Advancing the Value of Ethnography

Growing Communities: How Social Platforms Can Help Community Groups Achieve the Right Scale at the Right Time


Download PDF

Cite this article:

2020 EPIC Proceedings pp 87–97, ISSN 1559-8918,

Supporting communities on its platforms has been a part of Facebook’s core mission since 2017. Early understandings of the needs of groups and organizers largely centered around groups that began on Facebook itself. This paper is the result of ethnographic research conducted in 2019 to better understand the needs of different types of groups and the corresponding ways that technology platforms do and could support them. The initial orientation towards online groups led to the recognition of the difficulty of managing fast-growing groups but failed to consider whether groups might want to avoid growth in members altogether. We found in our research that many groups in fact did want to avoid or limit their growth in numbers. For these groups, growing as a community meant different things: offering more to existing members, raising awareness, or promoting the group to an outside audience, or simply maintaining over time. Our research was able to connect the dots of why organizers would have different aims between different groups or at different points in time. We ultimately presented our findings in a simple framework of three ‘toolkits’ that technology platforms can provide to meet the different needs of groups and organizers.


In 2017, Mark Zuckerberg, Facebook CEO, announced the company would have a focus on building the “social infrastructure for community,” or giving people the tools they need to build communities, a major shift from the original focus on connecting family and friends (Zuckerberg 2017). At the keynote for Facebook’s F8 Developer Conference in 2019, Zuckerberg noted that Facebook is “making communities as central as friends” (Bloomberg 2019). Using the lens of community, the company has released new products and design improvements such as putting Groups, Facebook’s product for people to engage with others who share their interests, front and center on the Facebook app experience. As part of this stream of work a new tab made it easier to access Groups and features were introduced to meet the needs of specific types of groups (Facebook 2019).

The work of weaving community into the Facebook app is iterative and ongoing. As such, the Facebook research and product team continues studying how people experience and build community and designing and building products that better support these behaviors. Part of the implicit challenge of this kind of large, long-term initiative is that a large product team needs to be brought along in the journey of learning about how people create and interact with their communities. Therefore, the research needs to shed light on new problem areas as well as helping the product team stakeholders internalize foundational insights that underpin multiple features. The study reported here was meant to further progress on this overall initiative by looking at specific questions, with the findings incorporated into the general narrative understanding about community.


Academic research on community groups and organizers has primarily focused on community organizing as a response to a problem (Escandón 2010, Mundell et al. 2015) or as a theoretical process (‘Community Development’) that needs to be figured out and perfected by those seeking a particular outcome (Schwartz 1981). In this body of research, ‘communities’ are often taken for granted as homogenous pre-existing units, defined at their limit of geography or demographics.

Studies of online groups have often focused exclusively on virtual communities (Rheingold 2000, Wilson and Peterson 2002), while some theorists of community have sought to deny the possibility of virtual communities at all (Calhoun 1991). In our research we were particularly interested in how groups span the offline and online world and how this balance shifted depending on the group’s primary orientation.

We were interested in groups that were not just responding to a problem or pre-defined by a certain geography or social unit. We wanted to explore as well the communities that are formed, joined and left in a purely voluntary manner. The community groups we encountered more closely resembled the grassroots campaigns characterized by Stokes Jones in an early EPIC paper (2005, 46) as “emergent in nature; as rooted in experiential being together; and as human projects driven by affect and effervescence as much as efficiency and purposiveness.” We saw groups that were emergent projects: overlapping and crisscrossing, coming into or dropping out of existence, of decidedly uncertain long-term viability.

Goodsell and Williamson (2008) provide a wonderful case study of a hybrid community group, a group that would have fit perfectly into our study. For online communities rooted in “geographic-place-as-practice” (253) they provide the following persuasive list of what members need to sustain community: information and explanations, “hot” topics, humor, maintenance of control over interactions, mutual encouragement and connection of online and offline worlds to facilitate interaction in both (260-1). These needs resonate with our findings and their in-depth explanation of a single group focused on urban rejuvenation is enlightening. They do not, however, explore how a group’s needs might change over time or how one group’s needs might differ from another, particularly the differences between groups that are more offline or online oriented.


