“What You Post in the Group Stays in the Group”: Examining the Affordances of Bounded Social Media Places

Abstract

This study focuses on the affordances of bounded social media places (BSMPs), low visibility places within social media platforms like private messaging and private groups. While researchers have focused on BSMPs within specific platforms, this study presents a systematic examination of BSMPs across multiple platforms to facilitate theoretical durability. Interviews with users of BSMPs across diverse platforms (N = 35) reveal that BSMPs discourage the affordance of visibility as they are considered private due to visibility management mechanisms and trust in known audiences. They encourage personalization as users believe they receive relevant content from and can send relevant content to specific audiences, in the absence of algorithms. BSMPs also encourage synchronicity by facilitating continuous conversations. The strength of encouragement or discouragement of these affordances varies across different BSMPs and is informed by users’ social positions. This study therefore contributes a framework and shared terminology for future research on BSMPs across social media.

Recent media coverage and public discussions have emphasized how our understanding and use of social media is changing, with a focus on factors like growing privacy concerns among users, changes in ownership and policies for platforms like Twitter (now X), and the COVID-19 pandemic (Molla, 2021; Roose, 2022). One major change is the growing shift toward the use of more private-feeling and community-focused spaces on social media such as private chats on mobile messaging services like WhatsApp, private Facebook groups, and private servers on services like Discord (Sinders, 2022). These are bounded social media places (BSMPs), defined as places within social media platforms and services characterized by low levels of visibility to broad and unintended audiences (Malhotra, 2023).

Given the increasing use of these BSMPs, it is important to empirically and systematically examine how they are perceived as used. Social media research has often focused on how people present themselves to a broad imagined audience in high visibility spaces (Litt, 2012; Marwick & boyd, 2011), with an emphasis on platforms like Facebook and Twitter (Stoycheff et al., 2017). But our understanding and use of social media is evolving. As Baulch et al. (2020) note, it is vital to “consider how the increasing tendency to communicate in closed networks or groups is cause for reconceptualising how the public/private divide plays out online” (p. 2). Studying the use of BSMPs can help us understand this divide better. To be sure, BSMPs have received some scholarly attention in recent years, especially private chats on mobile messaging services like WhatsApp (e.g., Cruz & Harindranath, 2020; Johns et al., 2023) and specific channels like private Facebook groups (e.g., Pruchniewska, 2019). However, these studies often focus on one particular platform or channel. There is also a need for a more systematic examination of BSMPs across multiple platforms to ensure theoretical durability and provide a common language for future research. While individual studies on WhatsApp or private Facebook groups are certainly valuable, systematically understanding how people use and perceive BSMPs across multiple platforms can lay the theoretical groundwork for future research on low visibility places within social media, even as existing platforms evolve and new platforms emerge.

To address this issue, through interviews (N = 35) with users of diverse BSMPs, I examine the affordances associated with these places, wherein affordances are actions made possible by a technology within a specific context (Evans et al., 2017). I utilize an affordances perspective as it produces “more generalizable and lasting findings because the mechanisms identified in affordance-focused scholarship can be transferred to newer platforms and other user practices and contexts” (Vitak & Ellison, 2018, p. 488). Drawing on Davis’s (2020) mechanisms and conditions framework of affordances, I find that BSMPs discourage the affordance of visibility as content shared within these places is perceived as being visible to only intended audiences, encourage the affordance of personalization as users believe they receive relevant content from and can send relevant content to specific individuals or groups, and encourage the affordance of synchronicity as these places are associated with facilitating continuous conversations. Furthermore, I highlight how the strength of this encouragement or discouragement varies across different types of BSMPs, with private mobile messaging chats and groups discouraging visibility and encouraging personalization and synchronicity to an especially strong degree. How these affordances operate also varies across different users; particularly, people whose identity or life circumstances result in them being part of private groups about sensitive topics are required to put in greater effort to protect themselves from harm. In presenting these findings, I therefore establish a systematic framework for future research on BSMPs, based on empirical findings and existing scholarship on a variety of BSMPs.

Literature Review

Bounded Social Media Places

The concept of BSMPs brings together low visibility places within broader social media platforms under one wide umbrella (Malhotra, 2023). Through mechanisms like requiring administrators to approve group membership and having people answer screening questions, participation in BSMPs is restricted to specific audiences. Content posted within BSMPs is also usually inaccessible via public search and is typically presented in chronological order rather than being algorithmically curated. There are four main types of BSMPs—private chats within mobile messaging services like WhatsApp and Signal, private communities within channel-based messaging services like Discord and Slack, private messaging within social networking sites like Facebook and Instagram, and private groups within social networking sites like Facebook. These types are not mutually exclusive or exhaustive, but they reflect the variety of BSMPs. To systematically examine how they are perceived and used, I adopt an affordances approach.

Affordances

Affordances are actions made possible by an object and are understood through examining the dynamic relationship between technological objects and users within a particular context (Evans et al., 2017). Gibson’s (1979) conceptualization of affordances, situated in ecological psychology, emphasized the role of the materiality of objects in shaping affordances. Meanwhile, Norman’s (1988) conceptualization, situated within human-computer interaction (HCI) and design studies, foregrounded the role of user perceptions. More recent scholarship contends that affordances must be viewed as relational rather than object- or subject-centered, dynamic and variable rather than static and binary, and contextually- and culturally-situated rather than existing in a vacuum (Davis, 2020).

