Retraining your ethnographic muscles for design ethnography
A journey from academia to practice.
The first ethnographic account I remember reading was a book called Body & Soul: Notebook of an Apprentice Boxer by Loïc Wacquant. It’s an ethnography about a young doctoral student stumbling onto a boxing gym in the South Side of Chicago that quickly became an obsession for the author who’s now been teaching his craft at UC Berkeley for over twenty years. In the book, Wacquant paints thick descriptions of a social world centered on a boxing gym where young men find a home, a sense of stability and safety that’s otherwise missing in their domestic environments. He draws on empirical evidence collected through years of participant observations to illustrate how our socio-spatial environment shapes our habits. Pierre Bourdieu calls this habitus. Wacquant who was still uninitiated to boxing in the opening pages of the book closes with a chapter describing his experience as a fighter in the Chicago Golden Gloves, one of the most competitive boxing tournaments anywhere. If this isn’t participant observation, I don’t know what is! Writing good ethnography takes time. Like boxing, it also takes training and practice. Theory only takes you so far. These are sports that you learn by doing.
Wacquant’s book unknowingly drew me as an environmental designer to ethnography, design ethnography, and design research; three investigative tools intricately related to UX and whose names are often misused to describe anything from phone interviews to discovery meetings. Here, I propose to draw connections between ‘ethnography’, ‘design ethnography’ and ‘design research’ by taking you through my own journey as I describe and illustrate what these fields of research are really about for me.
Stepping into ethnography
My interest for Body & Soul and fascination for ways in which design and science intersect brought me to Baltimore in 2011 to kick off a research project of my own. I was a fresh PhD student then and intent on learning about ways that environmental design influences matters like marginalization, urban poverty and racism. I came to Baltimore via the HBO show The Wire. This was just around the time when Harvard launched a sociology course developed around the tv program and University of Michigan organized its symposium, ‘Heart of the City: Black Urban Life on The Wire’. Academia was all over it. Although I didn’t know it at the time, my project would become a backstory to The Wire where I studied how different housing and neighborhood design programs helped create the type of segregated, over-policed and disproportionately poor neighborhoods presented in David Simon’s The Wire. The dissertation turned out to be an archival research, but I defined my topic thanks to an ethnography that I conducted in West Baltimore.
Building this research also meant building new muscles as an ethnographer. I had done enough fieldwork before, usually spending a few weeks in one place at a time, but the process wasn’t ethnographic per se. It was in fact closer to what I do now, design research. I direct my own design research agency called workroom B where we lean on design ethnography, spatial analysis, different types of interviews and participatory workshops to generate actionable insights for place shaping projects or ideas. Like my work in Baltimore, our approach at workroom B starts from a place of empathy and a stubborn curiosity with the connection between social and spatial, between human behaviors and man-made things. Whether these ‘things’ are policies, design or architectural objects doesn’t really matter to me as long as they affect each other. The philosopher and anthropologist Bruno Latour likes to say that ‘things’ are actors too. I think he made this point for the first time in a paper about the sociology of a door-closer! Following Latour, I study humans and things together and simply refer to them as ‘actors.’
How do you learn ethnography?
Mark-Anthony Falzon (2009) writes that ethnography ‘privileges an engaged, contextually rich and nuanced type of qualitative social research, in which fine grained daily interactions constitute the lifeblood of the data produced.’ Ethnography is a qualitative type of research that collects information through different means over months, sometimes years. (What Erika Hall calls ‘Minimum Viable Ethnography’ in another Medium post is arguably not ethnography at all. I don’t know of any trained researcher that would agree with her contention that ‘any efforts to learn about people in their social and cultural context can merit that label as long as the score and standard are clear’. What she describes are semi-structured telephone interviews, nothing more, nothing less. With deep training and experience, one could conduct a semantic ethnography of verbal answers, but her process leaves out any analysis of language and thoughts required in that approach.) With the fundamental necessity of time, ethnographers are able to gather insightful data through direct observations. They look for patterns, for behaviors. This yields a very different type of information compared to interviews as our data in ethnography is not always ‘verbal’. As Steven Levitt likes to remind his Freakonomics audience: people lie, to us and to themselves, or give us answers to questions they’ve never really thought about. That’s a big reason why it’s important to think critically about the data we collect in qualitative research.
Ethnography is unique in its ability to make sense of user behavior but that doesn’t mean that we can’t complement our data through other means of research. In fact, complementing participant observation with other types of data is a good way to create descriptions thick with valuable information. To create these thick descriptions, we use field notes, journals, audio and visual recordings, cultural artefacts, and unstructured interviews that when conducted well, sound just like a natural conversation. These conversations take place in the interviewee’s own environment and between people that have an established report. There’s a focus on exchanges taking place 1-on-1 in ethnography to grasp how people think, rationalize, act from and so forth, but the objective is always to get a wider view than only one person. All of this matters: the physical location, the people present, the relationship between researcher(s) and actor(s), and the relevance of what’s being study on larger groups of people.