The study’s goal described in this paper was to understand how people use multiple tools and techniques to create and support community. Our prior research suggested that communities cut across technology platforms. Although some communities may center on a particular digital platform like This Cat is CHONKY, a private Facebook Group for fanciers of plump cats (Kooser 2019), many other communities exist in a number of ways such as a college alumni association that can be supported by email lists, association magazines with updates about classmates, and in-person gatherings.

The study’s other goal was to understand more how community organizers managed their communities across platforms and how community members engaged across platforms. We wanted to understand how they thought of their communities and did their work with a portfolio of tools by both organizers and members. We were curious about how tools were chosen and also how digital tools supplement in-person meetings. The other area of investigation was whether the tools are a permanent part of the community’s communication repertoire or if they might be gradually added, much like the communication repertoire for a person’s social communication (Licoppe 2004). The product team would use this information to understand people’s underlying needs for building and engaging with community and brainstorm ways to improve its portfolio of products to serve unmet needs. We had previously highlighted Facebook Groups as a place for community, and there are other features of the Facebook app like Facebook Events and Pages for businesses and organizations that could be employed.


In light of the research questions around communities that spanned different platforms and wanting to understand both community members and organizers, we designed the study to reach a broad range of group types. Our goal was to get a variety of situations for diverse insights.

While community can be interpreted in many ways, we had a working definition: a collection of people, from which members receive a sense of belonging, connection, and feeling of safety, and to which they give trust and investment over time.

A feeling of safety means an environment that feels secure, where members understand the norms/culture/rules and how they should behave. Members can therefore do not need to worry about inadvertently breaking the rules and they can reveal more of themselves without fearing negative feedback.

Following this definition, whether a group is really a community is subjective and depends on the relationship between the individual and the group. One group member may feel strongly that it is a community for them and another may feel it is not.

Population & Diversity of Group Types

We knew from the outset that we wanted to include a diverse range of groups. We hypothesized that three key variables would carry a considerable impact on our subject matter: the purpose of the group, the role of technology for the group and, temporality.

  • The purpose of the group: hobby and interest-based groups, cause-based groups, values-based groups, experience-based groups
  • The role of technology: offline-oriented vs. online-oriented vs. hybrid groups
  • Temporality: maturity of group in its lifecycle (early-stage vs. well-established); periodicity (seasonal impact and frequency of interactions and activities)

We chose a diverse set of groups so as to uncover different and perhaps unexpected needs and opportunities.

  • Religious community groups
  • Neighborhood groups (city neighborhood, village groups)
  • Local political groups or issue-based groups
  • Hobby groups (e.g. cyclists, foodies, photographers, pet lovers)
  • Women’s groups, men’s groups, parenting groups
  • Immigrant groups
  • Charity/fundraising groups

We believed these types of groups would be likely to result in rich discussions of their needs and experiences. We anticipated that they would partake in a range of kinds of interactions (e.g. in-person and online) with different membership structures (e.g. flat vs. vertical organizational structure, single group vs subgroups) and models of participation (e.g. core and peripheral members).


We chose a multi-method approach for this project as we needed to get the most out of a short five days we had together in Madrid. We ran Mobile Diaries before the in-person fieldwork to get a sense of potential participants’ relationships to their groups and ensure we got the right mix of group types. We conducted in-depth interviews to capture detailed accounts of particular groups and the needs and descriptions of the tools used by each group. The team partook in observational immersions of community spaces and group meetings to get a feel for the cultural context and the emotional energy that these spaces and activities engender. Lastly we conducted brief unscheduled intercept interviews with people we approached in public spaces in order to get signal on an even wider array of groups and let us compare our specific sample with random representatives of the ‘general population’ so we could see if our respondents seemed to be outliers in their level of group engagement. Intercepts and immersions also gave us optionality, allowing us to take advantage of spare blocks of time as we felt best suited us in the moment.