Mechanisms and Conditions Framework

Through introducing the mechanisms and conditions framework, Davis (2020) presents an analytical tool informed by this modern view of affordances. She suggests that rather than asking what an object affords, which often results in a binary conceptualization of affordances, scholars should focus on how and for whom and under what circumstances objects afford. Davis (2020) states that answering the how question requires focusing on the mechanisms through which affordances operate. The encourage, discourage, and refuse mechanisms refer to how technologies accommodate lines of action initiated by users. Technologies encourage by making a user’s initiative easy to accomplish, discourage by creating barriers, and refuse by making the initiative impossible to accomplish. Meanwhile, request and demand are mechanisms that capture bids initiated by technologies on users. Technologies request by making one action preferable over others and demand by making an action unavoidable. Finally, allow is a neutral mechanism wherein no one line of action is imposed or made more easy or difficult. It is also important to note that

bids by the object (request and demand) are not empirically distinct from bids on the object (encourage, discourage, and refuse). Rather, each serves as a set of analytic pegs that represent distinct foci on particular parts of the human-technology relation (Davis, 2020, p. 71).

As this interview study foregrounds the perspectives of users and how they engage with a range of BSMPs, its primary focus is on bids placed by users, namely how BSMPs may encourage, discourage, or refuse certain affordances.

Davis (2020) also notes that answering the for whom and under what circumstances question necessitates focusing on how these mechanisms are “entangled with social and structural conditions” (p. 89). This includes the extent to which users perceive a technology’s function, the dexterity or skill they possess to fully engage with a technology, and how the user-technology relationship is influenced by cultural and institutional legitimacy, which refers to the power users possess within broader social and institutional structures. In this study, I examine these conditions by focusing on how differences in identity and demographic markers, cultural backgrounds, and individual life circumstances across the interview sample shape how the affordances of BSMPs operate. Furthermore, Davis’s (2020) focus on contextual variation in how affordances operate provides the theoretical flexibility required to recognize that the same affordance may operate differently across different types of BSMPs. For instance, private WhatsApp chats are typically encrypted and often consist of a small audience of close friends and family, while private Facebook groups are typically comprised of a larger audience of people with common interests or goals. Thus, the former may discourage visibility to a greater degree than the latter.

Overall, Davis’s (2020) framework serves as an analytical tool that allows me to systematically examine how the common affordances associated with BSMPs operate, accounting for variation across different BSMPs, users, and contexts. In the next section, I identify these common affordances through drawing on existing literature on social media affordances and BSMPs.

Affordances and BSMPs

Research on social media and affordances has primarily examined the use of Facebook and Twitter to publicly interact with intended and unintended audiences (Ronzhyn et al., 2022), with scholars focusing on affordances like visibility and persistence (boyd, 2010; Treem & Leonardi, 2013). Some studies also generate new typologies of affordances for each emerging social media platform, which is an issue because “narrowly identifying or simply listing purported affordances associated with a technology provides limited theoretical value and confines the contribution of work to a restrictive empirical space” (Evans et al., 2017, p. 46). Furthermore, Evans et al. (2017) argue that this can lead to scholars mislabeling technological features or outcomes as affordances.

Affordances-based research on BSMPs is nascent but follows some of the same patterns as broader research on social media and affordances. For example, research by Masip et al. (2021) on private WhatsApp chats finds that these chats afford privacy, personalization, replicability, and segmentation. Meanwhile, Yeshua-Katz (2021) outlines how private Facebook groups and WhatsApp chats afford visibility, availability, multimediality, surveillance, and synchronicity. Lou et al. (2021) focus on news consumption on Telegram and state that Telegram affords customizability, accessibility, aesthetics, and simultaneity. These studies are immensely valuable as they apply the affordances perspective to examine specific BSMPs that are gaining popularity. However, constructing new typologies of affordances for each new BSMP that emerges limits systematic and durable theorizing. Furthermore, as noted above, it is important not to confuse affordances with a technology’s features or outcomes that result from a technology’s use (Evans et al., 2017). For instance, it can be argued that surveillance is an outcome of using social media and aesthetics are a design feature. Much of this work also focuses on what social media affords rather than highlighting how affordances operate differently for different people, based on how they are socially situated (Davis, 2020).

To address these issues, I first identify three main affordances associated with BSMPs. To do so, I draw on well-established typologies of social media affordances (boyd, 2010; Evans et al., 2017; Ronzhyn et al., 2022; Treem & Leonardi, 2013). I also review existing literature on specific BSMPs, including the limited amount of research that adopts an affordances framework, as well as scholarship that does not explicitly adopt an affordances framework but gestures at the actions made possible by BSMPs. Through reviewing this literature and applying Evans et al.’s (2017) threshold criteria of affordances being variable and distinct from technological features and outcomes, I identify three main affordances associated with BSMPs—visibility, personalization, and synchronicity. I outline these affordances below, followed by an empirical examination of how and for whom and under what circumstances they operate through interviews with users of a variety of BSMPs.

Visibility

Visibility refers to what can be seen or perceived by the sense of sight and is both situated and reciprocal (Thompson, 2005). Mediated channels remove the spatiotemporal constraints associated with visibility as content on them can be visible across time and distance. Pearce et al. (2018) argue that on social media, the affordance of visibility requires both a seer and an object and is therefore relational; it is also strategic, in that people can manipulate it to their own ends by drawing boundaries between what is seen and not seen; and its outcomes are indeterminate, as its consequences can be harmful or beneficial. Visibility is key to understanding the current social media environment, where “information is not public or private but located somewhere on a spectrum of visibility” (Marwick, 2023, p. 27). It is a root affordance that facilitates other affordances associated with social media (Flyverbom et al., 2016).

