relationships that are unknown at the beginning. How do you go about selecting your clients at Mule?

Erika: We actually have a really clear set of criteria. Having applied these over the years means that a lot of the people who reach out to us at this point are aware of the kind of work we do and the kind of client relationships that are going to be successful. The most important thing is that the client has a clear goal. We can’t help anybody who doesn’t have a goal. For this reason, a lot of times we don’t work with too many early stage companies because it’s more like they want to run an experiment.


Mule Design Studio’s redesign for The Audubon Society’s website

Then the organization has to be doing something that we really feel good about having our name associated with. This doesn’t just mean like a mission-driven organization. It could be a game, it could be a business that’s just selling some useful software, providing a useful service to people.

The people we work with have to absolutely believe in what they’re doing and not just cynically come to us and say, “We found this niche that we really want to exploit, but we don’t really feel like we’re solving a real problem for these people,” because if the people we’re working with don’t really believe in what they’re doing, that won’t be a good client relationship. In conjunction with this, it has to be a problem that’s really essential to the business. If somebody comes to us and they’re like, “We’ve had a little extra budget in Q4 and we want to try something,” that project will go terribly because nobody will be really invested in it.

What clients don’t really get is how much they need to participate in the process. A lot of times people will come to us and say, “It’s this magic. You’re designers, which means that whatever you do is like magic sprinkles on hard business problems.” No, what we’re doing is we’re helping you do your business better. We’re working in partnership with you. It’s up to the clients to really bring the knowledge and challenge every one of our ideas just as much as we’ll challenge every one of their assumptions.

If somebody comes to us and they say, “The future of my organization depends on the success of this project. My job is on the line. This absolutely has to generate results,” that project will go well. If somebody’s just like, “Eh, we’re just going to try this,” it’ll be really, really frustrating. Then, there needs to be trust. I feel terrible when we meet new clients who’ve been burned by designers in the past. They made big promises, they didn’t deliver, or we talked to one set of people and then a totally different set of people did the work.

That’s why we’ve stayed really small. Everybody who works with us is fantastic and there’s no secret army of contractors on the other coast or something like that. We all work very collaboratively together to solve problems, but it’s really hard if somebody comes in and they don’t trust us because then it’s this huge overhead. People want to know what makes projects cost less and go faster, and it’s having that trust and it’s making decisions quickly. This true of any designer you’re working with. If you don’t just trust them, then they’re going to spend all their time trying to placate you and build your trust, and that is time away from solving your problem.

On hiring well

Stewart: I know you’ve got some openings at Mule Design too, so hiring is top of mind. How do you assess what makes a great designer outside of just having a slick portfolio?

Erika: We look at the portfolio fifth of all the things we look at. There’s a lot of emphasis on having a portfolio, but a portfolio is not necessarily meaningful, especially for our work. We’re not designing flat images. It’s really easy to look at applications or online services as a series of pictures. We’re interested in quality of thinking, and we’re interested in designers who can express themselves clearly, who are really intellectually curious, who really want to grow, and who not only want to think about the what.

So much of design education is focused on the how. There’s a magazine called How. As a designer, you need to be focused on the why. If you don’t know the why then you don’t know which how to apply. Good design, good craftsmanship is part of that. The artifacts are part of it, but it’s the ability to help people make good quality decisions because that’s the difference between good design and bad design. It’s not just the craft, but it’s the ability of the designer to get all the people involved to make a good decision. This is why there’s so much bad design in the world. It’s not because we have an enormous lack of competent designers. It’s because designers are often terrible at helping the people they work with and for make good decisions, and so terrible things get out in the world.

Stewart: One thing that designers, especially as they’re fresh out of school, have to learn how to work alongside other people in technical disciplines like engineering, or research. What can we do as founders or managers to help enable those designers succeed?

Erika: The most important thing is to be clear about what will make you successful and let the designer know. What we’ve run into, especially with earlier stage companies, is the idea of, “We want you to just sort of speculate. We want you to create this blue sky solution.” It has to work in the real world, so be really, really clear about what design work needs to accomplish so you don’t end up with something that’s beautiful and delightful and lovely but doesn’t meet a goal and won’t help you for your users or for your organization and then won’t succeed.

People talk about the business value of user experience. User experience has no business value inherently. You really have to have total clarity about what you want to accomplish and be able to convey that to the designer and then evaluate their work based on that. Don’t just come in and say, “I’m a genius founder. Guess what’s in my head.” It’s surprising how many people who are otherwise pretty smart have that sense of, “I have this thing inside me and the job of the designer is just to make a beautiful rendering of what’s in my head.” That’s not going to succeed no matter how beautiful that rendering is, so be honest about those things and then be willing to have that give and take. Also be open to being questioned and be open to being challenged and let the designers you’re working with get you to a place that they can challenge you and make your ideas stronger.

Designer empowerment

Stewart: You do a series of great workshops at Mule Design that empower designers to talk about their work. Have you seen anything in organizing and running those that points to maybe a deeper level of understanding of design and how designers should carry themselves within companies?

Erika: The thing that we really see through any of these workshops is how common some of these situations are and that they have nothing to do with the skills or the craft of the individual designer and everything to do with the organizational dynamic. If an organization has a culture in which it’s possible to question authority and have open, honest conversations, design can work and be successful.

@stewartsc at #goatsalon tonight!

A photo posted by Mule Gallery (@mulegallery) on


Workshops and seminars, like the above, are a staple at Mule Design Studio.

I’ve talked to people who come from very, very political organizations where you actually can’t question how things are done or you can’t speak candidly, and so I think that’s what really comes through. Politics are responsible for a lot of really, really bad design. Nothing makes a business run more efficiently than honesty. Because like I was saying with trust, there’s an enormous overhead to not being able to speak openly in an environment of mutual respect.

That’s something people aren’t talking about enough – it’s so much work to maintain this consensus that everything’s okay when it’s actually not, or that people are actually working towards the same goal when they’re not. I’ve seen people who come from environments where they have a lot of fear about speaking honestly. When your staff is thinking like that, that means that’s where their brains are going. Their brain is going to, “How do I work the politics?” not, “How do I make my business successful?”

Stewart: Designers have a responsibility to themselves and the end users with what they put out into the world. It’s my point of view that designers need to feel empowered to question certain things that they’re being asked to do, even if it doesn’t necessarily fall within the realm of the pixels that they’re putting on the screen. Have you seen that become a little bit more prevalent?

Erika: I think so. That has been one of the positive side effects of the hype around design thinking and designer founders – a general sense that design is important. Even if people misunderstand what design is or the role of design or don’t have a lot of clarity, there has been a sense of, “We should listen to designers because they are problem solvers. They’re not just a pair of hands. They’re not just pixel pushers.”

That sense has been kind of pervading the business and technology world and making it better. I’d say it’s on designers now to really take advantage of this and don’t just take it as, “Well now I can do whatever I want and I don’t need to ask questions. I can just do things and say you have to listen to me because I’m the designer,” but should take that as an invitation to really engage in a dialogue about what the best course of action is.

Stewart: Erika Hall, thank you so much.

Erika: Thank you.