The Role of Offline

Because of the history and culture of Facebook, the nature of online groups and the examples of online communities were always prominent in conversations. It was essential that we learn more about offline-oriented groups to understand how they differed from online-oriented groups and the different role that technology might play for them.

In our interviews, we asked participants to map the groups that were most important to them. Time and time again we saw the same groups closest to the stick figures representing our respondents: la familia, amigos/as de toda la vida, amigos/as del pueblo: family, lifelong friends, friends from my hometown. Many residents of Madrid are originally from small towns, and we often heard about the importance of pueblos, with town associations and annual parties, as a place to return to and be around one’s family and lifelong friends. When we asked about tools and platforms, these groups always ‘lived’ digitally in one place or another, most often as groups on WhatsApp. But while tools kept members of the group in touch, the groups existed in their ‘pure’ form outside and beyond any technology platform.

Immersions to community spaces showed us key differences between online and offline spaces for groups. We visited a cultural center run by local community groups and volunteers. The former tobacco factory had been transformed into spaces for lessons, rehearsals, and performances of dance, and music, a community garden, craft workshops and studios for sculpture and painting. A large calendar listed out the upcoming events and their locations while a bulletin board overflowed with flyers for events, classes, services and resources.

In online spaces, groups typically stand on their own as discrete and disconnected units. The cultural center in contrast was defined by a space of overlapping and intermingling groups. Regulars could drift from one activity, group or space to another. Non-members could come without a clear goal in mind to peruse the activities, observe groups in action and mingle with the other humans serendipitously sharing the space.

As we learned about groups which were oriented around offline activities, we were surprised to see that their interactions on digital platforms were often devoid of purely social light-hearted or humorous communication. There was little interaction that demonstrated intimacy and trust between members of the group. We realized that they had much less need for these types of interactions when online because their in-person activities provided the sufficient and appropriate time to bond and enjoy each other’s presence. In contrast, the conversation on platforms was nearly all sorting logistics and sharing necessary information to achieve their in-person goals.

The Role of Growth

As discussed above, teams at Facebook were very familiar with online groups, which could quickly grow from zero to hundreds of thousands of members. We’d seen how this could bring challenges for admins in moderating and managing such groups. We knew that not all organizers were trying to grow their groups all the time, but we didn’t have a clear understanding of why or why not. Nor had we thought of growth as a potential negative, to be actively avoided.

As we gained a deeper understanding of the diversity of groups that our participants were a part of, many revolving around offline interaction, we learned that many were quite content with the number of members they had and sought no growth or only limited growth in numbers. For some groups, an in-person activity at the center of the group limited from a logistical perspective the number of participants. This was the case with groups that came together to play sports. For other groups, the pool of potential participants was limited by characteristics required for the relevance of the group to a given individual, for example a group for immigrants from one country living in a particular part of Madrid. In a third scenario, some groups put tight limits on growth to ensure the quality and safety of their members. This was the case with a volunteering organization that had an extensive vetting process, requiring interested candidates to apply, attend an in-person meeting and run a trial of the service they intended on offering.

It was not only offline-oriented groups that were content without growth. We encountered online-based groups for sharing common interests, whether in collectible figurines or BMW 7-series, where organizers and members felt that no need for new members given that their groups regularly had members posting content and communication put out to the group resulted in engagement and feedback. Adding new members at this point would simply increase the likelihood of irrelevant or inappropriate content.

There was a handful of groups that were actively engaged in increasing the number of members. These were all groups that were centered around a particular cause or agenda. One group wanted to promote a particular philosophy within medical practice. Another was comprised of parents organizing to promote education for children with autism while another sought to improve the working conditions of prison guards. These groups were all hybrid groups, with significant activities both offline and online. The former was often used for planning and decision making, while the latter was essential for reaching new audiences.