There are some ground rules in ethnography and getting better at it also means knowing when to break them.
Analyzing these fragments through an informed and reflexive and theoretical lens is also central to ethnographic work. This allows us to write our ‘ethnography’, which is the name we give to accounts like Body & Soul that include thick descriptions and their analysis. There are some ground rules in ethnography and getting better at it also means knowing when to break them. To get better at it, we need to practice alongside a good supervisor that teaches us on the job, so to speak. Like sculpting and fighting, you learn the craft through practice, by doing it again and again under capable supervision. This is usually something that you learn in doctoral studies.
Taking ethnography to Baltimore
I spent roughly three months over a two-year period carrying out ethnographic work in West Baltimore, the area where most of The Wire unfolds. I wanted to observe how much of The Wire rang true. I wanted firsthand data to contribute to a conversation about the use of fiction to study real life situations. I typed up my findings in a journal article and shifted my attention to a different, albeit related topic. My next challenge was to find out how all the dystopia portrayed in the tv program affects the neighborhood as a whole. By dystopia, I mean the open-air drug trade, double digit unemployment rates, lack of basic public services, everyday violence and over imprisonment in neighborhoods pleagued by under-funded public schools and under-trained policemen.
Having a good feel of the city at that point, I decided to focus on a West Baltimore neighborhood called Sandtown. Rather than following a set of people, the actor I studied was the neighborhood itself. My daily routine usually consisted of spending mornings at the city archives in East Baltimore, breaking for a late lunch, and using afternoons and evenings to immerse myself in Sandtown. I trained at a Fells Point boxing gym for a couple hours almost every day, maybe to embody Wacquant and certainly for my love for the ‘sweet science’.
I met many of the people that I would develop a relationship with at the ‘Family Fun Festival’ held at Franklin Square where residents gathered around art, food and music. In tight knit communities, it often only takes one person to welcome you in and I was lucky to get two. The first was an older man married to one of the festival’s organizers. He let me into his world of builders, of communities and rehabilitated houses. The second was a little younger than me and seemed to know just about everyone. Together, they brokered introductions and allowed me to spend time with the neighborhood and the actors within it. With one, we moved through a world made up of older folks, community leaders and construction workers. With the other, we hung around the neighborhood waiting for time to pass. I knew enough to tell when it was time for me to disappear for a minute or to fetch snacks. I never became a true part of the neighborhood, or turned a ‘they’ into a ‘we’ as the anthropologist Bronisław Malinowski would have it. People felt more natural around me as time went by but my presence had an effect on the actors involved nevertheless. Recognizing and accounting for your effects on actors is another central part of the craft.
Because The Wire centers in on dysfunctional aspects of Baltimore, many actors are left out of the narrative. Having studied the tv show so intensely, I met people in Baltimore that already felt familiar but I also got to know people left out of HBO’s fictionalized Baltimore who added an invaluable layer of information to my understanding of the city. ‘Ethnography is inductive rather than deductive,’ (Cranz, 2016), so discovering new insights leading to hypothesis formation is part of what makes it a discovery science. In other words, ethnography is more about developing hypotheses than testing them. There were people like Martine who had just wrapped up a degree in pharmacy and decided to open up a practice right in the neighborhood. I spent time with mothers, some gainfully employed, others not. I spoke to recovered addicts for hours at a time over days if not weeks. They let me in on their daily lives, revealing their demons, blue skies and backstories. Some of them allowed me to record video interviews, while others I spoke to conversationally with no real objective other than getting to know them better. What struck me the most was how many actors regarded dysfunction in the neighborhood as normal, and regarded what we define as “normal” in middle-class America as being exceptional. People would speak of potential improvement in the humblest of terms, dreaming up a world with slightly less vacant housing (entire blocks are boarded up in Sandtown), fewer drug addicts and better schools, conjuring spatial imaginaries still miles away from even the most subdued versions of the American dream.
My ethnography of West Baltimore included videos, photographs and texts to interpret and support what I had captured as an observer, participant, interviewer, researcher. This was my multimedia thick description that I analyzed through social theories conceived to enhance our understanding of race and class issues. The books on my table had Anderson, Hirsch, Vale, Venkatesh, Wacquant, Williams and Wilson printed on their spine. Processing the material through a social scientific lens is a required part of the trade. This, to me, is the ‘minimum’ for ethnography to be viable. It’s a long way from Wacquant’s study at the Woodlawn Boys and Girls Club in Chicago, but it at least builds on the fundamental principles of the discipline.