Control over visibility is a defining characteristic of BSMPs. Studies find that BSMPs facilitate strategic management of visibility as they allow people to restrict membership within a chat or group to specific audiences (Malinen, 2021; Swart et al., 2019). This results in BSMPs being perceived as safe spaces for marginalized communities (Archer et al., 2021; Pruchniewska, 2019). It also means that they are used to interact with close friends and family (Cui, 2016), share knowledge with those who have common interests (Pi et al., 2013), and exchange social support (Yeshua-Katz, 2021). However, we know less about the social and technological factors that shape perceptions of visibility across different types of BSMPs. Furthermore, the studies outlined above seldom explicitly apply an affordances framework and recognize visibility as an affordance, limiting broader theorizing on visibility and BSMPs. Thus, drawing on Davis’s (2020) mechanisms and conditions framework, I address the following research question:

• Research Question 1 (RQ1). How and for whom and under what circumstances does the affordance of visibility operate across different types of BSMPs?

Personalization

There are two primary conceptualizations of the affordance of personalization—one that is receiver-oriented and emphasizes the tailoring of content, and another that is sender-oriented and is associated with the directedness of content. Both are relevant to BSMPs.

Receiver-Oriented Personalization

While some conceptualizations of personalization focus on how algorithms tailor content for users, Kümpel’s (2021) idea of explicit personalization refers to “instances in which the social media users themselves take action to customize their information environment” (Kümpel, 2021, p. 4). This is analogous to what Sundar and Marathe (2010) label as customization. On BSMPs, explicit personalization occurs as users can customize their information environment through controlling who they interact with. For instance, users believe that they receive relevant and personalized content on mobile messaging services as their audience typically consists of close ties who know them well and/or people with shared interests (Lou et al., 2021; Masip et al., 2021). However, much of the scholarship on receiver-oriented personalization and BSMPs focuses on these mobile messaging services, with little attention paid to other types of BSMPs. Furthermore, how broader structural factors impact the ways in which different people experience personalization remains understudied.

Sender-Oriented Personalization

Fox and McEwan (2017) define personalization as “the ability to direct a message to a specific individual or group” (p. 303), similar to the concept of directedness, which refers to “whether a message is targeted at a particular person or not” (Liu & Kang, 2017, p. 70). BSMPs are associated with such sender-oriented personalization as they enable users to easily segment their audiences into different groups and one-to-one chats (Ling & Lai, 2016; Masip et al., 2021; Mols & Pridmore, 2021). Even as researchers have discussed how different types of BSMPs allow for such segmentation, they have not linked this to the affordance of personalization, limiting systematic and durable theorizing. Furthermore, we need to understand the specific social and technological factors that inform the perception that BSMPs are associated with sender-oriented personalization.

To address these issues, I focus on both receiver- and sender-oriented personalization and examine the following research question:

• Research Question 2 (RQ2). How and for whom and under what circumstances does the affordance of personalization operate across different types of BSMPs?

Synchronicity

The affordance of synchronicity is associated with immediate feedback and “the timing of message exchange within a given time frame” (Burgoon et al., 2010, p. 347). It is shaped by technological features as well as social norms and expectations. As Rettie (2009) states, “[T]he perception of a medium as synchronous reflects socially shaped expectations that the interactants will maintain continuous focused attention and will engage in coordinated interaction for a sustained period” (p. 427).

In terms of technological features, the presence feature, which indicates whether a user is online and available for a focused interaction (Nardi et al., 2000), contributes to private chats within mobile messaging services being perceived as near-synchronous (Ling & Lai, 2016). Less is known about perceptions of synchronicity on other BSMPs like private Facebook groups, where such features are typically absent. In terms of norms and expectations, we know that on mediated channels, people expect close ties to respond quicker than strangers (Dogruel & Schnauber-Stockmann, 2021). Thus, different BSMPs may be perceived as more or less synchronous, depending on the technological features, norms and expectations, and audiences associated with them. In this study, I examine how these factors vary across different types of BSMPs, addressing the following research question:

• Research Question 3 (RQ3). How and for whom and under what circumstances does the affordance of synchronicity operate across different types of BSMPs?

Method

Recruitment and Sampling

In-depth interviews with U.S.-based adults who use at least one BSMP regularly were conducted between July and October of 2022. After the university institutional review board (IRB) deemed the study exempt, participants were recruited through purposive and snowball sampling. I posted a recruitment flyer on social media, and interviewees were also asked to refer people who use different BSMPs. Efforts were made to recruit people who use different social media platforms and people who are of diverse ages, genders, professions, and cultural backgrounds. This is because my objective was to conduct a systematic examination of BSMPs across multiple platforms, including examining how the affordances of BSMPs operate differently for different users across varied contexts. One sampling strategy geared toward this objective was intentionally oversampling for people belonging to different immigrant and diasporic groups as they come from diverse cultural backgrounds and use BSMPs within different platforms (Gursky et al., 2022). For instance, South Asian and Latin American immigrant groups are typically associated with the use of WhatsApp, while the Chinese immigrant community typically uses WeChat. The final sample therefore consisted of 35 interview participants of different ages, genders, professions, and cultural backgrounds, including people of Brazilian, Indian, Chinese, Thai, Mexican, and Costa Rican backgrounds (see Appendix A in Supplemental Material for demographic details). This also meant that the sample covered the use of the following diverse BSMPs: Private chats within WhatsApp, WeChat, Telegram, Signal, LINE, Kakao Talk, GroupMe, and Viber; private messaging within Facebook, Twitter, and Instagram; private servers and workspaces within Discord and Slack; and private visible and private hidden groups within Facebook. All participants received a $15 e-gift card as compensation. When referring to individual participants, I use pseudonyms chosen by them.