In general, provided there were enough members to make the core activity possible and to keep the group active and lively, online or offline, these groups did not seek growth for growth’s sake. It was only with groups where the core purpose of the group depended on increasing the number of members or reach of the group that gaining new members and raising awareness was a top priority.

Growing beyond Numbers

Sheryl A. Kunjawa-Holbrook (2017, 203) writes of groups partaking in interreligious learning that:

Some congregations experienced growth beyond numbers-profound spiritual growth, growth in the knowledge of their own tradition, growth in community involvement, growth in hospitality, and growth in relationships between members due to interreligious partnerships.

We saw many organizers striving after a similar ‘growth beyond numbers’. Groups like a soccer team and a collection of dancers were limited in the number of participants, but as they participated in exhibitive activities they did seek to promote a community of fans and drive attendance at performances.

Others sought to grow the richness of the experience, finding new sources of common ground with other members. One woman combined her love of a photography app with her hobby of collecting dolls by creating a subgroup of fellow members who shared both passions.

We met organizers who wanted to grow their groups in stature more than in numbers. Andrea had started a Facebook group to donate items to people in need in his neighborhood. At the time of research, the group had nearly 1,500 members, and Andrea had 13 volunteers working with him yet his primary goal was to obtain official status as a non-profit, a permanent physical location, and some full-time staff members.


A Framework for Technology Supporting Community Groups

After our research, we faced a large number of stakeholders across products and roles keen to learn about what community groups needed and how technology platforms could support them.

The findings had to be actionable – the ultimate objective was always creating a platform that better supported community groups. They had to be communicated in a way that was clear and consistent: with a large audience we couldn’t risk conflicting interpretations that could arise from an overly complicated analysis.

We also needed to present our findings in such a way that they would instantly gel with intuitive common sense and personal experience, for two reasons. Firstly, nearly everyone has some experience being a part of a community group. With such a topic a listener will naturally compare the findings of the research with her personal experience. If the two clash, the findings are likely to be regarded with uncertainty.

Second, was the scale of the internal initiative and the broad group of stakeholders. Unlike a product team working on a sprint, the initiative to support communities is a long-term strategy and mission across many different teams and products at Facebook. For our findings to have an impact, we needed the learnings to embed with many stakeholders such that it would stick with them over time. Our findings needed to be easy to understand without too much effort and they needed to be not just believable but deeply believed by our audience.

Although we needed actionable insights, the level of specific features and UI was going to be too granular. We had to think about what it was that organizers and members were trying to achieve. Communication was at the heart of what platforms provided to groups, but to what end?

Ultimately we formulated a framework of three broadly scoped ‘toolkits’ through which a technological platform could meet the different needs of community groups:

  • Managing logistics
  • Promoting the group or a cause
  • Facilitating discussion

Managing logistics: Who is doing what? What are the key dates and deadlines? Who is bringing what? Who has paid and who hasn’t? Who has an extra seat in their car?

Promoting the group or a cause: Adding new members, fundraising or collecting donations, raising awareness

Facilitating discussion: Discussion that is active, a responsive community where members answer each other’s questions, give advice and make suggestions, sharing ideas and inspiration, enjoying relationships and making new connections when not together in-person. Not too quiet and not too noisy.

All groups and organizers at certain points will have a need for all three toolkits. However, a group often has a much stronger need for one toolkit over the others, and further, the relative importance of each can change considerably depending on the stage of the group’s development.

A Platform that Evolves with the Group

Concrete understanding of these groups helped the product team refine its definition of community, with a greater focus on the purposes and blurred boundaries of communities. Influenced by this study, the team shifted from thinking about individual features to a portfolio approach to the features for community organizers, recognizing the need for toolkits of features that serve community organizers’ various goals and needs. The growth of the group is only one need served among many others. Further, the team began considering the development stage of a community as a factor that influences priorities. Most groups tend to start small. We saw how groups evolved and how leaders evolved with their groups. We understood that toolkits provided must be similarly flexible and able to evolve with a given group.