From Ethnography to Design Ethnography
My role as a doctoral student and researcher afforded me a kind of support, space, and freedom that I’d have a hard time finding in private practice. Design ethnography, on the other hand, allows us to gain behavioral insights into design matters relatively quickly. I mean, days or weeks instead of months to years. We’re able to do this by conducting immersive research through the experience sampling method where we frame our observations more narrowly to study how people interact with a single actor, oftentimes a design object or space not bigger than a room. Commenting on an earlier draft of this text, my colleague Dr. Fred Ariel Hernandez proposed the term ‘microethnography’ instead given the very limited set of circumstances being studied. I’m already liking this term better, even if I’ll continue using design ethnography until at least the end of this paper!
For Bella Martin and Bruce Hanington, ‘Design ethnography approximates the immersion methods of traditional ethnography, to deeply experience and understand the user’s world for design empathy and insight.’ The objective is to decipher ‘patterns and themes emerging from research materials, and articulated in a set of design implications or guidelines in preparation for generative research and concept development.’ (2019) While we could debate whether there are any hard and fast rules defining what design ethnography really is, I would propose that any research labeled with this term needs to include data gathered through behavioral observations in situ while providing insights that can be generalized to some degree. Simply put, if your data is fully verbal and coming in the form of answers to questions, or collected in a lab, your work is not likely to have much ethnographic value, if any. And the same goes for insights that only ring true to one person. It may still be incredible qualitative research (or user research like we often call it in the industry), but that doesn’t necessarily make it ethnographic.
Our research in design ethnography supports making and makers of ‘things’.
Because the stakes in design ethnography are different than with conventional ethnography and its derivatives (reflexive ethnography, autoethnography, multi-sited ethnography, etc.), the methods are different too. Design ethnography primarily intends to inform design decisions, not social theory — though it can have revelations and implications. Generally speaking, we do this by studying how people interact with a prototype, an object, or engage with an idea to figure out how we can create something of value. Our research in design ethnography supports making and makers of ‘things’.
While design ethnography departs from ethnography in particular ways, it nevertheless needs to align with the defining principles of the discipline to justify its use of the qualifying label, ethnography. This means, recording, describing and analyzing behavior observed through firsthand observation in situ. Like with ethnography, we thicken our findings through other modes of qualitative or quantitative research. At its core, design ethnography presents an emic perspective, meaning, a perspective derived from an insider’s point of view. For that emic perspective to emerge, the researcher needs to at least have a basic awareness of the conditions required by the field that have been defined through over 100 years of practice and discourse.
How I used design ethnography to redesign a chain of Corona stores in Mexico
In the fall of 2015 I had just embarked on a new project as a design consultant. Our challenge was to find ways to reduce the energy use for a chain of over 100 convenience stores owned and operated by Mexico’s most popular beer company, Corona. For the project, I was joined by a colleague that specializes in low-energy buildings. This was a true design research project, borrowing methods from different fields to generate insights, ideate interventions and implement a clearly defined design strategy. One of the methods we used in the research phase was design ethnography.
We defined a project schedule where we’d use one week to conduct desk-based research and plan out the project, followed by multiple weeks in the field to collect primary data and test out prototypes, and the final few days compiling a report from the Los Angeles office. Because we can only speak a little bit of Spanish, learning from observation came naturally. We knew that we couldn’t look at each of the stores individually, so we worked with the client to select a location that offered a representative model in terms of design, sales volume, building typology and energy use. That would become our case study and pilot store. We spent a few days looking at the extremes; stores that used significantly more or significantly less energy than any other, but it wouldn’t have been particularly productive to focus on the exception for a project like this one. These stores generally performed differently based on three indicators: sales volume, building types and poorly installed equipment. As for performance, of course cooling down more beer requires more energy. And it didn’t take long to figure out that convenience stores built through retrofitting old houses would be less energy performant than new buildings, or that ice machines dating from the 1990s that never had their door gasket replaced would leak cold air and need more energy.
To begin, we wanted to understand how energy moves inside the store. We connected some homemade sensors to create an energy audit that would give us hard data on energy uses for every electrical device (AC, freezers, lighting, etc.). Collecting this baseline information was not only critical to seeing the big picture, but also necessary to calculate the effectiveness of the different design experiments that we planned to conduct later on. That would become our control data. While the sensors did their work, we hung around the store, rigged prototypes, helped the staff here and there and observed. We took notes on everything and nothing. On how the team propped the doors open to receive dry deliveries; how staff members re-adjusted the AC temperature throughout the day; how they used their downtime; we looked at customers buying and popping popcorn bags in the store’s microwave, etc. Part of me wanted to get a bag of popcorn of my own, and part of me wanted to prop the door open to let the smell of cinema butter get out. It took self-control to resist doing either, not because intervening would have compromised the data (this is participant observation after all), but rather because I wanted to see how long it would take for anyone else to break!