Interviews

After completing an informed consent form and a form with questions about demographic details and how frequently they use different social media platforms, participants were interviewed via Zoom. Interviews were conducted in English and lasted for approximately 1 hour on average. As these interviews were part of a larger research project on social media and BSMPs, participants were asked about many different issues. In terms of the research questions this study focuses on, they were specifically asked about visibility, personalization, and synchronicity in relation to BSMPs. To allow for the possibility of additional emergent affordances being mentioned, they were also asked more general questions regarding how they perceive and use BSMPs (see Appendix B in Supplemental Material for full interview protocol). During interviews, participants were also encouraged to scroll through the BSMPs they use to help with recall (Robards & Lincoln, 2019).

Analysis and Verification

Interviews were recorded and transcribed. Transcripts were analyzed using the technique of theoretically-driven thematic analysis (Braun & Clarke, 2006). Analysis was guided by Davis’s (2020) definitions of the different mechanisms and conditions associated with affordances. I also drew on existing conceptualizations of visibility, personalization, and synchronicity. These definitions and conceptualizations are detailed in the literature review. Participants’ responses were analyzed to identify how visibility, personalization, and synchronicity operate on BSMPs, with a particular focus on the encourage, discourage, and refuse mechanisms. Thus, for each affordance, I identified if participants described BSMPs as making the affordance easy to execute (encourage), difficult to execute by creating barriers (discourage), or impossible to execute (refuse). Participants’ responses were also analyzed within the context of their identity and individual circumstances within existing social structures (cultural and institutional legitimacy), how they perceived BSMPs (perception), and the extent to which they had the skills to use BSMPs (dexterity). To confirm the trustworthiness of this analysis, I also engaged in data verification through peer debriefing and keeping an audit trail (Lincoln & Guba, 1985). Peer debriefing involved sharing a subset of the data and all the identified themes with an independent qualitatively trained researcher who evaluated the extent to which the themes accurately reflected the data. This was followed by a discussion wherein we came to a consensus regarding the final themes. Maintaining the audit trail involved taking detailed notes throughout the analysis process.

Findings

Across the diverse sample, BSMPs within multiple social media platforms were discussed. On being asked about what generally motivates them to use BSMPs, participants highlighted how these places are suitable for maintaining relationships with family and friends and/or joining groups and communities organized around shared interests. These include groups centered on sensitive issues that impact marginalized populations such as infertility issues, mental health–related topics, and LGBTQIA+ identity. As is outlined below, participants also reflected on the main affordances associated with BSMPs, including the social and technological factors that inform how these affordances operate across different types of BSMPs and different contexts.

RQ1: Visibility

Participants’ accounts revealed that BSMPs discourage visibility. The discourage mechanism refers to how an object makes a certain line of action more difficult to carry out (Davis, 2020). According to participants, BSMPs make it difficult for content to be visible to unintended and broad audiences as these places are private, difficult to access and search, and associated with known and intended audiences. For example, Bette said that BSMPs like private Facebook groups feel “more private,” are associated with “an intentional audience,” and “what is posted on there is going to stay in there.” Similarly, Cruz emphasized how private chats on WhatsApp facilitate control over accessibility and allow them to communicate with a known audience: “It’s a more controlled environment, it’s more private. I can see exactly who is in the group.” Here, it is important to note that in discussing visibility, many participants used the term private. Colloquially, people often refer to a lack of visibility as privacy; however, within affordances research, privacy is considered an outcome of a technology affording low visibility (Evans et al., 2017). As this study focuses on visibility as an affordance, I view privacy as a descriptor for low visibility rather than engaging with it as a distinct concept.

Content posted within most BSMPs can, of course, be captured through a screenshot and shared with wider and unintended audiences (Jaynes, 2020). However, Xavier highlighted how BSMPs discourage visibility by making this line of action difficult and time-consuming, rather than refusing visibility by making it impossible:

If you are going to take a screenshot from a more private group conversation or an individual direct messaging conversation, then you have to scroll to what you want to share, probably take multiple screenshots, and then decide how you’re going to share them out. Are you going to message them to somebody else? Are you going to post them? So, it’s a more tedious process.

As is outlined below, the level of trust users have in their audiences on BSMPs also informs the expectation that people will not take a screenshot of content posted within BSMPs and share it with a wider audience.

How visibility operates on BSMPs is also influenced by users’ social positions and broader conditions (Davis, 2020). These conditions include the extent to which people have the skills or dexterity to use technology. Some participants highlighted how controlling one’s visibility online used to require high levels of internet skills, but BSMPs have made this practice accessible to more people. For instance, Ben stated that low visibility places like Internet Relay chats were mainly accessible to technologically savvy people like him, but now people “can’t use technical ignorance as an excuse” since many BSMPs are accessible to people with limited technological skills. Allen, a 76-year-old man who described himself as having limited technological skills, demonstrated this while comparing WhatsApp to email: “I can still do the basic WhatsApp, send a message, receive a message. Whereas most email formats have gotten so complicated, I’m discouraged by them.”