The existing body of knowledge within Facebook was enriched by having a much strong appreciation of the notion of life stages in a community and in each person’s relationship or journey with the community – and thus differing needs for toolkits over time. The insights from the study helped frame the team’s community strategy, which includes yet to be released products that better support the different life stages of where a person might be in their journey to community, for example helping users discover communities of potential interest at the beginning of a journey or further along helping them engage more deeply in communities they already belong to.

From a researcher’s perspective, the study confirmed a hunch that we needed to think more holistically about the goals of community leaders. Although efficiency and growth are important at different times, considering their deeper motivations and aspirations for the community can help us provide a richer set of tools. With the themes of the toolkits, we have been shifting the narrative about leaders to one of longer-term goals and not just short-term efficiency. Likewise, with the more nuanced understanding around the desire for growth or lack of, we can consider other tools for preserving internal culture and nurturing.

From a storytelling perspective, we were able to tell more stories about the goals and vision of community organizers. More tactical research would highlight a narrow problem that Groups admins might experience (such as working through pending membership requests) and suggest possible fixes. This ethnographic fieldwork enabled us to discuss the bigger picture of goals and vision – what organizers were truly concerned about (often preserving culture rather than growing the group). We were then able to place the tactical problems within the context of broader goals to explain why they were problematic in the first place (for example sifting through membership requests from people who are not the target audience of the group doesn’t help achieve the ultimate goal of preserving group culture).

Equal Appreciation of Online and Offline

Previous research was heavily grounded on understanding around communities based on the Facebook platform. Our insights added much-needed clarity into group dynamics, functioning and priorities in more offline-oriented or hybrid offline-online contexts, areas that as discussed were less well-known to the team. We were able to explain to stakeholders why an organizer might use a mix of tools, whether a multi-purpose communication tool like WhatsApp or something perceived to be ideal for certain situations such as sharing photos on Instagram.

The sorting of offline logistics in particular showed that a good deal of the interactions that take place on digital platforms are about meeting basic needs and ensuring awareness between members, whether coordinating a potluck, organizing a gift exchange or finding a carpool to this week’s soccer practice. Fewer interactions are actually about the shared interest or purpose of the group. This insight set the stage for a design sprint exploring how we could better serve the unique needs of community leaders and inspired further exploration of how teams of community leaders coordinate and communicate amongst themselves. Ultimately the sprint gave specific product teams some action items to develop features that can help leaders be more proactive in meeting their vision for their community.

The vision of supporting communities underpins much of the product and touches many different teams and features. We generated knowledge with this research which was adopted and internalized by a huge team, ultimately making the research a success.


‘Obvious’ Findings and the Scaling of Findings to Large Numbers of Internal Stakeholders

The organizational psychologist Adam Grant has written about the value of ‘obvious’ insights (2019). Findings that resonate with common sense and intuitively ‘sound right’ to the audience can, Grant argues, be particularly effective in overcoming three barriers to change: resistance to new data, resistance to change and organizational uniqueness bias.

Obvious insights can motivate us to close the knowing-doing gap. Common sense is rarely common practice. If you ask managers what effectiveness looks like, they often can spell out the critical factors. The key is to get them to act on that insight, and that’s where the obvious can help.

That community group’s goals and needs change over time and depending on the group’s purpose is not particularly surprising or particularly profound. In our case, we believe it was an advantage rather than a weakness to present findings in a simple framework that was compatible with common sense experiences and well-established understanding of what community groups were like, rather than challenging conventional wisdom or introducing unfamiliar concepts.

As researchers, we often feel the desire or need for our findings to unveil a new paradigm, which overturns the previous understanding of the subject matter, shatters preconceived notions, and busts common-sense myths. In this work, we had to resist such temptations or risk the findings being taken opposing directions or falling by the wayside entirely.

In presenting a straightforward framework that was easy to grasp intuitively, we sought to provide a solid foundation of shared understanding that would serve as the basis for decisions and execution over the long-term across a multitude of different teams and products. It is our belief that this approach is likely the more effective one with research projects seeking to drive long-term impact with many diverse stakeholders, particularly when the initiative is an ongoing one touching on multiple parts of the product (as opposed to, for example, a brand new product being launched).