We also conducted some spatial analysis to understand the logic behind the hundreds of micro design decisions taken in building the store. Why build a store with wide (and costly) windows if you’re going to block them with floor to ceiling bags of chips? Why use reflective paint on certain walls but not others. Why keep the lights inside the fridge turned on overnight? We defined these architectural and system design questions through observations, but generally found answers by asking. Not everything is behavioral after all.
After observing, measuring, analyzing and having short discussions with staff and clients, we ideated over a dozen design interventions to test out. These ranged from changing the temperature settings of ice machines, to repainting the building’s roof from terracotta to white (see photo at the top), to making homemade stickers for the AC remote that asked for the temperature to be set to a certain degree. We implemented each of our prototypes one-by-one so that we could measure their individual impact on energy use (quantitatively) and on people (ethnographically). What we needed to know was fairly simple: whether the strategy would lower the carbon footprint, and whether it would have any impact on the staff and customers. Because the sensors measured how each device was being used whether we were inside the store or not, we were able to assess if our presence in the store changed the way actors used the equipment.
Our second cheapest prototype was the remote-control stickers. It was also one of the most successful! Prior to applying the stickers, employees re-adjusted the temperature throughout the day constantly. If you’ve ever shared an office where you can control the thermostat, you know exactly the kind of micro tensions that this can cause. One employee in particular loved it really cold. Like 18ºC (64ºF) cold. Moving from that kind of temperature to the outdoors is physically tiring for most. One manager wore her Corona parka, provided to restock the cold room (proudly kept just below freezing as per the customers preference), while she worked the cash register. This is Veracruz, Mexico where the average highs in December hover around 27ºC (80ºF). That really wasn’t just a carbon footprint problem, but a workplace comfort problem.
After testing different temperatures, we settled on 23ºC (74ºF). It was refreshing enough to draw clients in and warm enough for employees to dress normally. Sure, some people would have ideally liked it a little cooler or warmer, but 23ºC was the communal sweet spot. We could have used our insights to define a more advanced type of intervention, like creating a system to control all store temperatures remotely from the headquarters (that technology certainly exists), but the stickers worked just fine. They cost close to nothing and left employees with some level of autonomy rather than taking away the AC remote, a strategy that would have been demeaning and patronizing. A little guidance and direction was all that was missing.
In the end, we managed to implement a handful of solutions that lowered the carbon footprint by around 25% for over 100 stores, all with very low-to-no-cost interventions. This was a win for the client’s bottom line, for the environment, for employees and for the town as a whole as surges in demand triggered by big consumers takes the power grid down often in this part of the world.
Ethnography offers a window into people’s behaviors by observing and interpreting how actors engage and interact. Its lifeblood can often be non-verbal information collected in situ as ethnographers interpret actions, expressions, movement — what is said and what’s left unsaid. This takes training, experience and toolkits.
As David Travis writes on Medium, design ethnography ‘doesn’t follow all the rules but it gets the job done.’ As design ethnographers, we’re able to work relatively quickly by narrowing on specific interactions to study. Instead of looking at the entire social network of a small group of people, we focus our gaze on how people interact with one given actor or variable: an object, a space, a policy, etc. We can learn about how actors interact with a new piece of equipment, a chair, a waiting room without knowing much about their life outside of that meeting room. This allows us to generate invaluable insights relatively quickly, especially when we’re skilled in ethnography and design. Why? Because designing means thinking projectively. Designers find it easy to imagine things that don’t exist. Being able to do both of these things well makes it easier to identify what insight is actionable and imagine what that new action might look like. It allows us to quickly translate a research insight into a design prototype. This new prototype–a model, a mock-up, lo or hi fi–is nothing less than a new actor that we can deploy and study again through design ethnography. That’s one sure way to create things that work because they matter and sustain.
— — -
I wish to thank William Hunter, Dr. Fred Ariel Hernandez and Dr. Nerve Macaspac, who shared precise and productive comments on an earlier draft of this essay without necessarily agreeing with all of its arguments.
Dr. Benjamin Leclair-Paquet is a researcher currently leading the design research consultancy, workroom B. He honed his craft through doctoral research in environmental design at the Swiss Federal Institute of Technology (ETH Zurich), a 4-year research residency at the SEC Future Cities Laboratory, graduate work in design and development planning, professional work consulting for small and large companies, and as a Faculty member at University College London (UCL) and the University of California (UCLA).
