When I first started working at Facebook as a UX researcher, I tried to hide my background as an anthropologist. I felt like an odd duck, the one who, among a team of mostly psychologists, had the “other” type of academic background. To blend in with the rest of the research team, I distanced myself from my theoretical background, avoiding terms like structural inequality, embodiment, and objectivity and trying to forget about Foucault and Bourdieu.

But over time, I began to see how anthropology—and in particular, its focus on power dynamics and structural inequality—could help me come up with more creative and enduring ways to make technology more equitable. I saw how limiting it was to think that better technological futures would emerge through in-product changes alone. I began to think more broadly about how to foster inclusion through organizational change, by gradually shifting the ways the company approached research and product design.

For the past few years, my research at Facebook has focused on how digital technologies can have uneven impacts on different populations. For one person, the ability to sell goods on Facebook Marketplace might be the key to economic vitality, while for another person, it could be a threat to one's safety. Peoples’ abilities to derive value from technology can be dramatically impacted by their digital literacy: the ability to understand, navigate, and participate online in a safe manner.

Addressing digital literacy in product design, however, is no simple task. On the one hand, teams can try to redesign and simplify features to make it easier for people of all abilities to use their products. But that is not always enough. For some segments of the population, a lack of digital skills may always be a barrier to usage, no matter how well designed a digital product is. My anthropological training kicked in, as I saw first-hand how the “user’s” abilities with technology transcended their singular relationship with their phone and was influenced by their life history and environment (as Shaheen Amirebrahimi points out in his 2016 EPIC paper, “The Rise of the User and the Fall of People: Ethnographic Cooptation and a New Language of Globalization”). Through my work on digital literacy, I began to understand how product development would need to address not just peoples’ problems with technology, but also the power structures that prevented people from engaging fully with technology in the first place.

After this realization, I advocated fiercely for the company to invest in education: in-product solutions that would help people learn to use and understand technology. Education, I thought, would help the company build products that took the power structures and social context into account. Tutorials, help pages, digital assistants—they would be key to helping those with lower digital literacy use Facebook’s products fully and safely.

To test this hypothesis, I set out to conduct research on how Facebook could better support “learning”—about new Facebook features, new skills required to fully take advantage of Facebook—among people with low digital literacy. At the beginning of the project, we had selected Florida as a location to “find” people with lower digital literacy in the United States. As I carried out in-depth interviews in impoverished parts of Miami-Dade County, I saw how, in this population, digital literacy was deeply intertwined with structural inequality: systemic poverty and violence, barriers to educational attainment, and racism. But to my surprise—as naive as it seems today—the same structural inequality that prevented people from engaging with technology also prevented them from understanding how to learn about technology. Learning itself was a skill that had to be nurtured and developed.

At this point, I wasn’t sure what to do. For disadvantaged communities who were most in need of help using digital services, perhaps in-product education would never work. The people who I had spoken to lacked the confidence to explore new features, and had never learned how to Google for virtual help. They struggled to formulate the correct search queries, to navigate to documents and services that provided help, or to follow along with jargon-heavy tutorials.

This research, on both digital literacy and learning, fundamentally altered how I thought about my job. In all likelihood, people with low digital literacy would need not just simpler user interfaces, but also access to broader forms of help—tech support, communal forums, programs at libraries, and so on—that were not seen as being in scope for in-product educational features. If Facebook wasn’t planning to invest in resource-intensive help solutions, what could we do to make Facebook’s products more equitable and accessible?

Ultimately, I became convinced that addressing the inequalities inherent in technology usage would require thinking beyond in-product changes and focusing instead on organizational change. I set out to change the process of research and product design, providing researchers with new tools to recruit across the range of the digital skills spectrum, providing designers with guidelines to adhere to, and providing data scientists with metrics to evaluate how product usage varied by skill level. This was not just about making products more simple or making education more inclusive. It was about (slowly!) shifting which user groups the organization thought it needed to build products for, as well as helping the organization prioritize developing solutions for vulnerable communities. I knew that organizational change was not enough, that it was not the same as investing in local libraries or education systems. But it was a start in the right direction.

Although this research was conducted before the COVID-19 pandemic, these learnings have stuck with me, especially as I have seen an increasing number of people turn to online services, ranging from social media to government to education. What happens when people must rely on technology for their health and livelihood but have no access to in-person help or dedicated trainings? What happens when the online medium of learning prevents children and young adults from learning how to learn? I don’t have the answers, but this work has given me a renewed appreciation for how we must always account for power structures as we design technological solutions and how technology will impact the most vulnerable.