Furthermore, how people are embedded within social structures, or their cultural and institutional legitimacy (Davis, 2020), also shapes how they experience visibility on BSMPs. For those who use BSMPs within platforms with strong ties to authoritarian governments known for digital surveillance, there was a sense that they could not assume that what happens in the BSMP stays in the BSMP. For example, Elise described her awareness of Chinese government surveillance and how that impacts the extent to which she views private chats within WeChat as discouraging visibility: “There’s the underlying audience, as in, the Chinese government developers. Honestly, I’ve always been aware of that since I’ve used WeChat. I know that everything I say is going to be recorded.” Meanwhile, Alfonso highlighted that in times of heightened surveillance, BSMPs within specific platforms are perceived as discouraging visibility to a greater degree: “Since the war started in Ukraine, we’ve started using Signal more because there’s a perception it’s more secure when I’m talking to my Ukrainian friends.” These examples demonstrate that people faced with the threat of institutional surveillance are required to make extra efforts to protect themselves and find BSMPs that sufficiently discourage visibility. Alfonso’s quote also shows how perception influences the operation of affordances (Davis, 2020); Signal is perceived to discourage visibility to a greater degree than BSMPs within other platforms.

As noted above, visibility on social media is strategic, relational, and outcome-driven (Pearce et al., 2018). Participants also highlighted how these different dimensions of visibility operate across different types of BSMPs.

Strategic Management

BSMPs allow for the strategic management of visibility, which refers to the control individuals have over drawing boundaries between what is seen and not seen (Pearce et al., 2018). This is facilitated by features and mechanisms like encryption and administrator control over groups with tools and screening questions. These features and mechanisms differ across different types of BSMPs.

Encryption features associated with private chats on mobile messaging services have been especially noted as a key element of visibility management (Gursky et al., 2022). Participants said that they believe these features allow them to manage the level of access larger institutions like corporations and governments have over what they post. Cesar perceived that “Signal is also rock solid encrypted,” ensuring that messages sent are “safe from not only individuals or companies, but [also] from government entities.” As noted above, Signal was perceived by some as discouraging visibility to a greater degree than other platforms. Furthermore, as encryption is typically, though not always, a default feature that is imposed on users of certain BSMPs, it can also be argued that encrypted BSMPs demand limited visibility from institutions.

Administrator control over groups with tools and screening questions is the other main way in which visibility is strategically managed on BSMPs. Screening questions are particularly associated with private Facebook groups. These are the questions that people are asked in order to confirm that they meet the requirements for being added to a private group. As these screening questions are an option provided by the platform to private group administrators, they can also be viewed as allowing limited visibility. Participants said that these screening questions help to ensure that only intended audiences are part of these groups and can see what they post. Becky reflected on the role of screening questions when describing her participation in a private alumni group for her high school: “we have the entry questions to make sure that you actually went to [the school].” However, answers to such questions can be falsified. In groups on sensitive topics, the screening process can be stricter as the social position of group members requires visibility to be discouraged to an especially strong degree. As Audrey said about a private Facebook group on infertility issues: “The groups were even stricter, asking you to provide your lab results, to make sure that your lab results actually were indicative of this condition.” This is another example of social position or cultural and institutional legitimacy influencing how different people experience the affordances of BSMPs differently. Those who are in groups about sensitive topics because of difficult life circumstances are required to put in additional effort to protect themselves and ensure that BSMPs sufficiently discourage visibility.

The degree to which users have control over strategically managing their visibility also varies across different types of BSMPs. This was reflected in how participants compared private chats within mobile messaging services, particularly those comprising of close friends and family, to interest- and hobby-based private Facebook groups and Discord servers. They emphasized that the former discourage visibility to a greater degree because individuals have more personal agency in the strategic management of visibility. Francesca made this comparison, stating that “I have less control over my private Facebook groups because [members of] those [groups] are admitted [to the group] by someone else.” In comparison, only she and her family members have control over who is allowed to participate in their small family group chat on WhatsApp.

Relationality

On BSMPs like private chats within mobile messaging services, the audience is often restricted to close friends and family. Participants stated that their closeness with these audiences contributes to the perception that these places discourage visibility. They believe that there is an implicit understanding and trust between interpersonal ties that anything posted within BSMPs will not be shared with unintended audiences. As Ravi said, “the personal connection, the depth and strength of relationships, meeting those people every day, knowing about their lives, knowing who they are. That gives me [a] certain confidence that anything I share with them is private.”

This sense of trust also exists in interest- and hobby-based private Facebook groups and Discord servers, albeit not to the same degree. In some groups, a culture of trust can develop over time. As Bette stated, there is a “culture of private groups where it’s just this understanding that people have. What you post in the group stays in the group.” Groups can also have established norms about this, especially if they contain marginalized people. “[In] groups around LGBTQ identity, there are norms about not screenshotting and sharing because this is dealing with people’s safety,” said Mallory. This, again, demonstrates contextual variation in how affordances operate. Technology is developed for people with social status or cultural and institutional legitimacy (Davis, 2020); conversely, those with marginalized identities are required to expend extra effort and establish norms so that BSMPs discourage visibility in a way that protects them from harm.

Outcomes

The low visibility associated with BSMPs leads to divergent outcomes. Echoing extant literature (Yeshua-Katz, 2021), some participants associated BSMPs with beneficial outcomes, particularly the exchange of social support. For example, Anders mentioned how BSMPs like private Discord servers allow him and others to acquire informational support, which refers to factual information or advice about an issue (Xu & Burleson, 2001). He stated that people share “advice for physical health issues or mental health issues,” and this is only possible on BSMPs because one “wouldn’t want to share [health-related information] publicly just because you don’t know who’s going to see it.”

