By MIKKEL KRENCHEL, ReD Associates
Three strategies for designing research that captures the social forces shaping people's behavior.
Remember the days when a main challenge of the EPIC community was convincing executives that humans weren’t just rational actors all the time? Back when arguing for the value of ethnographic research, thick data, and so forth, started with getting executives to realize that there was more to people than what could be observed through a spreadsheet?
Fortunately, those days are long gone. Today, most successful leaders of large corporations readily embrace the idea that humans are complex, emotional creatures and that the success of their business in large part rests on making the right bets on how they will behave. In response, research departments across the corporate world have grown exponentially in both size and sophistication, and ‘ethnographic research’ as a term has almost gone mainstream.
It would be easy to conclude that it’s time to declare victory. But if you look a little closer at the growing volume of ‘ethnographic’ and other qualitative research conducted in industry, it soon becomes clear that something isn’t right. I’m sure you’ve seen it yourself: What starts as a big, juicy, and open-ended question about the social, cultural and/or ethical underpinnings of a product or business often ends up being answered with – wait for it – a list of needs? A set of ‘jobs-to-be-done’? Perhaps a description of the ‘user journey’ detailing the emotions, challenges, and aspirations of some persona?
All too often, the output of applied qualitative research seems to stop at a description of what people think and feel; it examines individuals and their emotions rather than group dynamics, culture, and the social forces and context that shape so much of people’s behavior.
You can see this dynamic in the evolution of research methods too. Researchers, often academically trained in deep, rigorous ethnographic explorations, end up doing studies consisting of X number of ‘ethnos’ (read: long semi-structured interviews), some card sorting exercises, and perhaps a diary. They focus on methods designed to elicit a rich portrait of the ‘respondent’s’ worldview. But that means they end up with little time left over for observing actual behavior and what happens when the ideas and ideals we hold in our minds clash with other people and the realities of everyday life.
We’ve taught business leaders to go beyond the ‘rational actor’ approach and ask the critical social and cultural questions. But somewhere along the way, the anthropological explorations often become a superficial psychological investigation. Why?
There is, of course, the obvious answer: because focusing on individuals seems easier and cheaper. Research teams are pushed to deliver answers fast. They don’t have the time and resources it takes to do real immersion into the social context of their respondents. Stakeholders on the product or the business side have clear hypotheses that need answers and don’t have patience for open-ended investigations. Or worse, the stakeholders don’t trust that the researchers are capable of providing useful, directional, or true insights without a very tight, explicit and replicable research frame.
I suspect there may also be a deeper issue at play: When (Western) executives, and to some extent researchers, try to make sense of people, it seems they often default to an implicit philosophy of individualism. We may have accepted and internalized the premise that humans aren’t always rational but have yet to fully accept the ramifications of the fact that humans aren’t always alone. Anyone who has had the least bit of exposure to the social sciences – or simply a well-trained eye for observing the world around them – knows that people are deeply influenced by their context; that social structure and power matters greatly to our behavior and experience of the world; that groups and cultures have emergent properties that are more than the sum of their parts. I could keep going. But those kinds of considerations not only have the potential to make research complicated and unwieldy; they also clash with our desire to see all people (including ourselves) as free agents who are ultimately in charge of our own actions.
Of course, focusing on the psychology of the individual is not bad per se. Making sense of the complexity of human experience is a big improvement over the previous rationalist approaches to market research. And there are many cases where describing the world from the point of view of a single person is instructive, both as a way to make sense of what’s going on and as a communication tool to build empathy. The point is simply that this approach alone is insufficient. It causes us to lose sight of all the ways the behaviors of individuals are influenced by our context and other forces around us we cannot see. And as importantly, it means research often overlooks emergent social phenomena with potentially massive implications for strategy. Think about some of the big changes that have been reshaping business and society just this year: emerging norms around hybrid work, the rise of meme-stocks and the changing power-dynamics in the stock market, the spread of Covid mis- and disinformation, to mention but a few. To make sense of how they play out we must observe not only what people think and experience but the fundamental changes taking place in our social fabric.
So what can we do? Getting the business world to embrace the idea that the world is socially constructed is no small task. As researchers, we need to show that this can be done pragmatically: that in an increasingly interconnected world, this leads to better outcomes and concrete decisions, not just deeper theorizing; that we have a language for talking about the research but also that the outcomes will be even more pragmatic and helpful. More work on this front is needed, yet one practical place to start is to reframe the ways in which we talk about and frame ethnographic research.
Here are three initial suggestions for how:
1. Study ecologies, not individuals
A good place to start could be to more systematically switch the unit of analysis in research from participants themselves to the social ecologies in which they take part. If you want to understand the role of pickup trucks in modern life, you are better off recruiting cars than people. If you want to understand the home as a living organism, recruit households, not customers. If you need to understand how small businesses make decisions, recruit companies, not COs. In all cases, snowball through the social ecology and ensure you get to observe how people within the ecology interacts.