The goal of this research was to understand the needs of community groups and group organizers and the role of tech platforms to inform the organization’s strategy of supporting community groups, ultimately seeking to embed insights with diverse roles on a multitude of different product teams. We learned about a range of group types, in particular gaining knowledge about offline-oriented and hybrid online-offline groups.

In our research we saw that growing the number of members was only a priority in a select number of groups. Others sought to limit or even actively avoid too much growth. They sought instead to grow by increasing the richness of the experience of their members by increasing the stature of the group or by increasing engagement with an outside audience.

We identified the three main ‘toolkits’ that community groups need from technology. Most instructive for our stakeholders was understanding how these needs varied depending on the purpose of the group and on the stage in development of the group.

Communicating our findings in a simple and straightforward framework that fit intuitively with stakeholders’ common sense and avoided abstract or unfamiliar concepts was key to the impact of the research. This framework along, with concrete examples from case studies, allowed us to embed a large number of stakeholders on a large number of teams with a shared foundational understanding of what community groups need from technology that would enable them to make decisions about product and feature development as this initiative continues over the years.


Bloomberg. 2019. “Facebook leans harder into Groups as it redesigns main app.” LA Times website, April 30. Accessed September 22, 2020.

Calhoun, Craig. 1991. “Indirect Relationships and Imagined Communities: Large-Scale Social Integration and the Transformation of Everyday Life.” In Social Theory for a Changing Society, edited by P. Bourdieu and J. S. Colemen, 95-121. Boulder: Westview Press.

Escandón, Socorro. 2010. “Theoretical Versus Grass-Roots Development of a Community Partnership.” The Qualitative Report 15 (1): 142–155.

Facebook. 2019. “Day 1 of F8 2019: Building New Products and Features for a Privacy-Focused Social Platform.”, April 30. Accessed September 22, 2020.

Goodsell, Todd L. and Owen Williamson. 2008. “The Case of the Brick Huggers: The Practice of an Online Community.” City & Community 7 (3): 251-71.

Grant, Adam. 2019. “The Surprising Value of Obvious Insights.” MIT Sloan Management Review, February 5. Accessed September 22, 2020.

Jones, Stokes. 2005. “Grass Roots Campaigning as Elective Sociality (or Maffesoli Meets ‘Social Software’): Lessons from the BBC iCan Project.” Ethnographic Praxis in Industry Conference Proceedings: pp. 31–52.

Kooser, Amanda. 2019. “This Cat is Chonky: The fat-cat online shrine lifting humans from despair.”, June 14. Accessed September 22, 2020.

Kujawa-Holbrook, Sheryl A. 2017. “‘The Sacrament of Human Life’ Cultivating Intentional Interreligious Learning in Congregations.” In Teaching for a Multifaith World, edited by E. S. Fernandez, 199-217. Eugene: Pickwick Publications.

Licoppe, Christian. 2004. “‘Connected’ Presence: The Emergence of a New Repertoire for Managing Social Relationships in a Changing Communication Technoscape.” Environment and Planning D: Society and Space 22 (1): 135-56.

Mundell, Leah, Lisa J. Hardy, Roxana Cardiel De Niz, and Michelle Thomas. 2015. “Defying Displacement: Organizing for a Beautiful Life in Flagstaff, Arizona.” Practicing Anthropology 37 (2): 35–39.

Rheingold, Howard. 2000. The virtual community: homesteading on the electronic frontier. Cambridge MA: MIT Press.

Schwartz, Norman B. 1981. “Anthropological Views of Community and Community Development.” Human Organization 40 (4): 313–322.

Wilson, Samuel M. and Leighton C. Peterson. 2002. “The Anthropology of Online Communities.” Annual Review of Anthropology, Vol. 31, 449-467.

Zuckerberg, Mark. 2017. “Building Global Community.”, February 16. Accessed September 22, 2020.