The low visibility associated with BSMPs can also result in more undesirable outcomes as people feel emboldened to engage in negative activities in the absence of platform moderation and a broad audience to hold them accountable (Matamoros-Fernández, 2020; Semenzin & Bainotti, 2020). Diane touched on this while describing her experience on a private Facebook group for nannies. When she posted about wanting to meet “other Black, other POC [People of Color] nannies,” she was met with derisive and angry responses from White group members, saying “why don’t you want to hang out with us?” This led to her viewing such groups as being “very aggressive.” Meanwhile, participants who identify as women also described receiving unwanted contact through direct messages on social networking sites like Twitter and Instagram. May, who is a U.S.-based Thai student, recounted her experience on Twitter: “I just posted a selfie and a bunch of my friends liked it. And I just got a few message requests in my DMs from these random guys.” Diane and May are women of color, and their accounts exhibit how those with marginalized identities experience the affordances of BSMPs differently. For them, BSMPs discouraging visibility can lead to negative outcomes like unwanted contact and harassment.

RQ2: Personalization

Participants highlighted how BSMPs encourage receiver-oriented and sender-oriented personalization as these places allow them to receive relevant content from and send relevant content to specific audiences.

Receiver-Oriented Personalization

Participants stated that they receive relevant content on BSMPs. One reason for this is that the interpersonal ties they interact with, especially on private chats within mobile messaging services, are familiar with their interests. As Ravi stated, “closer friends or people that know me personally will put some thought into . . . will this piece of news or information be interesting to the other person?”

On other types of BSMPs like private groups and servers, they receive relevant content because they can easily self-select into groups and servers organized around shared interests. For example, KC described participating in private groups related to Ultimate Frisbee: “Everybody in that group knows some other large portion of the group through doing an activity that we all enjoy. So, that’s probably why it feels more relevant. Because we’re all there for a shared interest.” Of course, people can also self-select into public groups and servers organized around topics they are interested in. However, many participants stated that these public spaces feel cluttered and include off-topic posts. Furthermore, as noted above, if people want to discuss sensitive topics, they prefer to do so in private groups.

The absence of content ranking algorithms on many BSMPs also contributes to the belief that these places encourage receiver-oriented personalization. Participants underscored that they believe they have personal control and agency over curating the audience they receive content from in the absence of algorithms. For instance, Amy Allen said:

I do have a lot more control over what I see on Telegram. No one is suggesting you should follow this person. It’s not going to start autoplaying someone’s Reels [a video format on Instagram] if I accidentally click something wrong. It’s not constantly trying to suggest new crap for me to keep me more engaged.

Johny also touched on this idea and said that they prefer engaging with content presented chronologically because “I don’t need the algorithm to decide for me.” These examples highlight how perception influences the operation of affordances. Echoing existing literature (Swart, 2021), many participants perceived that algorithms take away individual control over personalizing one’s information environment; conversely, non-algorithmic BSMPs were perceived as being associated with control and personalization.

Sender-Oriented Personalization

Participants revealed that BSMPs also encourage sender-oriented personalization as they make it easy to engage in targeted or directed communication. As Aryan said, “it is easier to reach a specific target audience within private spaces than in public spaces.” Unlike public social media feeds, where users can simply post content without directing it toward a specific individual or group, most BSMPs require people to send their posts or messages to a specific chat or group. Thus, BSMPs can also be viewed as demanding sender-oriented personalization.

One reason that BSMPs are associated with sender-oriented personalization is that these places allow users to easily self-select into specific groups and reach specific audiences. For instance, Mallory reflected on how private Facebook groups allow her to specifically reach foster parents for advice and why this is important for her:

These are people who are likely to share a perspective or find similar things important. So, if I asked a large group of foster parents about a parenting challenge I am having, they’re not going to come at it with advice from somebody who’s not a parent or is a parent of a biological kid.

This is another example of how people who need to use BSMPs to discuss sensitive topics experience the affordances of these places differently. They need to reach specific audiences to get specialized help and therefore have more to lose if BSMPs do not encourage sender-oriented personalization to an adequate degree. It also illustrates how visibility is a root affordance that underpins personalization (Flyverbom et al., 2016). The ability to restrict group membership to intended audiences through visibility management mechanisms allows users to receive content from and send content to specific audiences.

As with receiver-oriented personalization, participants noted that another reason they perceive they have control over reaching a specific audience on BSMPs is that content posted within these places is typically arranged chronologically rather than via algorithmic curation. Zoe made this argument while discussing private chats on WeChat:

For WeChat, I’m 100 percent sure people are going to see it, even if they don’t like it. I know that because there’s no algorithm. They don’t really block your content because it’s just for the people in your contacts.

The degree to which BSMPs encourage sender-oriented personalization also varies across different types of BSMPs. Some participants highlighted how private communities within channel-based messaging services like Discord and Slack strongly encourage sender-oriented personalization as they allow for the segmentation of audiences into different topic-based channels within a single private server or workspace. As Andrew noted, “the servers are broken out into different conversations, by topic, by events.” He said that this resulted in him valuing “the customizability of the way I can post in the private spaces of Discord.”

Overall, BSMPs encourage receiver- and sender-oriented personalization; however, the degree of encouragement can vary, based on how people are socially situated and the type of BSMP they are using.