Focusing research and analysis on ecologies rather than individuals also makes it easier to go beyond flat descriptions of what people need or want to do – to go from ‘the customer needs X’ to ‘Y makes the customer need X’. An emphasis on ecologies thus allows us to more easily observe emergent group properties and what venture capitalists like to talk about as the “flywheel” effects of how social groups interact – that is, the effects that sustain communities and build social networks. For example, if we want to understand what pet owners want, we could study their individual need for companionship but might miss the social role of pets, which often serves as an instrument of community and family building – an antidote to growing mental health and isolation in the West. Providing new opportunities for communities to come together around their pets would not only help pet owners get even more connection and companionship (from both humans and animals), but would also expose more non-pet owners to the social power of pets and thus make the flywheel spin. But exposing those kinds of insights requires a thick description at the level of how pet owners act as a group, not just who they are as participants.
2. Stop worrying about sample size and start talking about sample depth
Secondly, we need to end this obsession with sample size, or n. Qualitative researchers all know that the exact number of people in a sample beyond a certain minimum threshold doesn’t really matter. That they are looking for qualities and to explore variation, not quantities and sizing representation. Too often, researchers box themselves in by committing upfront to an n of, say exactly 18 participants (or ecologies), only to blindly drone on until they’ve done what they set out to do. They lose the explorative impulse, or the agility to dig deeper for the story because they obsess over mechanics. But stakeholders are looking for some way of making sense of the depth and rigor of the research, and sample size is the language they know. Instead of promising an exact number of individuals or interviews (or ethnographies), try reframing the conversation with stakeholders to focus on the quantity of data you will collect – that is, how many hours of videos, how many photos, how long you immersed yourself and your team within the ecologies. These are more important data points that can reveal the robustness of your research beyond your n, and in turn help break your audience away from the focus on individuals.
3. Focus on what people do together, to make the invisible visible
Researching the needs, wants, and desires of individuals is nice because they are often so eminently visible. All it takes is a salient quote, or a particularly touching video clip to get your audience on board. The social forces that surround us meanwhile – phenomena like group dynamics, social norms, network effects, power structures, flows of influence and information, moods – tend to be nearly invisible. That makes them difficult to identify and even more tricky to convincingly convey to others when approached solely from an ethnographic angle; because the social forces around us tend to be visible through what we do, rather than what we say. And when it comes to making sense of what people do, ethnographic research can only take us so far and will generally benefit from working in tandem with experimental and quantitative methods. We must take advantage wherever possible the tools of quantitative analysis that embrace the social (for example, experiments, social listening, network analysis, etc.). So consider this yet another argument for embracing a mixed methods approach and for researchers to be as fluent in quantitative reasoning as they are in qualitative inquiry.
Over the last few decades, most business leaders have learned to listen to their customers. Now it is time we teach them to see their communities; to make sense of our social worlds, not just our minds. It won’t happen automatically, but it also isn’t impossible. Even the most hardened CEO will remember from their schooling that reality is socially constructed, but most will default back to an individualist point of view when push comes to shove. It’s on us as a community of ethnographic researchers to build a language that is understandable but more precise to what it is we practice and are observing – to reframe the conversation and provide a viable, practical alternative. We need to reestablish a set of best practices to avoid shortcuts or catering to the narrow or hypothesis-driven demands of clients and stakeholders.
In other words, we need to put the ‘social’ back into social science.
As leader of ReD’s technology, communications, and media practice in North America, Mikkel has advised lead executive teams at the world’s largest social media companies, telecom providers and electronics manufacturers. He also has a variety of clients across the finance, energy, and industrials sectors on product development, sales/marketing, M&A and organizational change. As head of ReD’s emerging practice in integrated social science and data science, he brings expertise in applying both qualitative and quantitative research methods to uncover patterns in our social fabric and drive adoption of new ideas, technologies, and offerings. His work has been published in Wired, Foreign Affairs, and VentureBeat. Mikkel has a degree in Sociology and International Studies from Yale University and is a former national team rower.
ReD Associates is an EPIC2021 Sponsor. Sponsor support enables our unique annual conference program that is curated by independent committees and invites diverse, critical perspectives.
Image: "Silhouette" by Thomas Hawk via flickr https://flic.kr/p/ew5hJ (CC BY-NC 2.0)
Applying Theory to Applied Ethnography, Maria Cury & Daniel Bird
Contextual Analytics: Towards a Practical Integration of Human and Data Science Approaches in the Development of Algorithms, Millie P. Arora, Mikkel Krenchel, Jacob McAuliffe, Poornima Ramaswamy
The Adaptation of Everyday Work in an Age of Automation, Tamara Moellenberg, Morgan Ramsey-Elliot, Claire Straty