RQ3: Synchronicity

As noted above, synchronicity is typically associated with immediate feedback. Participants compared different types of BSMPs and highlighted how chat- and messaging-based BSMPs are especially associated with such immediacy. As May said, “I think in group chats, I expect the response to be really fast, within 20, maybe 30–40 minutes.” However, participants also emphasized that BSMPs more generally feel synchronous because they facilitate continuous and uninterrupted conversations, regardless of the transmission time between messages. For example, Ben said that interactions on BSMPs feel like a “continuous conversation,” and “even if there are large gaps, chances are, the next message in that channel, even if it’s 36 hours later than the previous message, is probably continuing that thread or something related to it.”

Participants also highlighted how technological features as well as social and relational norms and expectations impact the degree to which different types of BSMPs encourage synchronicity.

Technological Features

The presence function, which allows users to see if an interactant is online and/or has read their message, contributes to private chats within mobile messaging services being viewed as especially synchronous. Echoing extant literature (Mols & Pridmore, 2021), participants said that this is because the feature encourages immediate responses. As Bette put it, “[people are] seeing [your messages] often. You can see who’s viewed [your messages] based on what privacy notifications they have.” It can also be argued that by showing users when a message has been viewed, this feature requests synchronicity as it persuades users to respond with immediacy.

Another technological factor that impacts synchronicity is the absence of content ranking algorithms. Like they did with personalization, participants emphasized the perception that the absence of algorithms results in users having more control over the content they see on BSMPs. This makes it easier to encounter and respond to messages in a timley manner. As Amy Allen said, “any message that you post [on Telegram], it’s still going to be there. Even if you read through and then go off and do your own thing, and then come back later. The algorithm doesn’t just disappear it.”

To be sure, not every participant believed that it is easy to sustain continuous conversations on BSMPs. Aligning with existing literature (Ling & Lai, 2016; Yeshua-Katz, 2021), some participants mentioned how conversation threads can be difficult to follow in large groups due to the sheer volume of messages. For example, KC said that “sometimes it feels like [content] is lost in the large groups” because “if someone posts about something that’s gonna [sic] take place with all the details, and then 30 people react or respond to it,” it becomes “difficult to scroll through and find that information.” However, overall, features like the presence function and the lack of content ranking algorithms contribute to BSMPs being viewed as synchronous.

Norms and Expectations

Norms and expectations surrounding response times also shape perceptions of synchronicity, echoing Davis’s (2020) argument that social and cultural norms influence how people experience affordances. As these norms and expectations differ across relationships and groups, people view BSMPs as more or less synchronous based on who they are interacting with. Some participants said that the close interpersonal ties they interact with via mobile messaging expect near-immediate responses. As Matthew noted, “if somebody sent something to me, I have a responsibility to respond to it.” Some participants also said that their familiarity with these interpersonal ties means that both parties know when to expect a response. Even if responses take some time, it still feels like a continuous conversation as long as these expectations are not violated. Manaal reflected on this idea:

When I’m reaching out to someone on private chat, I already have an idea of how long they take to respond to messages on a general basis. So, I already have that expectation in mind. Okay, this person takes an hour, this person takes three business days. I’ll hear back soon enough.

Meanwhile, expectations regarding response times in interest- and hobby-based groups are primarily informed by the size and culture of the group. In larger groups, there is less familiarity between people, and therefore, there are no set expectations. However, in smaller groups, people generally expect quick responses. Bette discussed this within the context of private Facebook groups:

If it’s either something hyperlocal, I know them in person, or it’s a small group, under 100 people, then [there is] an expectation that people are going to see it and engage with it, versus if it’s a more impersonal, bigger group.

Overall, BSMPs encourage synchronicity. The strength of this encouragement varies across different relationships, groups, and BSMPs, with private mobile messaging chats containing close friends and family viewed as especially synchronous.

Limitations

As with any study, there are limitations. First, the voluntary convenience sample consisted of many people who are highly educated, technologically skilled, and experienced in using BSMPs, limiting generalizability. While there was an effort to diversify the sample in terms of cultural background, gender, age, and profession, this was not done systematically, and the sample could have been more diverse. Systematically sampling for diverse identities, backgrounds, and experience levels with using social media can help future researchers further examine contextual variation in how the affordances of BSMPs operate. Second, my choice to conduct interviews limits the analysis to a focus on user perspectives and the bids users place on BSMPs. Combining interviews with methods like the walkthrough method, where researchers focus on how a technology’s interface and underlying infrastructure guides user experience (Light et al., 2018), would help future researchers more fully examine both sides of the user-technology relationship. For this study, I chose to rely on interviews as it allowed me to feasibly cover the use of diverse social media platforms and develop a systematic framework for understanding BSMPs across multiple platforms. Third, while this focus on multiple platforms facilitates theoretical durability, it may have led to culturally- and contextually-specific understandings of particular platforms being overlooked. Finally, it is important to note that on certain BSMPs like private Facebook groups, there is algorithmic ranking of content; however, participants stated that this is not noticeable, and they generally associate BSMPs, including private Facebook groups, with the absence of algorithms.

Discussion and Conclusion

Interviews with users of diverse BSMPs reveal that these places are used to communicate with interpersonal ties and others with shared interests. In terms of affordances, BSMPs discourage visibility as they are considered private and difficult to access due to visibility management mechanisms and the presence of known audiences. BSMPs encourage receiver-oriented personalization because users believe they have control over receiving relevant content from only these known audiences, in the absence of algorithms. BSMPs also encourage sender-oriented personalization as they make it easy for users to directly reach these audiences. Finally, BSMPs encourage synchronicity as they are associated with facilitating continuous conversations among familiar interactants, in the absence of algorithms. The strength of encouragement or discouragement for these affordances varies across different types of BSMPs. Furthermore, broader social conditions also shape how these affordances operate. Particularly, people who need to use BSMPs to discuss sensitive topics expend additional effort so that these affordances operate in a way that keeps them protected.

Contribution to Social Media Research and Theorizing

Through identifying the three main affordances associated with BSMPs and outlining how and for whom and under what circumstances they operate across different types of BSMPs, I present a systematic framework for understanding low visibility places within social media, contributing to social media research and theorizing. This framework can facilitate future research on BSMPs across diverse social media platforms as it highlights the common affordances associated with these places at the same time as recognizing variations across different types of BSMPs. Even as BSMPs within different platforms may have unique characteristics, scholars can draw on this framework and systematically highlight these characteristics through analyzing how visibility, personalization, and synchronicity affordances operate on the BSMP they are studying. This also helps scholars avoid the practice of generating new typologies of affordances each time a new BSMP or social media platform gains popularity (Evans et al., 2017). To be sure, the emergence of certain BSMPs and social media platforms may require scholars to identify new affordances. In such cases, my framework can serve as a comparison point to justify the need for identifying new affordances, facilitating more systematic and durable research. Furthermore, drawing on this framework will help social media scholars better clarify their object of study (Malhotra, 2023). At present, studies of BSMPs are often framed as studies of entire platforms. For instance, some scholars discuss how they study private social media platforms like WhatsApp (Swart et al., 2018). This can result in a lack of conceptual clarity as such platforms often also include public groups and channels associated with broad audiences (Johns et al., 2023). Clarifying that one is focusing on BSMPs, which exist within broader platforms, rather than entire platforms, can help address this issue.

Beyond the utility of this framework, my empirical findings also contribute new insights regarding the use of BSMPs. First, much of the extant literature on BSMPs focuses on specific mobile messaging services or channels like private Facebook groups. This study also presents insights into understudied BSMPs like private Discord servers, highlighting how they are used to discuss common interests and sensitive topics. Second, not much attention has been paid to the absence of algorithmic curation on BSMPs. I highlight how this plays a key role in the way BSMPs are perceived. Participants distinguished between algorithmic and chronological timelines, with the former being perceived negatively, perhaps due to public discourse linking algorithms to polarization and misinformation (Fournier, 2021). This seems to result in chronological BSMPs conversely being viewed positively, particularly as places associated with control over personalizing one’s information environment and synchronously responding to messages. Third, these findings also demonstrate the need for more systematic research on how public discourse about social media impacts user perceptions of social media affordances. This includes examining how these perceptions are informed by the way social media companies publicly market their own platforms. For instance, how does Signal marketing itself as private and encrypted inform the perception that it discourages visibility to a greater degree than BSMPs within other platforms? Fourth, while much of the literature on BSMPs focuses on the positive outcomes associated with these places such as relational maintenance and social support exchange, this study also highlights how women and people of color can face harassment within BSMPs. This reinforces the argument that the harms associated with social media disproportionately impact marginalized communities (Bailey, 2021; Dhoest & Szulc, 2016).

Contribution to Affordances-Based Research

This study also enhances our understanding of affordances-based research on social media. First, my findings demonstrate how adopting an affordances perspective can help scholars avoid a technologically or socially deterministic lens. I highlight how technological features and social and relational factors act in concert to inform how BSMPs are perceived and used. For example, users believe that BSMPs discourage visibility because of technological features like encryption and the trust they have in the people they interact with within these places. Second, much of the affordances-based scholarship on social media stops short at identifying affordances associated with specific platforms; less attention is paid to the way individuals’ social positions shape how these affordances operate (Davis, 2020). I demonstrate the utility of focusing on this aspect as I highlight how those belonging to marginalized communities and/or using BSMPs to discuss sensitive topics expend extra effort so that BSMPs discourage visibility and encourage personalization in a way that protects them from harm.

These findings also enhance our understanding of the specific affordances associated with BSMPs, particularly how they operate in the current social media ecology. This opens up avenues for future research. In terms of visibility, much research has focused on how BSMPs allow users to strategically manage visibility because of encryption (Gursky et al., 2022) or administrator control over groups (Malinen, 2021; Swart et al., 2019). My findings highlight how trust in audiences on BSMPs to not share content outside of these places also impacts perceptions of visibility. Future research can look further into this sense of trust, including whether it is justified; how it differs across various types of relationships, groups, and BSMPs; and the factors that result in its breach. In terms of personalization, while the creation of algorithmic filter bubbles on social media has been a topic of contention (Fletcher & Nielsen, 2017; Pariser, 2011), we know less about whether the absence of content ranking algorithms results in users creating their own information filter bubbles. My findings suggest that this might be the case since users associate BSMPs with control over only receiving content that they view as personally relevant. Finally, future research can also look further into how this absence of algorithms informs perceptions of synchronicity on social media.

In conclusion, this study presents a systematic examination of BSMPs, including identifying the main affordances associated with these places and outlining how these affordances operate across different types of BSMPs. This helps to enhance our understanding of low visibility places within the current social media ecology and lays the theoretical groundwork for future scholarship on these places across a variety of platforms and contexts.

Supplemental Material

sj-docx-1-sms-10.1177_20563051241285777 – Supplemental material for “What You Post in the Group Stays in the Group”: Examining the Affordances of Bounded Social Media Places

Supplemental material, sj-docx-1-sms-10.1177_20563051241285777 for “What You Post in the Group Stays in the Group”: Examining the Affordances of Bounded Social Media Places by Pranav Malhotra in Social Media + Society